diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/added_tokens.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d3d3a1b74d87e381e471f7b33784015d2dc0ea --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/added_tokens.json @@ -0,0 +1,13 @@ +{ + "<|assistant|>": 32001, + "<|endoftext|>": 32000, + "<|end|>": 32007, + "<|placeholder1|>": 32002, + "<|placeholder2|>": 32003, + "<|placeholder3|>": 32004, + "<|placeholder4|>": 32005, + "<|placeholder5|>": 32008, + "<|placeholder6|>": 32009, + "<|system|>": 32006, + "<|user|>": 32010 +} diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/config.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/config.json new file mode 100644 index 0000000000000000000000000000000000000000..2056b063e11084a658d8f0047b62ff03ad341faf --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/config.json @@ -0,0 +1,181 @@ +{ + "_name_or_path": "/cm/archive/namnv78/checkpoints/phi35-siglip224/pft", + "architectures": [ + "LlavaPhiForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "auto_map": { + "AutoConfig": "configuration_phi3.Phi3Config", + "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM" + }, + "balance_loss_coef": 0.01, + "bos_token_id": 1, + "clip_smoe": true, + "dropout": false, + "embd_pdrop": 0.0, + "eos_token_id": 32000, + "freeze_mm_mlp_adapter": false, + "hidden_act": "silu", + "hidden_size": 3072, + "image_aspect_ratio": "pad", + "initializer_range": 0.02, + "intermediate_size": 8192, + "local_rank": 0, + "loss1": "balanceloss", + "loss2": "zloss", + "luna": false, + "max_position_embeddings": 131072, + "mlp_smoe": true, + "mm_hidden_size": 1152, + "mm_patch_merge_type": "flat", + "mm_projector_lr": null, + "mm_projector_type": "moe", + "mm_use_im_patch_token": false, + "mm_use_im_start_end": false, + "mm_vision_select_feature": "patch", + "mm_vision_select_layer": -2, + "mm_vision_tower": "google/siglip-so400m-patch14-224", + "model_type": "llava_phi", + "moe_name": "competesmoe", + "normalization": true, + "num_attention_heads": 32, + "num_experts": 4, + "num_hidden_layers": 32, + "num_key_value_heads": 32, + "num_layers": 3, + "num_selected": 2, + "number_of_previous_tokens": 2, + "original_max_position_embeddings": 4096, + "pad_token_id": 32000, + "rate_compete": 0.2, + "rate_flip": 0.07, + "resid_pdrop": 0.0, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "long_factor": [ + 1.0800000429153442, + 1.1100000143051147, + 1.1399999856948853, + 1.340000033378601, + 1.5899999141693115, + 1.600000023841858, + 1.6200000047683716, + 2.620000123977661, + 3.2300000190734863, + 3.2300000190734863, + 4.789999961853027, + 7.400000095367432, + 7.700000286102295, + 9.09000015258789, + 12.199999809265137, + 17.670000076293945, + 24.46000099182129, + 28.57000160217285, + 30.420001983642578, + 30.840002059936523, + 32.590003967285156, + 32.93000411987305, + 42.320003509521484, + 44.96000289916992, + 50.340003967285156, + 50.45000457763672, + 57.55000305175781, + 57.93000411987305, + 58.21000289916992, + 60.1400032043457, + 62.61000442504883, + 62.62000274658203, + 62.71000289916992, + 63.1400032043457, + 63.1400032043457, + 63.77000427246094, + 63.93000411987305, + 63.96000289916992, + 63.970001220703125, + 64.02999877929688, + 64.06999969482422, + 64.08000183105469, + 64.12000274658203, + 64.41000366210938, + 64.4800033569336, + 64.51000213623047, + 64.52999877929688, + 64.83999633789062 + ], + "short_factor": [ + 1.0, + 1.0199999809265137, + 1.0299999713897705, + 1.0299999713897705, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0699999332427979, + 1.0999999046325684, + 1.1099998950958252, + 1.1599998474121094, + 1.1599998474121094, + 1.1699998378753662, + 1.2899998426437378, + 1.339999794960022, + 1.679999828338623, + 1.7899998426437378, + 1.8199998140335083, + 1.8499997854232788, + 1.8799997568130493, + 1.9099997282028198, + 1.9399996995925903, + 1.9899996519088745, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0799996852874756, + 2.0899996757507324, + 2.189999580383301, + 2.2199995517730713, + 2.5899994373321533, + 2.729999542236328, + 2.749999523162842, + 2.8399994373321533 + ], + "type": "longrope" + }, + "rope_theta": 10000.0, + "router_loss_coef": 0.005, + "router_z_loss_coef": 0.001, + "scales": [ + 1, + 3 + ], + "sliding_window": 262144, + "sparse_upcycling": true, + "strategy_train": "base", + "tie_word_embeddings": false, + "tokenizer_model_max_length": 2048, + "tokenizer_padding_side": "right", + "topk_max": 2, + "topk_min": 1, + "torch_dtype": "bfloat16", + "training": true, + "transformers_version": "4.43.0", + "tune_mm_mlp_adapter": false, + "use_cache": true, + "use_mm_proj": true, + "vocab_size": 32064, + "warm_up": 0.05 +} diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/generation_config.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..dad5c4578f0dc5969b38755d095fc30c368bb54a --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/generation_config.json @@ -0,0 +1,12 @@ +{ + "_from_model_config": true, + "bos_token_id": 1, + "do_sample": true, + "eos_token_id": [ + 32007, + 32001, + 32000 + ], + "pad_token_id": 32000, + "transformers_version": "4.43.0" +} diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00001-of-00003.safetensors b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..de9d1162f1c218739cb7391227d07dee60ce2db9 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3d707f92ad4e42e71a04d1469f4d3e10f6bde1ba6d67686088896f4a65fd83af +size 4972489328 diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00002-of-00003.safetensors b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cf007715d0917cdf7cccf5c7b5f655c6e0f2350e --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c18871bc23ce26c6894257da39ed6f2a23700d04d44e78cef710d0f3559b0f15 +size 4985754844 diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00003-of-00003.safetensors b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..38ff4955c81562872e381e7195e499174b94e818 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:02e754ddcf4544dcf3b08eca2a3d118e9c3a7daa9a08b1f2f64dc2e4bdb50a09 +size 248943552 diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model.safetensors.index.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..a900cb68b39c8fe6eedc011196340060a750c9c8 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/model.safetensors.index.json @@ -0,0 +1,1033 @@ +{ + "metadata": { + "total_size": 10207040684 + }, + "weight_map": { + "lm_head.weight": "model-00003-of-00003.safetensors", + "model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.gate.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.norm.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.position_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.prob_flips": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors" + } +} diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/special_tokens_map.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..3e4d5a5bc1cb51753cc9ae0305ece0da60052b10 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/tokenizer.model b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/tokenizer_config.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d579bb0b91b24b214ea3c2e487e27a65017cdc4a --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/tokenizer_config.json @@ -0,0 +1,132 @@ +{ + "add_bos_token": false, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": false + }, + "32000": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "32001": { + "content": "<|assistant|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32002": { + "content": "<|placeholder1|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32003": { + "content": "<|placeholder2|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32004": { + "content": "<|placeholder3|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32005": { + "content": "<|placeholder4|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32006": { + "content": "<|system|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32007": { + "content": "<|end|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32008": { + "content": "<|placeholder5|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32009": { + "content": "<|placeholder6|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32010": { + "content": "<|user|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "chat_template": "{% for message in messages %}{% if message['role'] == 'system' and message['content'] %}{{'<|system|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'user' %}{{'<|user|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>\n' + message['content'] + '<|end|>\n'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}", + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "legacy": false, + "model_max_length": 2048, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +} diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/trainer_state.json b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..2cc98d093570b09a3b3e9682241643f065c94b9d --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/trainer_state.json @@ -0,0 +1,124783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999398785546805, + "eval_steps": 500, + "global_step": 8316, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "auxiliary_loss_clip": 0.04572364, + "auxiliary_loss_mlp": 0.02183614, + "balance_loss_clip": 2.29309511, + "balance_loss_mlp": 1.82579517, + "epoch": 0.00012024289063909097, + "flos": 24932483919360.0, + "grad_norm": 40.00472956498374, + "language_loss": 2.5840292, + "learning_rate": 0.0, + "loss": 1.9026171, + "num_input_tokens_seen": 20375, + "step": 1, + "time_per_iteration": 15.790460586547852 + }, + { + "auxiliary_loss_clip": 0.03044211, + "auxiliary_loss_mlp": 0.01400663, + "balance_loss_clip": 1.52520752, + "balance_loss_mlp": 1.16701245, + "epoch": 0.00024048578127818193, + "flos": 30664624377600.0, + "grad_norm": 54.57939535925913, + "language_loss": 1.88843393, + "learning_rate": 5.021476677069823e-07, + "loss": 1.93288279, + "num_input_tokens_seen": 39035, + "step": 2, + "time_per_iteration": 2.686408281326294 + }, + { + "auxiliary_loss_clip": 0.03042922, + "auxiliary_loss_mlp": 0.01417982, + "balance_loss_clip": 1.52280807, + "balance_loss_mlp": 1.17899108, + "epoch": 0.0003607286719172729, + "flos": 19026227969280.0, + "grad_norm": 40.519042837185054, + "language_loss": 1.61536074, + "learning_rate": 7.958852231401551e-07, + "loss": 1.65996981, + "num_input_tokens_seen": 57600, + "step": 3, + "time_per_iteration": 2.532782554626465 + }, + { + "auxiliary_loss_clip": 0.03055763, + "auxiliary_loss_mlp": 0.01469517, + "balance_loss_clip": 1.53424227, + "balance_loss_mlp": 1.22003627, + "epoch": 0.00048097156255636386, + "flos": 19316314206720.0, + "grad_norm": 48.47971477893619, + "language_loss": 1.64210403, + "learning_rate": 1.0042953354139647e-06, + "loss": 1.68735695, + "num_input_tokens_seen": 76465, + "step": 4, + "time_per_iteration": 2.560455083847046 + }, + { + "auxiliary_loss_clip": 0.03037919, + "auxiliary_loss_mlp": 0.01461923, + "balance_loss_clip": 1.52974153, + "balance_loss_mlp": 1.21721053, + "epoch": 0.0006012144531954548, + "flos": 13991264893440.0, + "grad_norm": 55.10407378092321, + "language_loss": 1.93294644, + "learning_rate": 1.1659507774310057e-06, + "loss": 1.97794497, + "num_input_tokens_seen": 94350, + "step": 5, + "time_per_iteration": 2.7617924213409424 + }, + { + "auxiliary_loss_clip": 0.03074488, + "auxiliary_loss_mlp": 0.01477083, + "balance_loss_clip": 1.53261709, + "balance_loss_mlp": 1.23275197, + "epoch": 0.0007214573438345458, + "flos": 23148988225920.0, + "grad_norm": 44.71555883813792, + "language_loss": 1.6105237, + "learning_rate": 1.2980328908471373e-06, + "loss": 1.65603936, + "num_input_tokens_seen": 114595, + "step": 6, + "time_per_iteration": 2.886570930480957 + }, + { + "auxiliary_loss_clip": 0.0337483, + "auxiliary_loss_mlp": 0.0154999, + "balance_loss_clip": 1.71423578, + "balance_loss_mlp": 1.29059064, + "epoch": 0.0008417002344736367, + "flos": 67663246170240.0, + "grad_norm": 4.667726499213422, + "language_loss": 0.81454074, + "learning_rate": 1.4097067265369432e-06, + "loss": 0.86378896, + "num_input_tokens_seen": 179590, + "step": 7, + "time_per_iteration": 3.325885057449341 + }, + { + "auxiliary_loss_clip": 0.03041844, + "auxiliary_loss_mlp": 0.01476645, + "balance_loss_clip": 1.53275275, + "balance_loss_mlp": 1.21495724, + "epoch": 0.0009619431251127277, + "flos": 21281381504640.0, + "grad_norm": 42.72795598935893, + "language_loss": 1.58787751, + "learning_rate": 1.506443003120947e-06, + "loss": 1.63306236, + "num_input_tokens_seen": 195090, + "step": 8, + "time_per_iteration": 2.9141995906829834 + }, + { + "auxiliary_loss_clip": 0.03044312, + "auxiliary_loss_mlp": 0.01439687, + "balance_loss_clip": 1.53356647, + "balance_loss_mlp": 1.1965003, + "epoch": 0.0010821860157518186, + "flos": 23331342597120.0, + "grad_norm": 17.978848764162976, + "language_loss": 1.47923779, + "learning_rate": 1.5917704462803102e-06, + "loss": 1.52407789, + "num_input_tokens_seen": 211635, + "step": 9, + "time_per_iteration": 2.8643643856048584 + }, + { + "auxiliary_loss_clip": 0.03052539, + "auxiliary_loss_mlp": 0.01465176, + "balance_loss_clip": 1.53505063, + "balance_loss_mlp": 1.21703017, + "epoch": 0.0012024289063909096, + "flos": 17010166337280.0, + "grad_norm": 13.534145569671143, + "language_loss": 1.5313127, + "learning_rate": 1.6680984451379884e-06, + "loss": 1.57648993, + "num_input_tokens_seen": 224705, + "step": 10, + "time_per_iteration": 2.861268997192383 + }, + { + "auxiliary_loss_clip": 0.0304412, + "auxiliary_loss_mlp": 0.01436349, + "balance_loss_clip": 1.53256202, + "balance_loss_mlp": 1.18896639, + "epoch": 0.0013226717970300007, + "flos": 21288133261440.0, + "grad_norm": 13.871834752796476, + "language_loss": 1.32458735, + "learning_rate": 1.7371455188905097e-06, + "loss": 1.36939192, + "num_input_tokens_seen": 244635, + "step": 11, + "time_per_iteration": 2.816436529159546 + }, + { + "auxiliary_loss_clip": 0.03031118, + "auxiliary_loss_mlp": 0.01411415, + "balance_loss_clip": 1.52330637, + "balance_loss_mlp": 1.17032588, + "epoch": 0.0014429146876690916, + "flos": 27237884935680.0, + "grad_norm": 13.617095245712934, + "language_loss": 1.25506067, + "learning_rate": 1.8001805585541196e-06, + "loss": 1.29948592, + "num_input_tokens_seen": 265765, + "step": 12, + "time_per_iteration": 2.8535280227661133 + }, + { + "auxiliary_loss_clip": 0.03007342, + "auxiliary_loss_mlp": 0.01428202, + "balance_loss_clip": 1.51923656, + "balance_loss_mlp": 1.19741249, + "epoch": 0.0015631575783081825, + "flos": 19062174504960.0, + "grad_norm": 6.77282829248454, + "language_loss": 1.29420471, + "learning_rate": 1.8581671739548328e-06, + "loss": 1.3385601, + "num_input_tokens_seen": 283500, + "step": 13, + "time_per_iteration": 2.760094404220581 + }, + { + "auxiliary_loss_clip": 0.03032659, + "auxiliary_loss_mlp": 0.0142465, + "balance_loss_clip": 1.52620697, + "balance_loss_mlp": 1.19080901, + "epoch": 0.0016834004689472734, + "flos": 48139473985920.0, + "grad_norm": 6.393212279907861, + "language_loss": 1.13536727, + "learning_rate": 1.9118543942439254e-06, + "loss": 1.17994034, + "num_input_tokens_seen": 305685, + "step": 14, + "time_per_iteration": 3.9871129989624023 + }, + { + "auxiliary_loss_clip": 0.03005914, + "auxiliary_loss_mlp": 0.01402078, + "balance_loss_clip": 1.5204494, + "balance_loss_mlp": 1.16995335, + "epoch": 0.0018036433595863645, + "flos": 34970026314240.0, + "grad_norm": 5.608063287018698, + "language_loss": 1.12691212, + "learning_rate": 1.961836000571161e-06, + "loss": 1.17099202, + "num_input_tokens_seen": 327340, + "step": 15, + "time_per_iteration": 3.870206832885742 + }, + { + "auxiliary_loss_clip": 0.03223239, + "auxiliary_loss_mlp": 0.01506455, + "balance_loss_clip": 1.68040586, + "balance_loss_mlp": 1.25468469, + "epoch": 0.0019238862502254555, + "flos": 59768284440960.0, + "grad_norm": 3.7874847240589076, + "language_loss": 0.64647704, + "learning_rate": 2.0085906708279293e-06, + "loss": 0.69377398, + "num_input_tokens_seen": 382710, + "step": 16, + "time_per_iteration": 3.202209949493408 + }, + { + "auxiliary_loss_clip": 0.02987394, + "auxiliary_loss_mlp": 0.01445679, + "balance_loss_clip": 1.52274573, + "balance_loss_mlp": 1.20497191, + "epoch": 0.0020441291408645466, + "flos": 20814543417600.0, + "grad_norm": 4.3424722229286745, + "language_loss": 1.16030765, + "learning_rate": 2.0525099325728135e-06, + "loss": 1.20463824, + "num_input_tokens_seen": 400890, + "step": 17, + "time_per_iteration": 2.749842405319214 + }, + { + "auxiliary_loss_clip": 0.03178932, + "auxiliary_loss_mlp": 0.01493771, + "balance_loss_clip": 1.67110932, + "balance_loss_mlp": 1.24505281, + "epoch": 0.0021643720315036373, + "flos": 63857001582720.0, + "grad_norm": 3.5334806388801017, + "language_loss": 0.72129905, + "learning_rate": 2.0939181139872922e-06, + "loss": 0.76802599, + "num_input_tokens_seen": 462605, + "step": 18, + "time_per_iteration": 3.2300312519073486 + }, + { + "auxiliary_loss_clip": 0.02950178, + "auxiliary_loss_mlp": 0.01387951, + "balance_loss_clip": 1.50940979, + "balance_loss_mlp": 1.15468216, + "epoch": 0.0022846149221427284, + "flos": 31284981192960.0, + "grad_norm": 5.035550768653259, + "language_loss": 1.01560211, + "learning_rate": 2.1330868934640175e-06, + "loss": 1.05898356, + "num_input_tokens_seen": 483280, + "step": 19, + "time_per_iteration": 2.921243906021118 + }, + { + "auxiliary_loss_clip": 0.03110088, + "auxiliary_loss_mlp": 0.01461645, + "balance_loss_clip": 1.65403223, + "balance_loss_mlp": 1.21903062, + "epoch": 0.002404857812781819, + "flos": 51083648161920.0, + "grad_norm": 3.5710994449972877, + "language_loss": 0.7643429, + "learning_rate": 2.170246112844971e-06, + "loss": 0.81006014, + "num_input_tokens_seen": 537620, + "step": 20, + "time_per_iteration": 3.0670604705810547 + }, + { + "auxiliary_loss_clip": 0.02930409, + "auxiliary_loss_mlp": 0.01394708, + "balance_loss_clip": 1.50349164, + "balance_loss_mlp": 1.15838766, + "epoch": 0.0025251007034209102, + "flos": 15815347309440.0, + "grad_norm": 4.284724244855363, + "language_loss": 1.01471734, + "learning_rate": 2.2055919496770983e-06, + "loss": 1.05796838, + "num_input_tokens_seen": 555760, + "step": 21, + "time_per_iteration": 2.798492670059204 + }, + { + "auxiliary_loss_clip": 0.02918437, + "auxiliary_loss_mlp": 0.01380698, + "balance_loss_clip": 1.50365365, + "balance_loss_mlp": 1.15257955, + "epoch": 0.0026453435940600014, + "flos": 37851857458560.0, + "grad_norm": 4.097825374468856, + "language_loss": 0.89520693, + "learning_rate": 2.2392931865974923e-06, + "loss": 0.93819827, + "num_input_tokens_seen": 578450, + "step": 22, + "time_per_iteration": 3.012450695037842 + }, + { + "auxiliary_loss_clip": 0.02860526, + "auxiliary_loss_mlp": 0.01363852, + "balance_loss_clip": 1.48607326, + "balance_loss_mlp": 1.14183712, + "epoch": 0.002765586484699092, + "flos": 21141976821120.0, + "grad_norm": 4.149210929528141, + "language_loss": 1.01616383, + "learning_rate": 2.271496085962064e-06, + "loss": 1.05840766, + "num_input_tokens_seen": 596145, + "step": 23, + "time_per_iteration": 2.8721323013305664 + }, + { + "auxiliary_loss_clip": 0.02839068, + "auxiliary_loss_mlp": 0.01397087, + "balance_loss_clip": 1.48214412, + "balance_loss_mlp": 1.16496277, + "epoch": 0.002885829375338183, + "flos": 20667381396480.0, + "grad_norm": 3.02194182241137, + "language_loss": 1.02574861, + "learning_rate": 2.3023282262611022e-06, + "loss": 1.06811011, + "num_input_tokens_seen": 614920, + "step": 24, + "time_per_iteration": 2.8006253242492676 + }, + { + "auxiliary_loss_clip": 0.02816373, + "auxiliary_loss_mlp": 0.01381298, + "balance_loss_clip": 1.47811663, + "balance_loss_mlp": 1.15203428, + "epoch": 0.003006072265977274, + "flos": 34823869873920.0, + "grad_norm": 3.2549345353257166, + "language_loss": 0.92539823, + "learning_rate": 2.3319015548620114e-06, + "loss": 0.96737492, + "num_input_tokens_seen": 636060, + "step": 25, + "time_per_iteration": 2.8593485355377197 + }, + { + "auxiliary_loss_clip": 0.02808954, + "auxiliary_loss_mlp": 0.01392666, + "balance_loss_clip": 1.48066258, + "balance_loss_mlp": 1.16092348, + "epoch": 0.003126315156616365, + "flos": 24422021118720.0, + "grad_norm": 2.4543619142790445, + "language_loss": 0.92846084, + "learning_rate": 2.3603148416618152e-06, + "loss": 0.97047704, + "num_input_tokens_seen": 655575, + "step": 26, + "time_per_iteration": 2.7814629077911377 + }, + { + "auxiliary_loss_clip": 0.02744167, + "auxiliary_loss_mlp": 0.01342963, + "balance_loss_clip": 1.4655726, + "balance_loss_mlp": 1.12695622, + "epoch": 0.003246558047255456, + "flos": 23622326674560.0, + "grad_norm": 2.3623098809716843, + "language_loss": 1.00851107, + "learning_rate": 2.3876556694204647e-06, + "loss": 1.04938257, + "num_input_tokens_seen": 675730, + "step": 27, + "time_per_iteration": 2.809587240219116 + }, + { + "auxiliary_loss_clip": 0.02751591, + "auxiliary_loss_mlp": 0.01358826, + "balance_loss_clip": 1.46475768, + "balance_loss_mlp": 1.14024401, + "epoch": 0.003366800937894547, + "flos": 17820275725440.0, + "grad_norm": 2.6402324044526173, + "language_loss": 0.90640938, + "learning_rate": 2.414002061950908e-06, + "loss": 0.94751358, + "num_input_tokens_seen": 694605, + "step": 28, + "time_per_iteration": 2.72129487991333 + }, + { + "auxiliary_loss_clip": 0.02743113, + "auxiliary_loss_mlp": 0.01330561, + "balance_loss_clip": 1.46346879, + "balance_loss_mlp": 1.1094991, + "epoch": 0.003487043828533638, + "flos": 24426115269120.0, + "grad_norm": 2.474830590718488, + "language_loss": 0.996337, + "learning_rate": 2.4394238264681557e-06, + "loss": 1.03707373, + "num_input_tokens_seen": 714340, + "step": 29, + "time_per_iteration": 2.780243158340454 + }, + { + "auxiliary_loss_clip": 0.02725712, + "auxiliary_loss_mlp": 0.01312025, + "balance_loss_clip": 1.47081804, + "balance_loss_mlp": 1.10069132, + "epoch": 0.003607286719172729, + "flos": 26140311002880.0, + "grad_norm": 2.133810537621703, + "language_loss": 0.99522847, + "learning_rate": 2.4639836682781433e-06, + "loss": 1.03560591, + "num_input_tokens_seen": 734470, + "step": 30, + "time_per_iteration": 2.8030662536621094 + }, + { + "auxiliary_loss_clip": 0.02712399, + "auxiliary_loss_mlp": 0.01301122, + "balance_loss_clip": 1.46409667, + "balance_loss_mlp": 1.09484243, + "epoch": 0.00372752960981182, + "flos": 20593082113920.0, + "grad_norm": 2.736263929512533, + "language_loss": 1.00104535, + "learning_rate": 2.487738122623307e-06, + "loss": 1.04118061, + "num_input_tokens_seen": 753380, + "step": 31, + "time_per_iteration": 2.8655929565429688 + }, + { + "auxiliary_loss_clip": 0.02649189, + "auxiliary_loss_mlp": 0.0132272, + "balance_loss_clip": 1.44053221, + "balance_loss_mlp": 1.11462831, + "epoch": 0.003847772500450911, + "flos": 22674608282880.0, + "grad_norm": 2.910129969304171, + "language_loss": 0.98779863, + "learning_rate": 2.510738338534912e-06, + "loss": 1.02751768, + "num_input_tokens_seen": 772105, + "step": 32, + "time_per_iteration": 2.7838339805603027 + }, + { + "auxiliary_loss_clip": 0.02540893, + "auxiliary_loss_mlp": 0.01281783, + "balance_loss_clip": 1.41621828, + "balance_loss_mlp": 1.0802716, + "epoch": 0.003968015391090002, + "flos": 17967796882560.0, + "grad_norm": 3.1496462304661024, + "language_loss": 1.02492142, + "learning_rate": 2.5330307420306648e-06, + "loss": 1.06314826, + "num_input_tokens_seen": 788955, + "step": 33, + "time_per_iteration": 2.752159357070923 + }, + { + "auxiliary_loss_clip": 0.02538433, + "auxiliary_loss_mlp": 0.01324525, + "balance_loss_clip": 1.41938305, + "balance_loss_mlp": 1.12005699, + "epoch": 0.004088258281729093, + "flos": 27304103658240.0, + "grad_norm": 2.578087635933368, + "language_loss": 0.88104582, + "learning_rate": 2.554657600279796e-06, + "loss": 0.91967535, + "num_input_tokens_seen": 810230, + "step": 34, + "time_per_iteration": 2.7929580211639404 + }, + { + "auxiliary_loss_clip": 0.02490683, + "auxiliary_loss_mlp": 0.01287906, + "balance_loss_clip": 1.41368747, + "balance_loss_mlp": 1.08811152, + "epoch": 0.004208501172368184, + "flos": 23258587599360.0, + "grad_norm": 2.1965142340839376, + "language_loss": 1.0324682, + "learning_rate": 2.5756575039679493e-06, + "loss": 1.07025409, + "num_input_tokens_seen": 829780, + "step": 35, + "time_per_iteration": 2.800037384033203 + }, + { + "auxiliary_loss_clip": 0.0246867, + "auxiliary_loss_mlp": 0.01291463, + "balance_loss_clip": 1.40215147, + "balance_loss_mlp": 1.09395742, + "epoch": 0.0043287440630072746, + "flos": 17312104062720.0, + "grad_norm": 1.9905237881472446, + "language_loss": 0.95030552, + "learning_rate": 2.5960657816942747e-06, + "loss": 0.98790681, + "num_input_tokens_seen": 848695, + "step": 36, + "time_per_iteration": 2.7629194259643555 + }, + { + "auxiliary_loss_clip": 0.02326254, + "auxiliary_loss_mlp": 0.01188091, + "balance_loss_clip": 1.46626616, + "balance_loss_mlp": 1.00498605, + "epoch": 0.004448986953646365, + "flos": 53092491160320.0, + "grad_norm": 1.366412633500659, + "language_loss": 0.60931289, + "learning_rate": 2.6159148575788668e-06, + "loss": 0.64445639, + "num_input_tokens_seen": 906730, + "step": 37, + "time_per_iteration": 3.2656307220458984 + }, + { + "auxiliary_loss_clip": 0.02405904, + "auxiliary_loss_mlp": 0.01271419, + "balance_loss_clip": 1.3968358, + "balance_loss_mlp": 1.09146094, + "epoch": 0.004569229844285457, + "flos": 13444165866240.0, + "grad_norm": 2.2608955261184343, + "language_loss": 0.98630786, + "learning_rate": 2.635234561171e-06, + "loss": 1.02308106, + "num_input_tokens_seen": 925125, + "step": 38, + "time_per_iteration": 2.7899603843688965 + }, + { + "auxiliary_loss_clip": 0.02356383, + "auxiliary_loss_mlp": 0.01261132, + "balance_loss_clip": 1.37771082, + "balance_loss_mlp": 1.07545209, + "epoch": 0.0046894727349245475, + "flos": 16209609966720.0, + "grad_norm": 2.9145499629055545, + "language_loss": 0.94142896, + "learning_rate": 2.6540523970949877e-06, + "loss": 0.97760415, + "num_input_tokens_seen": 939970, + "step": 39, + "time_per_iteration": 2.7888879776000977 + }, + { + "auxiliary_loss_clip": 0.02391579, + "auxiliary_loss_mlp": 0.01278151, + "balance_loss_clip": 1.39025998, + "balance_loss_mlp": 1.09495044, + "epoch": 0.004809715625563638, + "flos": 23914244505600.0, + "grad_norm": 2.470870959266708, + "language_loss": 0.9232226, + "learning_rate": 2.6723937805519533e-06, + "loss": 0.95991993, + "num_input_tokens_seen": 957470, + "step": 40, + "time_per_iteration": 4.669780969619751 + }, + { + "auxiliary_loss_clip": 0.0234275, + "auxiliary_loss_mlp": 0.01254006, + "balance_loss_clip": 1.37773645, + "balance_loss_mlp": 1.07280827, + "epoch": 0.00492995851620273, + "flos": 20773030273920.0, + "grad_norm": 2.254965800934219, + "language_loss": 0.92836779, + "learning_rate": 2.690282243737839e-06, + "loss": 0.96433544, + "num_input_tokens_seen": 976405, + "step": 41, + "time_per_iteration": 3.74471378326416 + }, + { + "auxiliary_loss_clip": 0.02299301, + "auxiliary_loss_mlp": 0.01244932, + "balance_loss_clip": 1.36131454, + "balance_loss_mlp": 1.07355666, + "epoch": 0.0050502014068418205, + "flos": 20338655103360.0, + "grad_norm": 3.674778332174137, + "language_loss": 0.99347931, + "learning_rate": 2.7077396173840807e-06, + "loss": 1.0289216, + "num_input_tokens_seen": 994690, + "step": 42, + "time_per_iteration": 2.7558839321136475 + }, + { + "auxiliary_loss_clip": 0.02268767, + "auxiliary_loss_mlp": 0.01244567, + "balance_loss_clip": 1.35890698, + "balance_loss_mlp": 1.07452703, + "epoch": 0.005170444297480911, + "flos": 25994872834560.0, + "grad_norm": 2.1058930121977206, + "language_loss": 0.92696834, + "learning_rate": 2.7247861909342594e-06, + "loss": 0.96210176, + "num_input_tokens_seen": 1015615, + "step": 43, + "time_per_iteration": 2.8375353813171387 + }, + { + "auxiliary_loss_clip": 0.02240443, + "auxiliary_loss_mlp": 0.01227788, + "balance_loss_clip": 1.35222507, + "balance_loss_mlp": 1.07129002, + "epoch": 0.005290687188120003, + "flos": 20954055841920.0, + "grad_norm": 2.5514427224750054, + "language_loss": 0.82938576, + "learning_rate": 2.7414408543044743e-06, + "loss": 0.86406815, + "num_input_tokens_seen": 1031255, + "step": 44, + "time_per_iteration": 2.7561404705047607 + }, + { + "auxiliary_loss_clip": 0.02222924, + "auxiliary_loss_mlp": 0.01221044, + "balance_loss_clip": 1.34987235, + "balance_loss_mlp": 1.06092238, + "epoch": 0.005410930078759093, + "flos": 15851401585920.0, + "grad_norm": 6.11282134646746, + "language_loss": 0.79099721, + "learning_rate": 2.7577212237113157e-06, + "loss": 0.82543689, + "num_input_tokens_seen": 1048295, + "step": 45, + "time_per_iteration": 2.7272212505340576 + }, + { + "auxiliary_loss_clip": 0.02214011, + "auxiliary_loss_mlp": 0.0121046, + "balance_loss_clip": 1.34914899, + "balance_loss_mlp": 1.06187701, + "epoch": 0.005531172969398184, + "flos": 21104988791040.0, + "grad_norm": 2.1797588772799874, + "language_loss": 1.0406853, + "learning_rate": 2.7736437536690466e-06, + "loss": 1.07492995, + "num_input_tokens_seen": 1067925, + "step": 46, + "time_per_iteration": 2.8948214054107666 + }, + { + "auxiliary_loss_clip": 0.02215213, + "auxiliary_loss_mlp": 0.01218536, + "balance_loss_clip": 1.34789562, + "balance_loss_mlp": 1.06804669, + "epoch": 0.005651415860037276, + "flos": 20844887431680.0, + "grad_norm": 2.031992455888148, + "language_loss": 1.07692027, + "learning_rate": 2.789223836941131e-06, + "loss": 1.11125779, + "num_input_tokens_seen": 1088060, + "step": 47, + "time_per_iteration": 2.777007579803467 + }, + { + "auxiliary_loss_clip": 0.02159535, + "auxiliary_loss_mlp": 0.01223479, + "balance_loss_clip": 1.33661234, + "balance_loss_mlp": 1.0767082, + "epoch": 0.005771658750676366, + "flos": 13260195383040.0, + "grad_norm": 2.468732641350934, + "language_loss": 1.08648944, + "learning_rate": 2.8044758939680847e-06, + "loss": 1.1203196, + "num_input_tokens_seen": 1104130, + "step": 48, + "time_per_iteration": 2.801955461502075 + }, + { + "auxiliary_loss_clip": 0.02147834, + "auxiliary_loss_mlp": 0.01217081, + "balance_loss_clip": 1.33605218, + "balance_loss_mlp": 1.07517421, + "epoch": 0.005891901641315457, + "flos": 24425396997120.0, + "grad_norm": 2.7199158149418388, + "language_loss": 1.01741719, + "learning_rate": 2.8194134530738863e-06, + "loss": 1.0510664, + "num_input_tokens_seen": 1122900, + "step": 49, + "time_per_iteration": 2.885178327560425 + }, + { + "auxiliary_loss_clip": 0.02121368, + "auxiliary_loss_mlp": 0.01200827, + "balance_loss_clip": 1.32858229, + "balance_loss_mlp": 1.06302154, + "epoch": 0.006012144531954548, + "flos": 23076197314560.0, + "grad_norm": 2.6912563098943387, + "language_loss": 0.90300727, + "learning_rate": 2.834049222568994e-06, + "loss": 0.93622923, + "num_input_tokens_seen": 1140250, + "step": 50, + "time_per_iteration": 2.8197646141052246 + }, + { + "auxiliary_loss_clip": 0.02109138, + "auxiliary_loss_mlp": 0.01218809, + "balance_loss_clip": 1.32541275, + "balance_loss_mlp": 1.07261038, + "epoch": 0.006132387422593639, + "flos": 22528775064960.0, + "grad_norm": 3.260139462695256, + "language_loss": 0.92549098, + "learning_rate": 2.848395155712969e-06, + "loss": 0.95877045, + "num_input_tokens_seen": 1160470, + "step": 51, + "time_per_iteration": 2.7752420902252197 + }, + { + "auxiliary_loss_clip": 0.021245, + "auxiliary_loss_mlp": 0.01209963, + "balance_loss_clip": 1.33251858, + "balance_loss_mlp": 1.07129884, + "epoch": 0.00625263031323273, + "flos": 27628340751360.0, + "grad_norm": 2.4537501773528514, + "language_loss": 0.97511727, + "learning_rate": 2.8624625093687977e-06, + "loss": 1.00846195, + "num_input_tokens_seen": 1177605, + "step": 52, + "time_per_iteration": 3.0144083499908447 + }, + { + "auxiliary_loss_clip": 0.02082903, + "auxiliary_loss_mlp": 0.01223839, + "balance_loss_clip": 1.32008564, + "balance_loss_mlp": 1.0905149, + "epoch": 0.006372873203871821, + "flos": 23110671392640.0, + "grad_norm": 2.1816866615196666, + "language_loss": 0.88829887, + "learning_rate": 2.876261897070029e-06, + "loss": 0.92136621, + "num_input_tokens_seen": 1197735, + "step": 53, + "time_per_iteration": 2.847827196121216 + }, + { + "auxiliary_loss_clip": 0.0209024, + "auxiliary_loss_mlp": 0.01199685, + "balance_loss_clip": 1.3206979, + "balance_loss_mlp": 1.06845903, + "epoch": 0.006493116094510912, + "flos": 22856028900480.0, + "grad_norm": 2.5397966964818215, + "language_loss": 0.9240942, + "learning_rate": 2.889803337127447e-06, + "loss": 0.95699346, + "num_input_tokens_seen": 1216335, + "step": 54, + "time_per_iteration": 2.7499425411224365 + }, + { + "auxiliary_loss_clip": 0.02061162, + "auxiliary_loss_mlp": 0.01203618, + "balance_loss_clip": 1.31091022, + "balance_loss_mlp": 1.07477689, + "epoch": 0.006613358985150003, + "flos": 23071708114560.0, + "grad_norm": 2.8876868419214663, + "language_loss": 0.84803748, + "learning_rate": 2.903096296321516e-06, + "loss": 0.88068533, + "num_input_tokens_seen": 1234480, + "step": 55, + "time_per_iteration": 2.8103010654449463 + }, + { + "auxiliary_loss_clip": 0.0205085, + "auxiliary_loss_mlp": 0.01211888, + "balance_loss_clip": 1.31119072, + "balance_loss_mlp": 1.08285618, + "epoch": 0.006733601875789094, + "flos": 26537662229760.0, + "grad_norm": 1.9275057701988028, + "language_loss": 0.91576374, + "learning_rate": 2.9161497296578907e-06, + "loss": 0.94839108, + "num_input_tokens_seen": 1253870, + "step": 56, + "time_per_iteration": 2.7818305492401123 + }, + { + "auxiliary_loss_clip": 0.02037442, + "auxiliary_loss_mlp": 0.01190075, + "balance_loss_clip": 1.30842376, + "balance_loss_mlp": 1.06752753, + "epoch": 0.006853844766428185, + "flos": 15523178083200.0, + "grad_norm": 2.3955964590105574, + "language_loss": 0.85921645, + "learning_rate": 2.928972116604173e-06, + "loss": 0.89149165, + "num_input_tokens_seen": 1270145, + "step": 57, + "time_per_iteration": 2.825333595275879 + }, + { + "auxiliary_loss_clip": 0.02033603, + "auxiliary_loss_mlp": 0.01192228, + "balance_loss_clip": 1.30945742, + "balance_loss_mlp": 1.070539, + "epoch": 0.006974087657067276, + "flos": 24243760897920.0, + "grad_norm": 2.014815905683033, + "language_loss": 1.01930165, + "learning_rate": 2.9415714941751377e-06, + "loss": 1.05156004, + "num_input_tokens_seen": 1291365, + "step": 58, + "time_per_iteration": 2.7875442504882812 + }, + { + "auxiliary_loss_clip": 0.02031446, + "auxiliary_loss_mlp": 0.01198263, + "balance_loss_clip": 1.30686939, + "balance_loss_mlp": 1.08067465, + "epoch": 0.007094330547706367, + "flos": 25772513690880.0, + "grad_norm": 2.0571724622564753, + "language_loss": 0.93616259, + "learning_rate": 2.9539554871897396e-06, + "loss": 0.96845967, + "num_input_tokens_seen": 1311535, + "step": 59, + "time_per_iteration": 2.8632333278656006 + }, + { + "auxiliary_loss_clip": 0.01987077, + "auxiliary_loss_mlp": 0.01175094, + "balance_loss_clip": 1.29431605, + "balance_loss_mlp": 1.05760169, + "epoch": 0.007214573438345458, + "flos": 21319015979520.0, + "grad_norm": 1.9615296153167685, + "language_loss": 0.97331339, + "learning_rate": 2.9661313359851253e-06, + "loss": 1.00493503, + "num_input_tokens_seen": 1329420, + "step": 60, + "time_per_iteration": 2.7719740867614746 + }, + { + "auxiliary_loss_clip": 0.01989364, + "auxiliary_loss_mlp": 0.01181125, + "balance_loss_clip": 1.29395902, + "balance_loss_mlp": 1.0640142, + "epoch": 0.007334816328984549, + "flos": 24937088192640.0, + "grad_norm": 1.9711097065001755, + "language_loss": 0.93934, + "learning_rate": 2.978105921839922e-06, + "loss": 0.97104484, + "num_input_tokens_seen": 1349965, + "step": 61, + "time_per_iteration": 2.796774387359619 + }, + { + "auxiliary_loss_clip": 0.01961948, + "auxiliary_loss_mlp": 0.0119399, + "balance_loss_clip": 1.29378796, + "balance_loss_mlp": 1.07592511, + "epoch": 0.00745505921962364, + "flos": 18510586277760.0, + "grad_norm": 2.3021488560624617, + "language_loss": 0.72184837, + "learning_rate": 2.9898857903302893e-06, + "loss": 0.75340778, + "num_input_tokens_seen": 1368915, + "step": 62, + "time_per_iteration": 2.776263475418091 + }, + { + "auxiliary_loss_clip": 0.01963942, + "auxiliary_loss_mlp": 0.01174981, + "balance_loss_clip": 1.29308486, + "balance_loss_mlp": 1.07355738, + "epoch": 0.007575302110262731, + "flos": 18477656484480.0, + "grad_norm": 2.739542028813954, + "language_loss": 0.88025796, + "learning_rate": 3.001477172817253e-06, + "loss": 0.91164714, + "num_input_tokens_seen": 1386805, + "step": 63, + "time_per_iteration": 2.7053489685058594 + }, + { + "auxiliary_loss_clip": 0.01966282, + "auxiliary_loss_mlp": 0.01178195, + "balance_loss_clip": 1.29289746, + "balance_loss_mlp": 1.06823671, + "epoch": 0.007695545000901822, + "flos": 24973178382720.0, + "grad_norm": 2.4972162272937695, + "language_loss": 0.96093184, + "learning_rate": 3.012886006241894e-06, + "loss": 0.99237657, + "num_input_tokens_seen": 1406190, + "step": 64, + "time_per_iteration": 2.8906729221343994 + }, + { + "auxiliary_loss_clip": 0.01925394, + "auxiliary_loss_mlp": 0.01200523, + "balance_loss_clip": 1.28544974, + "balance_loss_mlp": 1.09599996, + "epoch": 0.007815787891540913, + "flos": 21324223451520.0, + "grad_norm": 1.9637699439523113, + "language_loss": 0.88189352, + "learning_rate": 3.0241179513858383e-06, + "loss": 0.91315269, + "num_input_tokens_seen": 1425500, + "step": 65, + "time_per_iteration": 2.803135395050049 + }, + { + "auxiliary_loss_clip": 0.01916597, + "auxiliary_loss_mlp": 0.0116581, + "balance_loss_clip": 1.27807283, + "balance_loss_mlp": 1.06386209, + "epoch": 0.007936030782180003, + "flos": 21575777374080.0, + "grad_norm": 2.0971997248440615, + "language_loss": 0.87762749, + "learning_rate": 3.035178409737647e-06, + "loss": 0.90845156, + "num_input_tokens_seen": 1442950, + "step": 66, + "time_per_iteration": 3.7264137268066406 + }, + { + "auxiliary_loss_clip": 0.0189271, + "auxiliary_loss_mlp": 0.01171627, + "balance_loss_clip": 1.27621758, + "balance_loss_mlp": 1.07687938, + "epoch": 0.008056273672819095, + "flos": 20120785159680.0, + "grad_norm": 3.2586284163628916, + "language_loss": 0.88600874, + "learning_rate": 3.046072539090907e-06, + "loss": 0.91665208, + "num_input_tokens_seen": 1460915, + "step": 67, + "time_per_iteration": 4.569115161895752 + }, + { + "auxiliary_loss_clip": 0.01870741, + "auxiliary_loss_mlp": 0.01180816, + "balance_loss_clip": 1.26529706, + "balance_loss_mlp": 1.08406603, + "epoch": 0.008176516563458186, + "flos": 18333116156160.0, + "grad_norm": 2.210634482454444, + "language_loss": 1.0454824, + "learning_rate": 3.056805267986779e-06, + "loss": 1.07599795, + "num_input_tokens_seen": 1478385, + "step": 68, + "time_per_iteration": 2.76755428314209 + }, + { + "auxiliary_loss_clip": 0.01852933, + "auxiliary_loss_mlp": 0.01166041, + "balance_loss_clip": 1.26136875, + "balance_loss_mlp": 1.07272387, + "epoch": 0.008296759454097276, + "flos": 21872076664320.0, + "grad_norm": 2.248054652702724, + "language_loss": 0.95350039, + "learning_rate": 3.0673813091022194e-06, + "loss": 0.98369008, + "num_input_tokens_seen": 1497605, + "step": 69, + "time_per_iteration": 2.7368640899658203 + }, + { + "auxiliary_loss_clip": 0.01750643, + "auxiliary_loss_mlp": 0.01118684, + "balance_loss_clip": 1.36587405, + "balance_loss_mlp": 1.03743112, + "epoch": 0.008417002344736368, + "flos": 63408228036480.0, + "grad_norm": 1.3075316269289197, + "language_loss": 0.62127608, + "learning_rate": 3.0778051716749317e-06, + "loss": 0.64996928, + "num_input_tokens_seen": 1561150, + "step": 70, + "time_per_iteration": 3.384437084197998 + }, + { + "auxiliary_loss_clip": 0.01818636, + "auxiliary_loss_mlp": 0.01152841, + "balance_loss_clip": 1.24919868, + "balance_loss_mlp": 1.06319582, + "epoch": 0.008537245235375458, + "flos": 22966454286720.0, + "grad_norm": 2.0308293039081255, + "language_loss": 0.90311009, + "learning_rate": 3.0880811730470094e-06, + "loss": 0.93282485, + "num_input_tokens_seen": 1580605, + "step": 71, + "time_per_iteration": 2.7604939937591553 + }, + { + "auxiliary_loss_clip": 0.01706497, + "auxiliary_loss_mlp": 0.01094997, + "balance_loss_clip": 1.34610689, + "balance_loss_mlp": 1.01946592, + "epoch": 0.008657488126014549, + "flos": 61984046712960.0, + "grad_norm": 1.1709019509445167, + "language_loss": 0.58585769, + "learning_rate": 3.098213449401257e-06, + "loss": 0.61387265, + "num_input_tokens_seen": 1647535, + "step": 72, + "time_per_iteration": 3.261636972427368 + }, + { + "auxiliary_loss_clip": 0.01793517, + "auxiliary_loss_mlp": 0.01142827, + "balance_loss_clip": 1.24476933, + "balance_loss_mlp": 1.05623353, + "epoch": 0.00877773101665364, + "flos": 30296791152000.0, + "grad_norm": 6.464222330740392, + "language_loss": 0.98644197, + "learning_rate": 3.1082059657570015e-06, + "loss": 1.01580536, + "num_input_tokens_seen": 1666770, + "step": 73, + "time_per_iteration": 2.8573076725006104 + }, + { + "auxiliary_loss_clip": 0.01792565, + "auxiliary_loss_mlp": 0.01161617, + "balance_loss_clip": 1.25008321, + "balance_loss_mlp": 1.07979155, + "epoch": 0.00889797390729273, + "flos": 23514056104320.0, + "grad_norm": 3.036925251874556, + "language_loss": 0.96520907, + "learning_rate": 3.1180625252858496e-06, + "loss": 0.99475086, + "num_input_tokens_seen": 1685200, + "step": 74, + "time_per_iteration": 2.797562599182129 + }, + { + "auxiliary_loss_clip": 0.01764327, + "auxiliary_loss_mlp": 0.01146123, + "balance_loss_clip": 1.24038887, + "balance_loss_mlp": 1.06262922, + "epoch": 0.009018216797931822, + "flos": 23075838178560.0, + "grad_norm": 2.9692665309820736, + "language_loss": 0.80106759, + "learning_rate": 3.1277867780021663e-06, + "loss": 0.83017206, + "num_input_tokens_seen": 1701835, + "step": 75, + "time_per_iteration": 2.8285348415374756 + }, + { + "auxiliary_loss_clip": 0.01754783, + "auxiliary_loss_mlp": 0.01142356, + "balance_loss_clip": 1.23892689, + "balance_loss_mlp": 1.06496572, + "epoch": 0.009138459688570914, + "flos": 15918877284480.0, + "grad_norm": 2.239690515227744, + "language_loss": 0.9559927, + "learning_rate": 3.1373822288779824e-06, + "loss": 0.98496401, + "num_input_tokens_seen": 1718415, + "step": 76, + "time_per_iteration": 2.7086715698242188 + }, + { + "auxiliary_loss_clip": 0.01753644, + "auxiliary_loss_mlp": 0.01150377, + "balance_loss_clip": 1.24100995, + "balance_loss_mlp": 1.06898069, + "epoch": 0.009258702579210003, + "flos": 27016531372800.0, + "grad_norm": 2.296988922495127, + "language_loss": 0.79536629, + "learning_rate": 3.1468522454274533e-06, + "loss": 0.8244065, + "num_input_tokens_seen": 1738770, + "step": 77, + "time_per_iteration": 2.7888739109039307 + }, + { + "auxiliary_loss_clip": 0.01745446, + "auxiliary_loss_mlp": 0.01137342, + "balance_loss_clip": 1.23528409, + "balance_loss_mlp": 1.05918884, + "epoch": 0.009378945469849095, + "flos": 26903196984960.0, + "grad_norm": 1.9236652964208127, + "language_loss": 0.91852468, + "learning_rate": 3.15620006480197e-06, + "loss": 0.94735253, + "num_input_tokens_seen": 1758040, + "step": 78, + "time_per_iteration": 2.835317373275757 + }, + { + "auxiliary_loss_clip": 0.0174441, + "auxiliary_loss_mlp": 0.01139184, + "balance_loss_clip": 1.23519874, + "balance_loss_mlp": 1.06670499, + "epoch": 0.009499188360488187, + "flos": 35694236327040.0, + "grad_norm": 3.059235943052791, + "language_loss": 0.75050932, + "learning_rate": 3.1654288004333087e-06, + "loss": 0.77934527, + "num_input_tokens_seen": 1776705, + "step": 79, + "time_per_iteration": 2.90307354927063 + }, + { + "auxiliary_loss_clip": 0.01722531, + "auxiliary_loss_mlp": 0.01130586, + "balance_loss_clip": 1.23009825, + "balance_loss_mlp": 1.05562735, + "epoch": 0.009619431251127276, + "flos": 21503201944320.0, + "grad_norm": 3.4198367928986766, + "language_loss": 0.75846994, + "learning_rate": 3.1745414482589353e-06, + "loss": 0.78700101, + "num_input_tokens_seen": 1795915, + "step": 80, + "time_per_iteration": 2.824453592300415 + }, + { + "auxiliary_loss_clip": 0.01722542, + "auxiliary_loss_mlp": 0.01137714, + "balance_loss_clip": 1.23166597, + "balance_loss_mlp": 1.06800044, + "epoch": 0.009739674141766368, + "flos": 17421056991360.0, + "grad_norm": 3.2229969535540857, + "language_loss": 0.87179977, + "learning_rate": 3.1835408925606204e-06, + "loss": 0.90040237, + "num_input_tokens_seen": 1814055, + "step": 81, + "time_per_iteration": 2.7774605751037598 + }, + { + "auxiliary_loss_clip": 0.01709043, + "auxiliary_loss_mlp": 0.01130823, + "balance_loss_clip": 1.22828269, + "balance_loss_mlp": 1.06516266, + "epoch": 0.00985991703240546, + "flos": 27527109246720.0, + "grad_norm": 2.1441743967801807, + "language_loss": 0.89227402, + "learning_rate": 3.1924299114448214e-06, + "loss": 0.92067271, + "num_input_tokens_seen": 1834535, + "step": 82, + "time_per_iteration": 2.9237935543060303 + }, + { + "auxiliary_loss_clip": 0.01705586, + "auxiliary_loss_mlp": 0.01131644, + "balance_loss_clip": 1.22534657, + "balance_loss_mlp": 1.06235993, + "epoch": 0.00998015992304455, + "flos": 13808084509440.0, + "grad_norm": 2.256116635675401, + "language_loss": 0.83491474, + "learning_rate": 3.2012111819909055e-06, + "loss": 0.86328703, + "num_input_tokens_seen": 1851865, + "step": 83, + "time_per_iteration": 2.7475533485412598 + }, + { + "auxiliary_loss_clip": 0.01682366, + "auxiliary_loss_mlp": 0.01137744, + "balance_loss_clip": 1.2161597, + "balance_loss_mlp": 1.07465887, + "epoch": 0.010100402813683641, + "flos": 20191385341440.0, + "grad_norm": 3.2901707592289444, + "language_loss": 0.95135707, + "learning_rate": 3.2098872850910627e-06, + "loss": 0.97955823, + "num_input_tokens_seen": 1868540, + "step": 84, + "time_per_iteration": 2.7455923557281494 + }, + { + "auxiliary_loss_clip": 0.01683848, + "auxiliary_loss_mlp": 0.01138444, + "balance_loss_clip": 1.22173142, + "balance_loss_mlp": 1.072927, + "epoch": 0.010220645704322733, + "flos": 17201642762880.0, + "grad_norm": 1.9415579235020548, + "language_loss": 0.89120209, + "learning_rate": 3.2184607100038194e-06, + "loss": 0.91942501, + "num_input_tokens_seen": 1887180, + "step": 85, + "time_per_iteration": 2.7493715286254883 + }, + { + "auxiliary_loss_clip": 0.01686627, + "auxiliary_loss_mlp": 0.01112062, + "balance_loss_clip": 1.2205404, + "balance_loss_mlp": 1.05093169, + "epoch": 0.010340888594961822, + "flos": 21470415805440.0, + "grad_norm": 2.404231380693293, + "language_loss": 0.9300704, + "learning_rate": 3.2269338586412414e-06, + "loss": 0.95805728, + "num_input_tokens_seen": 1904765, + "step": 86, + "time_per_iteration": 2.832045316696167 + }, + { + "auxiliary_loss_clip": 0.01663843, + "auxiliary_loss_mlp": 0.0112413, + "balance_loss_clip": 1.21356153, + "balance_loss_mlp": 1.06242776, + "epoch": 0.010461131485600914, + "flos": 23002831785600.0, + "grad_norm": 2.3102650660150443, + "language_loss": 0.9639374, + "learning_rate": 3.2353090496083106e-06, + "loss": 0.99181712, + "num_input_tokens_seen": 1922600, + "step": 87, + "time_per_iteration": 2.85151743888855 + }, + { + "auxiliary_loss_clip": 0.01649515, + "auxiliary_loss_mlp": 0.01134822, + "balance_loss_clip": 1.20966125, + "balance_loss_mlp": 1.0728327, + "epoch": 0.010581374376240005, + "flos": 33546850571520.0, + "grad_norm": 2.0921861759342852, + "language_loss": 0.81157243, + "learning_rate": 3.2435885220114572e-06, + "loss": 0.83941579, + "num_input_tokens_seen": 1943950, + "step": 88, + "time_per_iteration": 2.8560028076171875 + }, + { + "auxiliary_loss_clip": 0.01655858, + "auxiliary_loss_mlp": 0.01109197, + "balance_loss_clip": 1.21267664, + "balance_loss_mlp": 1.05366969, + "epoch": 0.010701617266879095, + "flos": 21763087822080.0, + "grad_norm": 2.1431670450371842, + "language_loss": 0.93944585, + "learning_rate": 3.2517744390519113e-06, + "loss": 0.96709645, + "num_input_tokens_seen": 1962815, + "step": 89, + "time_per_iteration": 2.7945780754089355 + }, + { + "auxiliary_loss_clip": 0.01650624, + "auxiliary_loss_mlp": 0.01101801, + "balance_loss_clip": 1.20668697, + "balance_loss_mlp": 1.04696476, + "epoch": 0.010821860157518187, + "flos": 19060199256960.0, + "grad_norm": 1.9942618180136416, + "language_loss": 0.7494424, + "learning_rate": 3.259868891418298e-06, + "loss": 0.77696669, + "num_input_tokens_seen": 1980580, + "step": 90, + "time_per_iteration": 2.772324562072754 + }, + { + "auxiliary_loss_clip": 0.01648506, + "auxiliary_loss_mlp": 0.01126369, + "balance_loss_clip": 1.20815587, + "balance_loss_mlp": 1.0698874, + "epoch": 0.010942103048157278, + "flos": 25447378757760.0, + "grad_norm": 1.979606189288899, + "language_loss": 0.85118377, + "learning_rate": 3.2678739004917757e-06, + "loss": 0.87893248, + "num_input_tokens_seen": 2000315, + "step": 91, + "time_per_iteration": 2.7983713150024414 + }, + { + "auxiliary_loss_clip": 0.01635367, + "auxiliary_loss_mlp": 0.01121355, + "balance_loss_clip": 1.20575655, + "balance_loss_mlp": 1.06518364, + "epoch": 0.011062345938796368, + "flos": 27493928058240.0, + "grad_norm": 2.0109409182865963, + "language_loss": 0.92106134, + "learning_rate": 3.275791421376029e-06, + "loss": 0.94862854, + "num_input_tokens_seen": 2023760, + "step": 92, + "time_per_iteration": 3.7902071475982666 + }, + { + "auxiliary_loss_clip": 0.01622123, + "auxiliary_loss_mlp": 0.01137117, + "balance_loss_clip": 1.1984868, + "balance_loss_mlp": 1.08316255, + "epoch": 0.01118258882943546, + "flos": 16071210864000.0, + "grad_norm": 1.9667972053585663, + "language_loss": 0.95906734, + "learning_rate": 3.2836233457634622e-06, + "loss": 0.98665971, + "num_input_tokens_seen": 2041895, + "step": 93, + "time_per_iteration": 3.680514335632324 + }, + { + "auxiliary_loss_clip": 0.01640213, + "auxiliary_loss_mlp": 0.01111662, + "balance_loss_clip": 1.20704103, + "balance_loss_mlp": 1.06283343, + "epoch": 0.011302831720074551, + "flos": 20668602458880.0, + "grad_norm": 2.6185541954149225, + "language_loss": 0.85403013, + "learning_rate": 3.2913715046481135e-06, + "loss": 0.88154888, + "num_input_tokens_seen": 2061640, + "step": 94, + "time_per_iteration": 3.7056121826171875 + }, + { + "auxiliary_loss_clip": 0.01623585, + "auxiliary_loss_mlp": 0.01103778, + "balance_loss_clip": 1.19737911, + "balance_loss_mlp": 1.05104017, + "epoch": 0.011423074610713641, + "flos": 13072238490240.0, + "grad_norm": 2.373832221529278, + "language_loss": 0.88754809, + "learning_rate": 3.299037670895023e-06, + "loss": 0.91482174, + "num_input_tokens_seen": 2078255, + "step": 95, + "time_per_iteration": 2.715282440185547 + }, + { + "auxiliary_loss_clip": 0.01614729, + "auxiliary_loss_mlp": 0.01111658, + "balance_loss_clip": 1.1953311, + "balance_loss_mlp": 1.06097007, + "epoch": 0.011543317501352733, + "flos": 30335646689280.0, + "grad_norm": 1.921272734418009, + "language_loss": 0.80263513, + "learning_rate": 3.3066235616750667e-06, + "loss": 0.82989895, + "num_input_tokens_seen": 2099490, + "step": 96, + "time_per_iteration": 2.9072201251983643 + }, + { + "auxiliary_loss_clip": 0.01614805, + "auxiliary_loss_mlp": 0.01091496, + "balance_loss_clip": 1.19986105, + "balance_loss_mlp": 1.04536247, + "epoch": 0.011663560391991824, + "flos": 15522962601600.0, + "grad_norm": 3.5729178442994027, + "language_loss": 0.92093945, + "learning_rate": 3.3141308407736276e-06, + "loss": 0.94800246, + "num_input_tokens_seen": 2116125, + "step": 97, + "time_per_iteration": 2.757974624633789 + }, + { + "auxiliary_loss_clip": 0.01603445, + "auxiliary_loss_mlp": 0.01104195, + "balance_loss_clip": 1.18962383, + "balance_loss_mlp": 1.05615377, + "epoch": 0.011783803282630914, + "flos": 19902125116800.0, + "grad_norm": 2.007854718039964, + "language_loss": 0.86699188, + "learning_rate": 3.321561120780869e-06, + "loss": 0.89406824, + "num_input_tokens_seen": 2134835, + "step": 98, + "time_per_iteration": 2.7972960472106934 + }, + { + "auxiliary_loss_clip": 0.01599815, + "auxiliary_loss_mlp": 0.01088874, + "balance_loss_clip": 1.19388008, + "balance_loss_mlp": 1.0462929, + "epoch": 0.011904046173270006, + "flos": 22340674517760.0, + "grad_norm": 11.86685586654566, + "language_loss": 1.01378489, + "learning_rate": 3.3289159651708192e-06, + "loss": 1.04067183, + "num_input_tokens_seen": 2152410, + "step": 99, + "time_per_iteration": 2.778303861618042 + }, + { + "auxiliary_loss_clip": 0.01590174, + "auxiliary_loss_mlp": 0.01094446, + "balance_loss_clip": 1.18938947, + "balance_loss_mlp": 1.05217433, + "epoch": 0.012024289063909096, + "flos": 19100060375040.0, + "grad_norm": 1.996137877525179, + "language_loss": 0.97496271, + "learning_rate": 3.3361968902759768e-06, + "loss": 1.00180888, + "num_input_tokens_seen": 2172090, + "step": 100, + "time_per_iteration": 2.841456651687622 + }, + { + "auxiliary_loss_clip": 0.01580301, + "auxiliary_loss_mlp": 0.01110611, + "balance_loss_clip": 1.18712068, + "balance_loss_mlp": 1.06671786, + "epoch": 0.012144531954548187, + "flos": 15012205159680.0, + "grad_norm": 2.1653407545043906, + "language_loss": 0.93925709, + "learning_rate": 3.343405367163663e-06, + "loss": 0.9661662, + "num_input_tokens_seen": 2189020, + "step": 101, + "time_per_iteration": 2.802994966506958 + }, + { + "auxiliary_loss_clip": 0.01602812, + "auxiliary_loss_mlp": 0.01105022, + "balance_loss_clip": 1.19709301, + "balance_loss_mlp": 1.06296468, + "epoch": 0.012264774845187279, + "flos": 15122020014720.0, + "grad_norm": 2.5599023615709546, + "language_loss": 0.81210625, + "learning_rate": 3.350542823419951e-06, + "loss": 0.83918458, + "num_input_tokens_seen": 2205620, + "step": 102, + "time_per_iteration": 2.732969284057617 + }, + { + "auxiliary_loss_clip": 0.01578658, + "auxiliary_loss_mlp": 0.01093936, + "balance_loss_clip": 1.18506241, + "balance_loss_mlp": 1.05116355, + "epoch": 0.012385017735826368, + "flos": 13949248959360.0, + "grad_norm": 3.0879611295779372, + "language_loss": 0.87387645, + "learning_rate": 3.3576106448465615e-06, + "loss": 0.90060234, + "num_input_tokens_seen": 2219000, + "step": 103, + "time_per_iteration": 2.7675552368164062 + }, + { + "auxiliary_loss_clip": 0.0157756, + "auxiliary_loss_mlp": 0.01107421, + "balance_loss_clip": 1.1845578, + "balance_loss_mlp": 1.06698561, + "epoch": 0.01250526062646546, + "flos": 23623260428160.0, + "grad_norm": 2.7227320394107846, + "language_loss": 0.88039041, + "learning_rate": 3.3646101770757797e-06, + "loss": 0.90724015, + "num_input_tokens_seen": 2237790, + "step": 104, + "time_per_iteration": 2.731309175491333 + }, + { + "auxiliary_loss_clip": 0.01579204, + "auxiliary_loss_mlp": 0.01111907, + "balance_loss_clip": 1.18996131, + "balance_loss_mlp": 1.06956351, + "epoch": 0.012625503517104552, + "flos": 34640078958720.0, + "grad_norm": 1.90333381934037, + "language_loss": 0.85644734, + "learning_rate": 3.371542727108104e-06, + "loss": 0.88335848, + "num_input_tokens_seen": 2259965, + "step": 105, + "time_per_iteration": 2.9264981746673584 + }, + { + "auxiliary_loss_clip": 0.01571278, + "auxiliary_loss_mlp": 0.01093078, + "balance_loss_clip": 1.18362212, + "balance_loss_mlp": 1.05450225, + "epoch": 0.012745746407743641, + "flos": 17821891837440.0, + "grad_norm": 2.5758322493886925, + "language_loss": 0.89957523, + "learning_rate": 3.3784095647770114e-06, + "loss": 0.92621881, + "num_input_tokens_seen": 2278610, + "step": 106, + "time_per_iteration": 2.8100011348724365 + }, + { + "auxiliary_loss_clip": 0.01560427, + "auxiliary_loss_mlp": 0.01091332, + "balance_loss_clip": 1.18038797, + "balance_loss_mlp": 1.05335212, + "epoch": 0.012865989298382733, + "flos": 20595057361920.0, + "grad_norm": 2.1195193398880656, + "language_loss": 0.88396829, + "learning_rate": 3.3852119241449547e-06, + "loss": 0.91048586, + "num_input_tokens_seen": 2297730, + "step": 107, + "time_per_iteration": 2.936950922012329 + }, + { + "auxiliary_loss_clip": 0.01564176, + "auxiliary_loss_mlp": 0.01096647, + "balance_loss_clip": 1.17937946, + "balance_loss_mlp": 1.06007397, + "epoch": 0.012986232189021825, + "flos": 23948969978880.0, + "grad_norm": 2.584811816029044, + "language_loss": 0.96511918, + "learning_rate": 3.3919510048344295e-06, + "loss": 0.99172741, + "num_input_tokens_seen": 2315740, + "step": 108, + "time_per_iteration": 2.856675624847412 + }, + { + "auxiliary_loss_clip": 0.01539512, + "auxiliary_loss_mlp": 0.01092309, + "balance_loss_clip": 1.17268014, + "balance_loss_mlp": 1.05680799, + "epoch": 0.013106475079660914, + "flos": 23725425686400.0, + "grad_norm": 2.8532392555189614, + "language_loss": 0.86634797, + "learning_rate": 3.3986279732976907e-06, + "loss": 0.89266622, + "num_input_tokens_seen": 2334215, + "step": 109, + "time_per_iteration": 2.800995111465454 + }, + { + "auxiliary_loss_clip": 0.01546177, + "auxiliary_loss_mlp": 0.01107292, + "balance_loss_clip": 1.17766786, + "balance_loss_mlp": 1.07365096, + "epoch": 0.013226717970300006, + "flos": 21102438925440.0, + "grad_norm": 2.0291708209078045, + "language_loss": 0.95501816, + "learning_rate": 3.4052439640284983e-06, + "loss": 0.98155284, + "num_input_tokens_seen": 2353130, + "step": 110, + "time_per_iteration": 2.7434537410736084 + }, + { + "auxiliary_loss_clip": 0.01550042, + "auxiliary_loss_mlp": 0.01087338, + "balance_loss_clip": 1.181247, + "balance_loss_mlp": 1.0532918, + "epoch": 0.013346960860939098, + "flos": 24863902231680.0, + "grad_norm": 1.729399251865375, + "language_loss": 0.81158888, + "learning_rate": 3.4118000807190217e-06, + "loss": 0.83796263, + "num_input_tokens_seen": 2374010, + "step": 111, + "time_per_iteration": 2.766143321990967 + }, + { + "auxiliary_loss_clip": 0.01542353, + "auxiliary_loss_mlp": 0.01093173, + "balance_loss_clip": 1.17392898, + "balance_loss_mlp": 1.05664754, + "epoch": 0.013467203751578187, + "flos": 28181940140160.0, + "grad_norm": 2.0396549881468, + "language_loss": 0.76140022, + "learning_rate": 3.4182973973648723e-06, + "loss": 0.78775549, + "num_input_tokens_seen": 2395220, + "step": 112, + "time_per_iteration": 2.752223014831543 + }, + { + "auxiliary_loss_clip": 0.01532672, + "auxiliary_loss_mlp": 0.0110173, + "balance_loss_clip": 1.17452121, + "balance_loss_mlp": 1.0682317, + "epoch": 0.013587446642217279, + "flos": 18916233546240.0, + "grad_norm": 6.4787845585367725, + "language_loss": 0.94965303, + "learning_rate": 3.424736959321014e-06, + "loss": 0.97599697, + "num_input_tokens_seen": 2413025, + "step": 113, + "time_per_iteration": 2.889324903488159 + }, + { + "auxiliary_loss_clip": 0.01538392, + "auxiliary_loss_mlp": 0.01098469, + "balance_loss_clip": 1.17755342, + "balance_loss_mlp": 1.06492329, + "epoch": 0.01370768953285637, + "flos": 23988615615360.0, + "grad_norm": 2.3477436381209222, + "language_loss": 0.88668787, + "learning_rate": 3.431119784311155e-06, + "loss": 0.91305649, + "num_input_tokens_seen": 2432700, + "step": 114, + "time_per_iteration": 2.7489993572235107 + }, + { + "auxiliary_loss_clip": 0.01533495, + "auxiliary_loss_mlp": 0.01086977, + "balance_loss_clip": 1.17806232, + "balance_loss_mlp": 1.05498135, + "epoch": 0.01382793242349546, + "flos": 39202565512320.0, + "grad_norm": 1.8688489861506037, + "language_loss": 0.7743181, + "learning_rate": 3.43744686339307e-06, + "loss": 0.8005228, + "num_input_tokens_seen": 2455020, + "step": 115, + "time_per_iteration": 2.8983659744262695 + }, + { + "auxiliary_loss_clip": 0.01518172, + "auxiliary_loss_mlp": 0.01093664, + "balance_loss_clip": 1.16536069, + "balance_loss_mlp": 1.06063104, + "epoch": 0.013948175314134552, + "flos": 41353506714240.0, + "grad_norm": 17.816743709766648, + "language_loss": 0.90690529, + "learning_rate": 3.44371916188212e-06, + "loss": 0.93302363, + "num_input_tokens_seen": 2475775, + "step": 116, + "time_per_iteration": 2.8805453777313232 + }, + { + "auxiliary_loss_clip": 0.01514324, + "auxiliary_loss_mlp": 0.01089559, + "balance_loss_clip": 1.1645484, + "balance_loss_mlp": 1.056633, + "epoch": 0.014068418204773643, + "flos": 22453542028800.0, + "grad_norm": 2.0513734312010774, + "language_loss": 0.86207402, + "learning_rate": 3.449937620235143e-06, + "loss": 0.8881129, + "num_input_tokens_seen": 2496370, + "step": 117, + "time_per_iteration": 2.795118570327759 + }, + { + "auxiliary_loss_clip": 0.01518491, + "auxiliary_loss_mlp": 0.01090708, + "balance_loss_clip": 1.17040884, + "balance_loss_mlp": 1.05904603, + "epoch": 0.014188661095412733, + "flos": 23805147922560.0, + "grad_norm": 1.7459716732171116, + "language_loss": 0.89392579, + "learning_rate": 3.456103154896722e-06, + "loss": 0.92001784, + "num_input_tokens_seen": 2517645, + "step": 118, + "time_per_iteration": 3.67635440826416 + }, + { + "auxiliary_loss_clip": 0.01513364, + "auxiliary_loss_mlp": 0.01070334, + "balance_loss_clip": 1.16938853, + "balance_loss_mlp": 1.04169989, + "epoch": 0.014308903986051825, + "flos": 23660248458240.0, + "grad_norm": 1.8992078059639352, + "language_loss": 0.92193186, + "learning_rate": 3.462216659109757e-06, + "loss": 0.94776881, + "num_input_tokens_seen": 2537825, + "step": 119, + "time_per_iteration": 4.614622354507446 + }, + { + "auxiliary_loss_clip": 0.01523678, + "auxiliary_loss_mlp": 0.01091443, + "balance_loss_clip": 1.17057133, + "balance_loss_mlp": 1.06021047, + "epoch": 0.014429146876690916, + "flos": 20667991927680.0, + "grad_norm": 4.188243129687955, + "language_loss": 0.85524547, + "learning_rate": 3.4682790036921077e-06, + "loss": 0.88139665, + "num_input_tokens_seen": 2556485, + "step": 120, + "time_per_iteration": 3.631289005279541 + }, + { + "auxiliary_loss_clip": 0.01495776, + "auxiliary_loss_mlp": 0.01095088, + "balance_loss_clip": 1.16473198, + "balance_loss_mlp": 1.06734848, + "epoch": 0.014549389767330006, + "flos": 20229199384320.0, + "grad_norm": 1.8787831197776652, + "language_loss": 0.83217287, + "learning_rate": 3.4742910377810193e-06, + "loss": 0.85808146, + "num_input_tokens_seen": 2573945, + "step": 121, + "time_per_iteration": 2.729886770248413 + }, + { + "auxiliary_loss_clip": 0.01497617, + "auxiliary_loss_mlp": 0.01105134, + "balance_loss_clip": 1.16217089, + "balance_loss_mlp": 1.07517636, + "epoch": 0.014669632657969098, + "flos": 18004174381440.0, + "grad_norm": 2.2572454923291194, + "language_loss": 0.88716537, + "learning_rate": 3.4802535895469042e-06, + "loss": 0.91319287, + "num_input_tokens_seen": 2592695, + "step": 122, + "time_per_iteration": 2.7558400630950928 + }, + { + "auxiliary_loss_clip": 0.01494708, + "auxiliary_loss_mlp": 0.01088993, + "balance_loss_clip": 1.1617595, + "balance_loss_mlp": 1.0598104, + "epoch": 0.01478987554860819, + "flos": 22741796672640.0, + "grad_norm": 2.1199095985242886, + "language_loss": 0.89521635, + "learning_rate": 3.4861674668779934e-06, + "loss": 0.92105341, + "num_input_tokens_seen": 2610925, + "step": 123, + "time_per_iteration": 2.7713732719421387 + }, + { + "auxiliary_loss_clip": 0.01491666, + "auxiliary_loss_mlp": 0.01096311, + "balance_loss_clip": 1.161798, + "balance_loss_mlp": 1.06787944, + "epoch": 0.01491011843924728, + "flos": 17198590106880.0, + "grad_norm": 4.078430605412225, + "language_loss": 0.84043062, + "learning_rate": 3.492033458037272e-06, + "loss": 0.86631036, + "num_input_tokens_seen": 2629495, + "step": 124, + "time_per_iteration": 2.682034492492676 + }, + { + "auxiliary_loss_clip": 0.0148585, + "auxiliary_loss_mlp": 0.01070758, + "balance_loss_clip": 1.15875173, + "balance_loss_mlp": 1.04282749, + "epoch": 0.01503036132988637, + "flos": 17673867889920.0, + "grad_norm": 2.3766289809136523, + "language_loss": 0.86952746, + "learning_rate": 3.497852332293018e-06, + "loss": 0.89509356, + "num_input_tokens_seen": 2645070, + "step": 125, + "time_per_iteration": 2.6742866039276123 + }, + { + "auxiliary_loss_clip": 0.01496098, + "auxiliary_loss_mlp": 0.01085887, + "balance_loss_clip": 1.1625849, + "balance_loss_mlp": 1.0575031, + "epoch": 0.015150604220525462, + "flos": 18878239935360.0, + "grad_norm": 3.058555880483079, + "language_loss": 0.96587431, + "learning_rate": 3.5036248405242356e-06, + "loss": 0.99169415, + "num_input_tokens_seen": 2663825, + "step": 126, + "time_per_iteration": 2.6963913440704346 + }, + { + "auxiliary_loss_clip": 0.01487041, + "auxiliary_loss_mlp": 0.01067958, + "balance_loss_clip": 1.15907121, + "balance_loss_mlp": 1.0418874, + "epoch": 0.015270847111164552, + "flos": 39420184060800.0, + "grad_norm": 2.0827096574271518, + "language_loss": 0.82779652, + "learning_rate": 3.509351715802146e-06, + "loss": 0.85334653, + "num_input_tokens_seen": 2684710, + "step": 127, + "time_per_iteration": 2.862795114517212 + }, + { + "auxiliary_loss_clip": 0.0148624, + "auxiliary_loss_mlp": 0.01076051, + "balance_loss_clip": 1.1568594, + "balance_loss_mlp": 1.04691672, + "epoch": 0.015391090001803644, + "flos": 43762466286720.0, + "grad_norm": 2.105988931544277, + "language_loss": 0.78418779, + "learning_rate": 3.5150336739488763e-06, + "loss": 0.8098107, + "num_input_tokens_seen": 2706995, + "step": 128, + "time_per_iteration": 2.9210822582244873 + }, + { + "auxiliary_loss_clip": 0.01479319, + "auxiliary_loss_mlp": 0.01094987, + "balance_loss_clip": 1.15618277, + "balance_loss_mlp": 1.0668416, + "epoch": 0.015511332892442733, + "flos": 18916341287040.0, + "grad_norm": 1.8755131379491867, + "language_loss": 0.84034145, + "learning_rate": 3.5206714140744143e-06, + "loss": 0.86608446, + "num_input_tokens_seen": 2727050, + "step": 129, + "time_per_iteration": 2.768864870071411 + }, + { + "auxiliary_loss_clip": 0.01483573, + "auxiliary_loss_mlp": 0.01082435, + "balance_loss_clip": 1.16078687, + "balance_loss_mlp": 1.05625713, + "epoch": 0.015631575783081827, + "flos": 24535283679360.0, + "grad_norm": 4.310873434187784, + "language_loss": 0.87491357, + "learning_rate": 3.5262656190928208e-06, + "loss": 0.90057367, + "num_input_tokens_seen": 2745350, + "step": 130, + "time_per_iteration": 2.747046709060669 + }, + { + "auxiliary_loss_clip": 0.01416957, + "auxiliary_loss_mlp": 0.01038016, + "balance_loss_clip": 1.21044803, + "balance_loss_mlp": 1.02323365, + "epoch": 0.015751818673720917, + "flos": 62328536098560.0, + "grad_norm": 1.0278222438858746, + "language_loss": 0.7146225, + "learning_rate": 3.5318169562186737e-06, + "loss": 0.73917222, + "num_input_tokens_seen": 2814195, + "step": 131, + "time_per_iteration": 3.300032377243042 + }, + { + "auxiliary_loss_clip": 0.01483809, + "auxiliary_loss_mlp": 0.01082231, + "balance_loss_clip": 1.15807748, + "balance_loss_mlp": 1.05635047, + "epoch": 0.015872061564360006, + "flos": 23878549365120.0, + "grad_norm": 1.8536288049340655, + "language_loss": 0.82085282, + "learning_rate": 3.5373260774446292e-06, + "loss": 0.84651321, + "num_input_tokens_seen": 2834645, + "step": 132, + "time_per_iteration": 2.7401225566864014 + }, + { + "auxiliary_loss_clip": 0.0147685, + "auxiliary_loss_mlp": 0.01057798, + "balance_loss_clip": 1.15678191, + "balance_loss_mlp": 1.03552938, + "epoch": 0.0159923044549991, + "flos": 23367899664000.0, + "grad_norm": 1.865425564022484, + "language_loss": 0.90146029, + "learning_rate": 3.542793620000961e-06, + "loss": 0.92680669, + "num_input_tokens_seen": 2854120, + "step": 133, + "time_per_iteration": 2.8912439346313477 + }, + { + "auxiliary_loss_clip": 0.01474343, + "auxiliary_loss_mlp": 0.01080865, + "balance_loss_clip": 1.15678823, + "balance_loss_mlp": 1.06101668, + "epoch": 0.01611254734563819, + "flos": 17858305249920.0, + "grad_norm": 2.581248477302466, + "language_loss": 0.86804879, + "learning_rate": 3.5482202067978894e-06, + "loss": 0.89360094, + "num_input_tokens_seen": 2871330, + "step": 134, + "time_per_iteration": 2.718918800354004 + }, + { + "auxiliary_loss_clip": 0.01468403, + "auxiliary_loss_mlp": 0.0107604, + "balance_loss_clip": 1.15394425, + "balance_loss_mlp": 1.05201924, + "epoch": 0.01623279023627728, + "flos": 20954774113920.0, + "grad_norm": 2.15312230359652, + "language_loss": 0.75942039, + "learning_rate": 3.553606446851471e-06, + "loss": 0.78486478, + "num_input_tokens_seen": 2888070, + "step": 135, + "time_per_iteration": 2.677715539932251 + }, + { + "auxiliary_loss_clip": 0.01465748, + "auxiliary_loss_mlp": 0.01065048, + "balance_loss_clip": 1.15520573, + "balance_loss_mlp": 1.04330444, + "epoch": 0.016353033126916373, + "flos": 15742412743680.0, + "grad_norm": 2.280958986165735, + "language_loss": 0.83494365, + "learning_rate": 3.5589529356937613e-06, + "loss": 0.86025161, + "num_input_tokens_seen": 2906465, + "step": 136, + "time_per_iteration": 2.6662375926971436 + }, + { + "auxiliary_loss_clip": 0.01461214, + "auxiliary_loss_mlp": 0.01102305, + "balance_loss_clip": 1.14844441, + "balance_loss_mlp": 1.07692552, + "epoch": 0.016473276017555463, + "flos": 18807280617600.0, + "grad_norm": 3.4364554014137503, + "language_loss": 0.77046704, + "learning_rate": 3.5642602557679627e-06, + "loss": 0.79610223, + "num_input_tokens_seen": 2924915, + "step": 137, + "time_per_iteration": 2.687528610229492 + }, + { + "auxiliary_loss_clip": 0.01466281, + "auxiliary_loss_mlp": 0.01069128, + "balance_loss_clip": 1.15756524, + "balance_loss_mlp": 1.04489279, + "epoch": 0.016593518908194552, + "flos": 24352641999360.0, + "grad_norm": 2.0813243110345074, + "language_loss": 0.84211564, + "learning_rate": 3.569528976809202e-06, + "loss": 0.86746967, + "num_input_tokens_seen": 2942130, + "step": 138, + "time_per_iteration": 2.6932921409606934 + }, + { + "auxiliary_loss_clip": 0.0145949, + "auxiliary_loss_mlp": 0.01065674, + "balance_loss_clip": 1.14894092, + "balance_loss_mlp": 1.04294038, + "epoch": 0.016713761798833646, + "flos": 22346133384960.0, + "grad_norm": 1.6529591911382966, + "language_loss": 0.89890182, + "learning_rate": 3.5747596562115522e-06, + "loss": 0.92415345, + "num_input_tokens_seen": 2962745, + "step": 139, + "time_per_iteration": 2.715885877609253 + }, + { + "auxiliary_loss_clip": 0.01466461, + "auxiliary_loss_mlp": 0.01072261, + "balance_loss_clip": 1.1523217, + "balance_loss_mlp": 1.04863334, + "epoch": 0.016834004689472735, + "flos": 17821820010240.0, + "grad_norm": 2.2648990804843425, + "language_loss": 0.90975928, + "learning_rate": 3.5799528393819138e-06, + "loss": 0.93514645, + "num_input_tokens_seen": 2981825, + "step": 140, + "time_per_iteration": 2.7501652240753174 + }, + { + "auxiliary_loss_clip": 0.01449983, + "auxiliary_loss_mlp": 0.01084381, + "balance_loss_clip": 1.14444661, + "balance_loss_mlp": 1.06135011, + "epoch": 0.016954247580111825, + "flos": 20519501103360.0, + "grad_norm": 2.009964765712365, + "language_loss": 0.88211703, + "learning_rate": 3.585109060081286e-06, + "loss": 0.90746069, + "num_input_tokens_seen": 3001625, + "step": 141, + "time_per_iteration": 2.712169647216797 + }, + { + "auxiliary_loss_clip": 0.01461514, + "auxiliary_loss_mlp": 0.01076261, + "balance_loss_clip": 1.14904189, + "balance_loss_mlp": 1.05346799, + "epoch": 0.017074490470750915, + "flos": 22088869200000.0, + "grad_norm": 1.9479756173064928, + "language_loss": 0.78453207, + "learning_rate": 3.590228840753992e-06, + "loss": 0.80990982, + "num_input_tokens_seen": 3022055, + "step": 142, + "time_per_iteration": 2.7825212478637695 + }, + { + "auxiliary_loss_clip": 0.01448473, + "auxiliary_loss_mlp": 0.01080047, + "balance_loss_clip": 1.14443004, + "balance_loss_mlp": 1.05857706, + "epoch": 0.01719473336139001, + "flos": 15997270717440.0, + "grad_norm": 2.1053555517245512, + "language_loss": 0.87140381, + "learning_rate": 3.5953126928453423e-06, + "loss": 0.89668906, + "num_input_tokens_seen": 3039605, + "step": 143, + "time_per_iteration": 2.674713611602783 + }, + { + "auxiliary_loss_clip": 0.01452722, + "auxiliary_loss_mlp": 0.01069982, + "balance_loss_clip": 1.14742541, + "balance_loss_mlp": 1.04821408, + "epoch": 0.017314976252029098, + "flos": 22492038430080.0, + "grad_norm": 2.2971328262229953, + "language_loss": 0.80525708, + "learning_rate": 3.600361117108239e-06, + "loss": 0.83048415, + "num_input_tokens_seen": 3059405, + "step": 144, + "time_per_iteration": 2.7032968997955322 + }, + { + "auxiliary_loss_clip": 0.01445276, + "auxiliary_loss_mlp": 0.01083262, + "balance_loss_clip": 1.14307404, + "balance_loss_mlp": 1.06117296, + "epoch": 0.017435219142668188, + "flos": 22018053536640.0, + "grad_norm": 2.166060253615211, + "language_loss": 0.97094035, + "learning_rate": 3.6053746038991616e-06, + "loss": 0.99622571, + "num_input_tokens_seen": 3078490, + "step": 145, + "time_per_iteration": 5.409006357192993 + }, + { + "auxiliary_loss_clip": 0.01359112, + "auxiliary_loss_mlp": 0.0100782, + "balance_loss_clip": 1.16963792, + "balance_loss_mlp": 0.99775833, + "epoch": 0.01755546203330728, + "flos": 72240526149120.0, + "grad_norm": 1.0563749353786351, + "language_loss": 0.5842849, + "learning_rate": 3.6103536334639843e-06, + "loss": 0.60795426, + "num_input_tokens_seen": 3131755, + "step": 146, + "time_per_iteration": 4.1931774616241455 + }, + { + "auxiliary_loss_clip": 0.01437987, + "auxiliary_loss_mlp": 0.01061936, + "balance_loss_clip": 1.14085603, + "balance_loss_mlp": 1.04019248, + "epoch": 0.01767570492394637, + "flos": 25337061112320.0, + "grad_norm": 1.9698376934976205, + "language_loss": 0.85410798, + "learning_rate": 3.615298676214041e-06, + "loss": 0.87910712, + "num_input_tokens_seen": 3152035, + "step": 147, + "time_per_iteration": 2.7616169452667236 + }, + { + "auxiliary_loss_clip": 0.01444475, + "auxiliary_loss_mlp": 0.01061171, + "balance_loss_clip": 1.14229584, + "balance_loss_mlp": 1.04140639, + "epoch": 0.01779594781458546, + "flos": 20449188230400.0, + "grad_norm": 2.0209068489637283, + "language_loss": 0.88885391, + "learning_rate": 3.6202101929928317e-06, + "loss": 0.91391045, + "num_input_tokens_seen": 3170625, + "step": 148, + "time_per_iteration": 2.738894462585449 + }, + { + "auxiliary_loss_clip": 0.01437015, + "auxiliary_loss_mlp": 0.01062971, + "balance_loss_clip": 1.14057636, + "balance_loss_mlp": 1.04361105, + "epoch": 0.017916190705224554, + "flos": 16253601148800.0, + "grad_norm": 2.0044782373441676, + "language_loss": 0.88486552, + "learning_rate": 3.6250886353337413e-06, + "loss": 0.90986538, + "num_input_tokens_seen": 3188155, + "step": 149, + "time_per_iteration": 2.7145726680755615 + }, + { + "auxiliary_loss_clip": 0.01434978, + "auxiliary_loss_mlp": 0.01065885, + "balance_loss_clip": 1.13796747, + "balance_loss_mlp": 1.04465437, + "epoch": 0.018036433595863644, + "flos": 23330588411520.0, + "grad_norm": 2.0438493123875325, + "language_loss": 0.8640157, + "learning_rate": 3.6299344457091488e-06, + "loss": 0.88902432, + "num_input_tokens_seen": 3209015, + "step": 150, + "time_per_iteration": 2.821089506149292 + }, + { + "auxiliary_loss_clip": 0.0143405, + "auxiliary_loss_mlp": 0.01069507, + "balance_loss_clip": 1.1393882, + "balance_loss_mlp": 1.05085123, + "epoch": 0.018156676486502734, + "flos": 18588010043520.0, + "grad_norm": 2.43222259332194, + "language_loss": 0.93902171, + "learning_rate": 3.634748057771256e-06, + "loss": 0.96405727, + "num_input_tokens_seen": 3224955, + "step": 151, + "time_per_iteration": 2.7888877391815186 + }, + { + "auxiliary_loss_clip": 0.01431878, + "auxiliary_loss_mlp": 0.01062213, + "balance_loss_clip": 1.13885713, + "balance_loss_mlp": 1.04273462, + "epoch": 0.018276919377141827, + "flos": 25448707560960.0, + "grad_norm": 1.6230131659021398, + "language_loss": 0.85562265, + "learning_rate": 3.639529896584965e-06, + "loss": 0.88056356, + "num_input_tokens_seen": 3246330, + "step": 152, + "time_per_iteration": 2.754772424697876 + }, + { + "auxiliary_loss_clip": 0.01430521, + "auxiliary_loss_mlp": 0.01068008, + "balance_loss_clip": 1.13825822, + "balance_loss_mlp": 1.05024552, + "epoch": 0.018397162267780917, + "flos": 20047311889920.0, + "grad_norm": 4.428381152506106, + "language_loss": 0.88629001, + "learning_rate": 3.6442803788531233e-06, + "loss": 0.91127527, + "num_input_tokens_seen": 3264290, + "step": 153, + "time_per_iteration": 2.754067897796631 + }, + { + "auxiliary_loss_clip": 0.0142838, + "auxiliary_loss_mlp": 0.01078111, + "balance_loss_clip": 1.13552761, + "balance_loss_mlp": 1.05876303, + "epoch": 0.018517405158420007, + "flos": 27565282425600.0, + "grad_norm": 7.683943058718452, + "language_loss": 0.96017623, + "learning_rate": 3.6489999131344357e-06, + "loss": 0.98524117, + "num_input_tokens_seen": 3287065, + "step": 154, + "time_per_iteration": 2.768587589263916 + }, + { + "auxiliary_loss_clip": 0.01418181, + "auxiliary_loss_mlp": 0.01061984, + "balance_loss_clip": 1.13434935, + "balance_loss_mlp": 1.04385233, + "epoch": 0.0186376480490591, + "flos": 19354056422400.0, + "grad_norm": 1.8546286018359346, + "language_loss": 0.9070192, + "learning_rate": 3.653688900054313e-06, + "loss": 0.93182081, + "num_input_tokens_seen": 3305595, + "step": 155, + "time_per_iteration": 2.6791701316833496 + }, + { + "auxiliary_loss_clip": 0.01427084, + "auxiliary_loss_mlp": 0.01061197, + "balance_loss_clip": 1.132231, + "balance_loss_mlp": 1.04301763, + "epoch": 0.01875789093969819, + "flos": 26687840993280.0, + "grad_norm": 1.9286916374199499, + "language_loss": 0.75992632, + "learning_rate": 3.6583477325089526e-06, + "loss": 0.78480911, + "num_input_tokens_seen": 3326135, + "step": 156, + "time_per_iteration": 2.7643330097198486 + }, + { + "auxiliary_loss_clip": 0.01415558, + "auxiliary_loss_mlp": 0.01067543, + "balance_loss_clip": 1.13159347, + "balance_loss_mlp": 1.0504601, + "epoch": 0.01887813383033728, + "flos": 24353001135360.0, + "grad_norm": 2.531564039369785, + "language_loss": 1.04289293, + "learning_rate": 3.6629767958628916e-06, + "loss": 1.06772375, + "num_input_tokens_seen": 3343510, + "step": 157, + "time_per_iteration": 2.7242398262023926 + }, + { + "auxiliary_loss_clip": 0.01423836, + "auxiliary_loss_mlp": 0.01079055, + "balance_loss_clip": 1.13581812, + "balance_loss_mlp": 1.05965948, + "epoch": 0.018998376720976373, + "flos": 14647532330880.0, + "grad_norm": 2.422245585224597, + "language_loss": 0.85496145, + "learning_rate": 3.667576468140291e-06, + "loss": 0.87999034, + "num_input_tokens_seen": 3361325, + "step": 158, + "time_per_iteration": 2.675607919692993 + }, + { + "auxiliary_loss_clip": 0.01414119, + "auxiliary_loss_mlp": 0.0105521, + "balance_loss_clip": 1.12814236, + "balance_loss_mlp": 1.03679204, + "epoch": 0.019118619611615463, + "flos": 29305261146240.0, + "grad_norm": 2.300903987762929, + "language_loss": 0.88948572, + "learning_rate": 3.672147120210184e-06, + "loss": 0.91417897, + "num_input_tokens_seen": 3377925, + "step": 159, + "time_per_iteration": 2.7247307300567627 + }, + { + "auxiliary_loss_clip": 0.01415551, + "auxiliary_loss_mlp": 0.01064295, + "balance_loss_clip": 1.13302207, + "balance_loss_mlp": 1.04609144, + "epoch": 0.019238862502254553, + "flos": 20886723797760.0, + "grad_norm": 2.2368126890260265, + "language_loss": 0.8668021, + "learning_rate": 3.6766891159659177e-06, + "loss": 0.89160055, + "num_input_tokens_seen": 3396335, + "step": 160, + "time_per_iteration": 2.723555326461792 + }, + { + "auxiliary_loss_clip": 0.01413066, + "auxiliary_loss_mlp": 0.01061471, + "balance_loss_clip": 1.13208127, + "balance_loss_mlp": 1.04416203, + "epoch": 0.019359105392893646, + "flos": 21360672777600.0, + "grad_norm": 3.4416405446653564, + "language_loss": 0.87812048, + "learning_rate": 3.6812028124990075e-06, + "loss": 0.90286589, + "num_input_tokens_seen": 3413605, + "step": 161, + "time_per_iteration": 2.7417805194854736 + }, + { + "auxiliary_loss_clip": 0.01410575, + "auxiliary_loss_mlp": 0.01072007, + "balance_loss_clip": 1.12930286, + "balance_loss_mlp": 1.05436385, + "epoch": 0.019479348283532736, + "flos": 16283729681280.0, + "grad_norm": 2.7778668622955753, + "language_loss": 0.81684864, + "learning_rate": 3.6856885602676016e-06, + "loss": 0.84167445, + "num_input_tokens_seen": 3429640, + "step": 162, + "time_per_iteration": 2.713186025619507 + }, + { + "auxiliary_loss_clip": 0.01413551, + "auxiliary_loss_mlp": 0.01063381, + "balance_loss_clip": 1.13031769, + "balance_loss_mlp": 1.04490328, + "epoch": 0.019599591174171826, + "flos": 22091239497600.0, + "grad_norm": 2.512429231631629, + "language_loss": 0.94595498, + "learning_rate": 3.6901467032597733e-06, + "loss": 0.97072428, + "num_input_tokens_seen": 3448125, + "step": 163, + "time_per_iteration": 2.785623788833618 + }, + { + "auxiliary_loss_clip": 0.0141156, + "auxiliary_loss_mlp": 0.01063343, + "balance_loss_clip": 1.12859368, + "balance_loss_mlp": 1.0467726, + "epoch": 0.01971983406481092, + "flos": 19609668581760.0, + "grad_norm": 2.5105230218464922, + "language_loss": 0.87361419, + "learning_rate": 3.694577579151804e-06, + "loss": 0.89836323, + "num_input_tokens_seen": 3466535, + "step": 164, + "time_per_iteration": 2.6948752403259277 + }, + { + "auxiliary_loss_clip": 0.01407608, + "auxiliary_loss_mlp": 0.01061692, + "balance_loss_clip": 1.12537348, + "balance_loss_mlp": 1.04406059, + "epoch": 0.01984007695545001, + "flos": 19099342103040.0, + "grad_norm": 2.9043137236240333, + "language_loss": 0.73799461, + "learning_rate": 3.6989815194616703e-06, + "loss": 0.76268762, + "num_input_tokens_seen": 3483730, + "step": 165, + "time_per_iteration": 2.7205522060394287 + }, + { + "auxiliary_loss_clip": 0.01410042, + "auxiliary_loss_mlp": 0.01070892, + "balance_loss_clip": 1.12655973, + "balance_loss_mlp": 1.05224764, + "epoch": 0.0199603198460891, + "flos": 20848406964480.0, + "grad_norm": 2.2098774399712293, + "language_loss": 0.79934692, + "learning_rate": 3.703358849697888e-06, + "loss": 0.82415628, + "num_input_tokens_seen": 3503640, + "step": 166, + "time_per_iteration": 2.775329828262329 + }, + { + "auxiliary_loss_clip": 0.01405553, + "auxiliary_loss_mlp": 0.01061662, + "balance_loss_clip": 1.12666273, + "balance_loss_mlp": 1.04553354, + "epoch": 0.020080562736728192, + "flos": 21870747861120.0, + "grad_norm": 1.8519501446835394, + "language_loss": 0.82693487, + "learning_rate": 3.7077098895038803e-06, + "loss": 0.85160702, + "num_input_tokens_seen": 3523010, + "step": 167, + "time_per_iteration": 2.716157913208008 + }, + { + "auxiliary_loss_clip": 0.01405052, + "auxiliary_loss_mlp": 0.01067625, + "balance_loss_clip": 1.12670302, + "balance_loss_mlp": 1.04968452, + "epoch": 0.020200805627367282, + "flos": 21688788539520.0, + "grad_norm": 25.128506044402435, + "language_loss": 0.97213089, + "learning_rate": 3.712034952798045e-06, + "loss": 0.99685764, + "num_input_tokens_seen": 3541125, + "step": 168, + "time_per_iteration": 2.6387271881103516 + }, + { + "auxiliary_loss_clip": 0.01400073, + "auxiliary_loss_mlp": 0.01071394, + "balance_loss_clip": 1.11867595, + "balance_loss_mlp": 1.05246353, + "epoch": 0.02032104851800637, + "flos": 33543043729920.0, + "grad_norm": 2.2986720030803633, + "language_loss": 0.84620273, + "learning_rate": 3.7163343479096656e-06, + "loss": 0.87091744, + "num_input_tokens_seen": 3562700, + "step": 169, + "time_per_iteration": 2.7436635494232178 + }, + { + "auxiliary_loss_clip": 0.01403319, + "auxiliary_loss_mlp": 0.0105648, + "balance_loss_clip": 1.12668514, + "balance_loss_mlp": 1.04021978, + "epoch": 0.020441291408645465, + "flos": 31686965274240.0, + "grad_norm": 2.2496263833537054, + "language_loss": 0.82828194, + "learning_rate": 3.720608377710802e-06, + "loss": 0.85288, + "num_input_tokens_seen": 3582790, + "step": 170, + "time_per_iteration": 2.7517659664154053 + }, + { + "auxiliary_loss_clip": 0.01396984, + "auxiliary_loss_mlp": 0.01059674, + "balance_loss_clip": 1.12128234, + "balance_loss_mlp": 1.04191172, + "epoch": 0.020561534299284555, + "flos": 20886687884160.0, + "grad_norm": 2.095020631789394, + "language_loss": 0.86895978, + "learning_rate": 3.7248573397443277e-06, + "loss": 0.89352632, + "num_input_tokens_seen": 3601715, + "step": 171, + "time_per_iteration": 4.565625190734863 + }, + { + "auxiliary_loss_clip": 0.01397484, + "auxiliary_loss_mlp": 0.01067238, + "balance_loss_clip": 1.12252712, + "balance_loss_mlp": 1.05021477, + "epoch": 0.020681777189923645, + "flos": 20996610480000.0, + "grad_norm": 2.2087237760096246, + "language_loss": 0.97688687, + "learning_rate": 3.729081526348224e-06, + "loss": 1.0015341, + "num_input_tokens_seen": 3620245, + "step": 172, + "time_per_iteration": 4.567037343978882 + }, + { + "auxiliary_loss_clip": 0.01394605, + "auxiliary_loss_mlp": 0.01061142, + "balance_loss_clip": 1.11858273, + "balance_loss_mlp": 1.04301059, + "epoch": 0.020802020080562738, + "flos": 28257532312320.0, + "grad_norm": 1.9404563877451293, + "language_loss": 0.85078949, + "learning_rate": 3.7332812247762777e-06, + "loss": 0.87534696, + "num_input_tokens_seen": 3641545, + "step": 173, + "time_per_iteration": 2.8146612644195557 + }, + { + "auxiliary_loss_clip": 0.01400134, + "auxiliary_loss_mlp": 0.0106133, + "balance_loss_clip": 1.12523079, + "balance_loss_mlp": 1.04408073, + "epoch": 0.020922262971201828, + "flos": 19681274344320.0, + "grad_norm": 2.5169736208016418, + "language_loss": 0.95642328, + "learning_rate": 3.737456717315293e-06, + "loss": 0.98103791, + "num_input_tokens_seen": 3660510, + "step": 174, + "time_per_iteration": 2.7653772830963135 + }, + { + "auxiliary_loss_clip": 0.01399486, + "auxiliary_loss_mlp": 0.01055188, + "balance_loss_clip": 1.12655938, + "balance_loss_mlp": 1.03868914, + "epoch": 0.021042505861840918, + "flos": 15666353694720.0, + "grad_norm": 1.7503080453917685, + "language_loss": 0.90720505, + "learning_rate": 3.7416082813989552e-06, + "loss": 0.93175179, + "num_input_tokens_seen": 3677505, + "step": 175, + "time_per_iteration": 2.6905086040496826 + }, + { + "auxiliary_loss_clip": 0.01391165, + "auxiliary_loss_mlp": 0.0106623, + "balance_loss_clip": 1.11714315, + "balance_loss_mlp": 1.04993963, + "epoch": 0.02116274875248001, + "flos": 21142012734720.0, + "grad_norm": 1.976102758694034, + "language_loss": 0.89392883, + "learning_rate": 3.745736189718439e-06, + "loss": 0.91850281, + "num_input_tokens_seen": 3696760, + "step": 176, + "time_per_iteration": 2.882361650466919 + }, + { + "auxiliary_loss_clip": 0.01382646, + "auxiliary_loss_mlp": 0.01051036, + "balance_loss_clip": 1.11569095, + "balance_loss_mlp": 1.03627765, + "epoch": 0.0212829916431191, + "flos": 24715770543360.0, + "grad_norm": 3.26847303379397, + "language_loss": 0.72656226, + "learning_rate": 3.749840710329894e-06, + "loss": 0.75089908, + "num_input_tokens_seen": 3717465, + "step": 177, + "time_per_iteration": 2.734989881515503 + }, + { + "auxiliary_loss_clip": 0.01390623, + "auxiliary_loss_mlp": 0.01063766, + "balance_loss_clip": 1.11795354, + "balance_loss_mlp": 1.04640961, + "epoch": 0.02140323453375819, + "flos": 16645493508480.0, + "grad_norm": 4.336749990755572, + "language_loss": 0.97878122, + "learning_rate": 3.7539221067588938e-06, + "loss": 1.0033251, + "num_input_tokens_seen": 3731440, + "step": 178, + "time_per_iteration": 2.7055039405822754 + }, + { + "auxiliary_loss_clip": 0.01392564, + "auxiliary_loss_mlp": 0.01068662, + "balance_loss_clip": 1.11945987, + "balance_loss_mlp": 1.05067301, + "epoch": 0.021523477424397284, + "flos": 20299332689280.0, + "grad_norm": 3.994200438427755, + "language_loss": 0.93881255, + "learning_rate": 3.757980638101964e-06, + "loss": 0.9634248, + "num_input_tokens_seen": 3744935, + "step": 179, + "time_per_iteration": 2.691927909851074 + }, + { + "auxiliary_loss_clip": 0.01389032, + "auxiliary_loss_mlp": 0.01067118, + "balance_loss_clip": 1.11620402, + "balance_loss_mlp": 1.04968929, + "epoch": 0.021643720315036374, + "flos": 26104005331200.0, + "grad_norm": 2.4164055507800373, + "language_loss": 0.89565539, + "learning_rate": 3.7620165591252806e-06, + "loss": 0.92021692, + "num_input_tokens_seen": 3763035, + "step": 180, + "time_per_iteration": 2.7967936992645264 + }, + { + "auxiliary_loss_clip": 0.01385167, + "auxiliary_loss_mlp": 0.0107803, + "balance_loss_clip": 1.11710668, + "balance_loss_mlp": 1.06089926, + "epoch": 0.021763963205675464, + "flos": 24787663614720.0, + "grad_norm": 1.9982514701464826, + "language_loss": 0.94487631, + "learning_rate": 3.766030120360636e-06, + "loss": 0.96950829, + "num_input_tokens_seen": 3782665, + "step": 181, + "time_per_iteration": 2.6932241916656494 + }, + { + "auxiliary_loss_clip": 0.0137958, + "auxiliary_loss_mlp": 0.01064429, + "balance_loss_clip": 1.11325932, + "balance_loss_mlp": 1.04546261, + "epoch": 0.021884206096314557, + "flos": 25813559957760.0, + "grad_norm": 2.396504573475085, + "language_loss": 0.90217137, + "learning_rate": 3.7700215681987578e-06, + "loss": 0.92661148, + "num_input_tokens_seen": 3802435, + "step": 182, + "time_per_iteration": 2.804126262664795 + }, + { + "auxiliary_loss_clip": 0.01383077, + "auxiliary_loss_mlp": 0.01053563, + "balance_loss_clip": 1.11561561, + "balance_loss_mlp": 1.03712416, + "epoch": 0.022004448986953647, + "flos": 20082719721600.0, + "grad_norm": 1.7716518309737963, + "language_loss": 0.82278413, + "learning_rate": 3.7739911449800767e-06, + "loss": 0.84715044, + "num_input_tokens_seen": 3822490, + "step": 183, + "time_per_iteration": 2.6714985370635986 + }, + { + "auxiliary_loss_clip": 0.01388719, + "auxiliary_loss_mlp": 0.01065039, + "balance_loss_clip": 1.1168859, + "balance_loss_mlp": 1.0480274, + "epoch": 0.022124691877592736, + "flos": 20480609652480.0, + "grad_norm": 2.8711278327893854, + "language_loss": 0.80891806, + "learning_rate": 3.7779390890830114e-06, + "loss": 0.83345556, + "num_input_tokens_seen": 3841140, + "step": 184, + "time_per_iteration": 2.760399103164673 + }, + { + "auxiliary_loss_clip": 0.01381139, + "auxiliary_loss_mlp": 0.01059113, + "balance_loss_clip": 1.11276412, + "balance_loss_mlp": 1.04349649, + "epoch": 0.02224493476823183, + "flos": 23586847015680.0, + "grad_norm": 1.7533705167742542, + "language_loss": 0.85867763, + "learning_rate": 3.7818656350098723e-06, + "loss": 0.88308012, + "num_input_tokens_seen": 3862090, + "step": 185, + "time_per_iteration": 2.721395492553711 + }, + { + "auxiliary_loss_clip": 0.01377071, + "auxiliary_loss_mlp": 0.01055663, + "balance_loss_clip": 1.1111989, + "balance_loss_mlp": 1.03858089, + "epoch": 0.02236517765887092, + "flos": 16909940413440.0, + "grad_norm": 2.758267755328107, + "language_loss": 0.77204657, + "learning_rate": 3.7857710134704447e-06, + "loss": 0.79637384, + "num_input_tokens_seen": 3881025, + "step": 186, + "time_per_iteration": 2.70103120803833 + }, + { + "auxiliary_loss_clip": 0.01375348, + "auxiliary_loss_mlp": 0.01062702, + "balance_loss_clip": 1.11230755, + "balance_loss_mlp": 1.04673982, + "epoch": 0.02248542054951001, + "flos": 43508182930560.0, + "grad_norm": 1.968219860555976, + "language_loss": 0.792768, + "learning_rate": 3.7896554514633234e-06, + "loss": 0.81714857, + "num_input_tokens_seen": 3905310, + "step": 187, + "time_per_iteration": 3.0431885719299316 + }, + { + "auxiliary_loss_clip": 0.01373514, + "auxiliary_loss_mlp": 0.01062586, + "balance_loss_clip": 1.11188579, + "balance_loss_mlp": 1.04693401, + "epoch": 0.022605663440149103, + "flos": 23367648268800.0, + "grad_norm": 2.028903452549462, + "language_loss": 0.84118307, + "learning_rate": 3.7935191723550955e-06, + "loss": 0.86554408, + "num_input_tokens_seen": 3924265, + "step": 188, + "time_per_iteration": 2.8781771659851074 + }, + { + "auxiliary_loss_clip": 0.01371959, + "auxiliary_loss_mlp": 0.01071444, + "balance_loss_clip": 1.11061752, + "balance_loss_mlp": 1.0558511, + "epoch": 0.022725906330788193, + "flos": 29019915504000.0, + "grad_norm": 2.2517496607355416, + "language_loss": 0.88799286, + "learning_rate": 3.797362395957408e-06, + "loss": 0.91242695, + "num_input_tokens_seen": 3944830, + "step": 189, + "time_per_iteration": 2.7573230266571045 + }, + { + "auxiliary_loss_clip": 0.01377779, + "auxiliary_loss_mlp": 0.010653, + "balance_loss_clip": 1.11333621, + "balance_loss_mlp": 1.04911172, + "epoch": 0.022846149221427282, + "flos": 24496176746880.0, + "grad_norm": 2.12988571339474, + "language_loss": 0.781654, + "learning_rate": 3.8011853386020055e-06, + "loss": 0.80608475, + "num_input_tokens_seen": 3965735, + "step": 190, + "time_per_iteration": 2.7418322563171387 + }, + { + "auxiliary_loss_clip": 0.0137102, + "auxiliary_loss_mlp": 0.01059666, + "balance_loss_clip": 1.10992217, + "balance_loss_mlp": 1.0445503, + "epoch": 0.022966392112066376, + "flos": 15523537219200.0, + "grad_norm": 2.8819206628413587, + "language_loss": 0.89627492, + "learning_rate": 3.804988213213804e-06, + "loss": 0.92058176, + "num_input_tokens_seen": 3983975, + "step": 191, + "time_per_iteration": 2.713193893432617 + }, + { + "auxiliary_loss_clip": 0.01321368, + "auxiliary_loss_mlp": 0.01012341, + "balance_loss_clip": 1.15047002, + "balance_loss_mlp": 1.00547409, + "epoch": 0.023086635002705466, + "flos": 55650408433920.0, + "grad_norm": 1.014760996274798, + "language_loss": 0.63178027, + "learning_rate": 3.808771229382049e-06, + "loss": 0.65511739, + "num_input_tokens_seen": 4043440, + "step": 192, + "time_per_iteration": 3.167344808578491 + }, + { + "auxiliary_loss_clip": 0.01375371, + "auxiliary_loss_mlp": 0.01061787, + "balance_loss_clip": 1.1120429, + "balance_loss_mlp": 1.04638493, + "epoch": 0.023206877893344555, + "flos": 19313441118720.0, + "grad_norm": 2.6705511444070495, + "language_loss": 0.84686369, + "learning_rate": 3.8125345934296324e-06, + "loss": 0.87123525, + "num_input_tokens_seen": 4061750, + "step": 193, + "time_per_iteration": 2.7078335285186768 + }, + { + "auxiliary_loss_clip": 0.01369603, + "auxiliary_loss_mlp": 0.01062176, + "balance_loss_clip": 1.10704875, + "balance_loss_mlp": 1.04627967, + "epoch": 0.02332712078398365, + "flos": 23072965090560.0, + "grad_norm": 2.106540662265893, + "language_loss": 0.87939942, + "learning_rate": 3.81627850848061e-06, + "loss": 0.90371728, + "num_input_tokens_seen": 4082345, + "step": 194, + "time_per_iteration": 2.7206146717071533 + }, + { + "auxiliary_loss_clip": 0.01367923, + "auxiliary_loss_mlp": 0.01053361, + "balance_loss_clip": 1.10837293, + "balance_loss_mlp": 1.03829265, + "epoch": 0.02344736367462274, + "flos": 24425971614720.0, + "grad_norm": 2.4369839338344965, + "language_loss": 0.86395639, + "learning_rate": 3.820003174525994e-06, + "loss": 0.88816917, + "num_input_tokens_seen": 4101770, + "step": 195, + "time_per_iteration": 2.6707067489624023 + }, + { + "auxiliary_loss_clip": 0.01368594, + "auxiliary_loss_mlp": 0.01058498, + "balance_loss_clip": 1.10968137, + "balance_loss_mlp": 1.0433104, + "epoch": 0.02356760656526183, + "flos": 21579799697280.0, + "grad_norm": 2.3844778636094786, + "language_loss": 0.82746482, + "learning_rate": 3.823708788487851e-06, + "loss": 0.85173577, + "num_input_tokens_seen": 4118770, + "step": 196, + "time_per_iteration": 2.773864507675171 + }, + { + "auxiliary_loss_clip": 0.01364825, + "auxiliary_loss_mlp": 0.01056981, + "balance_loss_clip": 1.10823083, + "balance_loss_mlp": 1.04286647, + "epoch": 0.02368784945590092, + "flos": 25193598192000.0, + "grad_norm": 1.947127639789531, + "language_loss": 0.84669024, + "learning_rate": 3.827395544281781e-06, + "loss": 0.87090832, + "num_input_tokens_seen": 4141110, + "step": 197, + "time_per_iteration": 2.736971616744995 + }, + { + "auxiliary_loss_clip": 0.01369052, + "auxiliary_loss_mlp": 0.01059501, + "balance_loss_clip": 1.11097777, + "balance_loss_mlp": 1.04407549, + "epoch": 0.02380809234654001, + "flos": 27562481164800.0, + "grad_norm": 1.9203811047886539, + "language_loss": 0.78950226, + "learning_rate": 3.831063632877802e-06, + "loss": 0.81378782, + "num_input_tokens_seen": 4161430, + "step": 198, + "time_per_iteration": 4.62250828742981 + }, + { + "auxiliary_loss_clip": 0.01370298, + "auxiliary_loss_mlp": 0.01056398, + "balance_loss_clip": 1.11475766, + "balance_loss_mlp": 1.04226017, + "epoch": 0.0239283352371791, + "flos": 18259786540800.0, + "grad_norm": 2.6139292135865393, + "language_loss": 0.75942528, + "learning_rate": 3.834713242359712e-06, + "loss": 0.78369224, + "num_input_tokens_seen": 4179260, + "step": 199, + "time_per_iteration": 3.58005428314209 + }, + { + "auxiliary_loss_clip": 0.01365847, + "auxiliary_loss_mlp": 0.01060264, + "balance_loss_clip": 1.10648704, + "balance_loss_mlp": 1.0441227, + "epoch": 0.02404857812781819, + "flos": 21395110942080.0, + "grad_norm": 3.143066947586685, + "language_loss": 0.87106884, + "learning_rate": 3.838344557982959e-06, + "loss": 0.89532995, + "num_input_tokens_seen": 4200640, + "step": 200, + "time_per_iteration": 2.7053656578063965 + }, + { + "auxiliary_loss_clip": 0.01357324, + "auxiliary_loss_mlp": 0.01062096, + "balance_loss_clip": 1.10248566, + "balance_loss_mlp": 1.04814887, + "epoch": 0.024168821018457284, + "flos": 16654256426880.0, + "grad_norm": 3.1236754245402794, + "language_loss": 0.84376091, + "learning_rate": 3.841957762231063e-06, + "loss": 0.86795509, + "num_input_tokens_seen": 4218170, + "step": 201, + "time_per_iteration": 2.701277256011963 + }, + { + "auxiliary_loss_clip": 0.01357486, + "auxiliary_loss_mlp": 0.01053472, + "balance_loss_clip": 1.10162556, + "balance_loss_mlp": 1.03796268, + "epoch": 0.024289063909096374, + "flos": 22820872464000.0, + "grad_norm": 1.9935856315476568, + "language_loss": 0.87764692, + "learning_rate": 3.8455530348706454e-06, + "loss": 0.90175653, + "num_input_tokens_seen": 4237770, + "step": 202, + "time_per_iteration": 2.708739995956421 + }, + { + "auxiliary_loss_clip": 0.01363511, + "auxiliary_loss_mlp": 0.01050405, + "balance_loss_clip": 1.1098907, + "balance_loss_mlp": 1.03573084, + "epoch": 0.024409306799735464, + "flos": 17748598135680.0, + "grad_norm": 2.1015233338579797, + "language_loss": 0.77252632, + "learning_rate": 3.849130553005099e-06, + "loss": 0.79666549, + "num_input_tokens_seen": 4255985, + "step": 203, + "time_per_iteration": 2.691192388534546 + }, + { + "auxiliary_loss_clip": 0.01361008, + "auxiliary_loss_mlp": 0.01054209, + "balance_loss_clip": 1.1058358, + "balance_loss_mlp": 1.03908122, + "epoch": 0.024529549690374557, + "flos": 21616213109760.0, + "grad_norm": 1.7481070008999944, + "language_loss": 0.83719546, + "learning_rate": 3.852690491126933e-06, + "loss": 0.86134768, + "num_input_tokens_seen": 4276035, + "step": 204, + "time_per_iteration": 2.7284510135650635 + }, + { + "auxiliary_loss_clip": 0.01362711, + "auxiliary_loss_mlp": 0.01061187, + "balance_loss_clip": 1.10587502, + "balance_loss_mlp": 1.04603553, + "epoch": 0.024649792581013647, + "flos": 25551662918400.0, + "grad_norm": 3.1750541450046716, + "language_loss": 0.91069967, + "learning_rate": 3.856233021168845e-06, + "loss": 0.93493867, + "num_input_tokens_seen": 4295730, + "step": 205, + "time_per_iteration": 2.737128496170044 + }, + { + "auxiliary_loss_clip": 0.01350741, + "auxiliary_loss_mlp": 0.01054018, + "balance_loss_clip": 1.10084009, + "balance_loss_mlp": 1.03959942, + "epoch": 0.024770035471652737, + "flos": 34495574544000.0, + "grad_norm": 2.2293687470857555, + "language_loss": 0.91149247, + "learning_rate": 3.859758312553544e-06, + "loss": 0.93554008, + "num_input_tokens_seen": 4317950, + "step": 206, + "time_per_iteration": 2.807807207107544 + }, + { + "auxiliary_loss_clip": 0.01360609, + "auxiliary_loss_mlp": 0.01055406, + "balance_loss_clip": 1.10787773, + "balance_loss_mlp": 1.04027796, + "epoch": 0.02489027836229183, + "flos": 21505428587520.0, + "grad_norm": 1.7990226888444933, + "language_loss": 0.91946191, + "learning_rate": 3.8632665322423735e-06, + "loss": 0.94362211, + "num_input_tokens_seen": 4337605, + "step": 207, + "time_per_iteration": 2.6524860858917236 + }, + { + "auxiliary_loss_clip": 0.01353387, + "auxiliary_loss_mlp": 0.01056676, + "balance_loss_clip": 1.10148585, + "balance_loss_mlp": 1.04231119, + "epoch": 0.02501052125293092, + "flos": 23219013790080.0, + "grad_norm": 2.426966164429914, + "language_loss": 0.85860586, + "learning_rate": 3.866757844782762e-06, + "loss": 0.88270652, + "num_input_tokens_seen": 4358110, + "step": 208, + "time_per_iteration": 2.726747512817383 + }, + { + "auxiliary_loss_clip": 0.01356936, + "auxiliary_loss_mlp": 0.01061816, + "balance_loss_clip": 1.10407984, + "balance_loss_mlp": 1.0471828, + "epoch": 0.02513076414357001, + "flos": 26388920010240.0, + "grad_norm": 2.2159538458786554, + "language_loss": 0.91186011, + "learning_rate": 3.870232412354527e-06, + "loss": 0.93604761, + "num_input_tokens_seen": 4374955, + "step": 209, + "time_per_iteration": 2.703935384750366 + }, + { + "auxiliary_loss_clip": 0.01355846, + "auxiliary_loss_mlp": 0.01049512, + "balance_loss_clip": 1.10270393, + "balance_loss_mlp": 1.03458667, + "epoch": 0.025251007034209103, + "flos": 13590430047360.0, + "grad_norm": 1.9343410711817621, + "language_loss": 0.92597818, + "learning_rate": 3.873690394815086e-06, + "loss": 0.95003182, + "num_input_tokens_seen": 4391535, + "step": 210, + "time_per_iteration": 2.6869688034057617 + }, + { + "auxiliary_loss_clip": 0.01352364, + "auxiliary_loss_mlp": 0.01064584, + "balance_loss_clip": 1.0998311, + "balance_loss_mlp": 1.04877687, + "epoch": 0.025371249924848193, + "flos": 15049229103360.0, + "grad_norm": 2.250389026877228, + "language_loss": 0.91373515, + "learning_rate": 3.877131949743587e-06, + "loss": 0.93790466, + "num_input_tokens_seen": 4408400, + "step": 211, + "time_per_iteration": 2.6407101154327393 + }, + { + "auxiliary_loss_clip": 0.01350895, + "auxiliary_loss_mlp": 0.01060625, + "balance_loss_clip": 1.09955192, + "balance_loss_mlp": 1.04549754, + "epoch": 0.025491492815487283, + "flos": 25553853648000.0, + "grad_norm": 2.116202532507836, + "language_loss": 0.77923107, + "learning_rate": 3.880557232483993e-06, + "loss": 0.80334628, + "num_input_tokens_seen": 4427840, + "step": 212, + "time_per_iteration": 2.727445602416992 + }, + { + "auxiliary_loss_clip": 0.01355653, + "auxiliary_loss_mlp": 0.01073516, + "balance_loss_clip": 1.10165727, + "balance_loss_mlp": 1.05766141, + "epoch": 0.025611735706126376, + "flos": 20630752502400.0, + "grad_norm": 2.305281130286858, + "language_loss": 0.86760783, + "learning_rate": 3.883966396187164e-06, + "loss": 0.89189953, + "num_input_tokens_seen": 4447110, + "step": 213, + "time_per_iteration": 2.6865460872650146 + }, + { + "auxiliary_loss_clip": 0.01353221, + "auxiliary_loss_mlp": 0.01058073, + "balance_loss_clip": 1.10262012, + "balance_loss_mlp": 1.0430764, + "epoch": 0.025731978596765466, + "flos": 19062282245760.0, + "grad_norm": 1.9109290989689054, + "language_loss": 0.90084136, + "learning_rate": 3.887359591851937e-06, + "loss": 0.92495435, + "num_input_tokens_seen": 4464715, + "step": 214, + "time_per_iteration": 2.679037570953369 + }, + { + "auxiliary_loss_clip": 0.01350696, + "auxiliary_loss_mlp": 0.01052534, + "balance_loss_clip": 1.10145235, + "balance_loss_mlp": 1.03920698, + "epoch": 0.025852221487404556, + "flos": 22163814927360.0, + "grad_norm": 1.6416418831948538, + "language_loss": 0.92327261, + "learning_rate": 3.890736968365265e-06, + "loss": 0.94730484, + "num_input_tokens_seen": 4485030, + "step": 215, + "time_per_iteration": 2.6640326976776123 + }, + { + "auxiliary_loss_clip": 0.01351373, + "auxiliary_loss_mlp": 0.01056772, + "balance_loss_clip": 1.1015085, + "balance_loss_mlp": 1.04218125, + "epoch": 0.02597246437804365, + "flos": 26541971861760.0, + "grad_norm": 9.460204320478706, + "language_loss": 0.84868574, + "learning_rate": 3.894098672541412e-06, + "loss": 0.87276721, + "num_input_tokens_seen": 4505935, + "step": 216, + "time_per_iteration": 2.7159528732299805 + }, + { + "auxiliary_loss_clip": 0.0134702, + "auxiliary_loss_mlp": 0.01054774, + "balance_loss_clip": 1.0973264, + "balance_loss_mlp": 1.03920507, + "epoch": 0.02609270726868274, + "flos": 32671671696000.0, + "grad_norm": 2.052670679499285, + "language_loss": 0.75290227, + "learning_rate": 3.89744484916025e-06, + "loss": 0.7769202, + "num_input_tokens_seen": 4527045, + "step": 217, + "time_per_iteration": 2.7957708835601807 + }, + { + "auxiliary_loss_clip": 0.01350511, + "auxiliary_loss_mlp": 0.01056168, + "balance_loss_clip": 1.09916961, + "balance_loss_mlp": 1.04079044, + "epoch": 0.02621295015932183, + "flos": 26243553669120.0, + "grad_norm": 2.166026634929153, + "language_loss": 0.87481594, + "learning_rate": 3.900775641004673e-06, + "loss": 0.89888275, + "num_input_tokens_seen": 4546360, + "step": 218, + "time_per_iteration": 2.842043399810791 + }, + { + "auxiliary_loss_clip": 0.01358814, + "auxiliary_loss_mlp": 0.01054375, + "balance_loss_clip": 1.10299373, + "balance_loss_mlp": 1.03864563, + "epoch": 0.026333193049960922, + "flos": 42921402353280.0, + "grad_norm": 2.7244196739388435, + "language_loss": 0.74189711, + "learning_rate": 3.904091188897156e-06, + "loss": 0.76602906, + "num_input_tokens_seen": 4565495, + "step": 219, + "time_per_iteration": 2.8627712726593018 + }, + { + "auxiliary_loss_clip": 0.01350067, + "auxiliary_loss_mlp": 0.0105667, + "balance_loss_clip": 1.10173917, + "balance_loss_mlp": 1.04225791, + "epoch": 0.026453435940600012, + "flos": 17963846386560.0, + "grad_norm": 2.115099690714036, + "language_loss": 0.81795448, + "learning_rate": 3.90739163173548e-06, + "loss": 0.84202194, + "num_input_tokens_seen": 4583330, + "step": 220, + "time_per_iteration": 2.6409943103790283 + }, + { + "auxiliary_loss_clip": 0.01347996, + "auxiliary_loss_mlp": 0.01051823, + "balance_loss_clip": 1.09861612, + "balance_loss_mlp": 1.03819704, + "epoch": 0.026573678831239102, + "flos": 18984319776000.0, + "grad_norm": 2.4989597652970623, + "language_loss": 0.88617784, + "learning_rate": 3.910677106527646e-06, + "loss": 0.91017604, + "num_input_tokens_seen": 4600520, + "step": 221, + "time_per_iteration": 2.6702141761779785 + }, + { + "auxiliary_loss_clip": 0.01348353, + "auxiliary_loss_mlp": 0.01066946, + "balance_loss_clip": 1.10157871, + "balance_loss_mlp": 1.05105567, + "epoch": 0.026693921721878195, + "flos": 29241448634880.0, + "grad_norm": 2.1749976003806664, + "language_loss": 0.841658, + "learning_rate": 3.913947748426004e-06, + "loss": 0.86581099, + "num_input_tokens_seen": 4617340, + "step": 222, + "time_per_iteration": 2.724642276763916 + }, + { + "auxiliary_loss_clip": 0.01344916, + "auxiliary_loss_mlp": 0.01050226, + "balance_loss_clip": 1.09891748, + "balance_loss_mlp": 1.03562844, + "epoch": 0.026814164612517285, + "flos": 14128083797760.0, + "grad_norm": 3.8669336641155896, + "language_loss": 0.76537514, + "learning_rate": 3.9172036907606136e-06, + "loss": 0.78932655, + "num_input_tokens_seen": 4630820, + "step": 223, + "time_per_iteration": 3.5696048736572266 + }, + { + "auxiliary_loss_clip": 0.01351941, + "auxiliary_loss_mlp": 0.01057332, + "balance_loss_clip": 1.1022439, + "balance_loss_mlp": 1.04219198, + "epoch": 0.026934407503156375, + "flos": 23511973115520.0, + "grad_norm": 5.304073168014867, + "language_loss": 0.95104957, + "learning_rate": 3.920445065071855e-06, + "loss": 0.97514236, + "num_input_tokens_seen": 4651985, + "step": 224, + "time_per_iteration": 3.619943857192993 + }, + { + "auxiliary_loss_clip": 0.01350372, + "auxiliary_loss_mlp": 0.01053221, + "balance_loss_clip": 1.10036182, + "balance_loss_mlp": 1.03762865, + "epoch": 0.027054650393795468, + "flos": 28950356816640.0, + "grad_norm": 2.9089003696025886, + "language_loss": 0.79842222, + "learning_rate": 3.923672001142322e-06, + "loss": 0.82245815, + "num_input_tokens_seen": 4672295, + "step": 225, + "time_per_iteration": 3.591242551803589 + }, + { + "auxiliary_loss_clip": 0.01349635, + "auxiliary_loss_mlp": 0.01052866, + "balance_loss_clip": 1.10019803, + "balance_loss_mlp": 1.03825045, + "epoch": 0.027174893284434558, + "flos": 31431568596480.0, + "grad_norm": 1.9198712683571613, + "language_loss": 0.8457309, + "learning_rate": 3.926884627027996e-06, + "loss": 0.86975592, + "num_input_tokens_seen": 4696065, + "step": 226, + "time_per_iteration": 2.7085719108581543 + }, + { + "auxiliary_loss_clip": 0.01343514, + "auxiliary_loss_mlp": 0.01052363, + "balance_loss_clip": 1.09650767, + "balance_loss_mlp": 1.03675854, + "epoch": 0.027295136175073648, + "flos": 22054466949120.0, + "grad_norm": 2.7471556865772584, + "language_loss": 0.77543211, + "learning_rate": 3.930083069088744e-06, + "loss": 0.79939085, + "num_input_tokens_seen": 4716065, + "step": 227, + "time_per_iteration": 2.7152795791625977 + }, + { + "auxiliary_loss_clip": 0.01294442, + "auxiliary_loss_mlp": 0.01010351, + "balance_loss_clip": 1.13552809, + "balance_loss_mlp": 1.00405633, + "epoch": 0.02741537906571274, + "flos": 60800752972800.0, + "grad_norm": 0.982946146488047, + "language_loss": 0.59285885, + "learning_rate": 3.933267452018137e-06, + "loss": 0.61590677, + "num_input_tokens_seen": 4775860, + "step": 228, + "time_per_iteration": 3.226308822631836 + }, + { + "auxiliary_loss_clip": 0.01346216, + "auxiliary_loss_mlp": 0.01058412, + "balance_loss_clip": 1.09895706, + "balance_loss_mlp": 1.04319525, + "epoch": 0.02753562195635183, + "flos": 24606278910720.0, + "grad_norm": 2.547810458863279, + "language_loss": 0.84476227, + "learning_rate": 3.936437898872622e-06, + "loss": 0.86880851, + "num_input_tokens_seen": 4795835, + "step": 229, + "time_per_iteration": 2.6901509761810303 + }, + { + "auxiliary_loss_clip": 0.01344473, + "auxiliary_loss_mlp": 0.01060243, + "balance_loss_clip": 1.09930861, + "balance_loss_mlp": 1.04368472, + "epoch": 0.02765586484699092, + "flos": 34094236907520.0, + "grad_norm": 3.141017320831649, + "language_loss": 0.79737258, + "learning_rate": 3.9395945311000525e-06, + "loss": 0.82141972, + "num_input_tokens_seen": 4817460, + "step": 230, + "time_per_iteration": 2.762923002243042 + }, + { + "auxiliary_loss_clip": 0.01345257, + "auxiliary_loss_mlp": 0.01058089, + "balance_loss_clip": 1.09991872, + "balance_loss_mlp": 1.04223394, + "epoch": 0.027776107737630014, + "flos": 14829922615680.0, + "grad_norm": 2.731867278947274, + "language_loss": 0.9098345, + "learning_rate": 3.942737468567608e-06, + "loss": 0.93386793, + "num_input_tokens_seen": 4835475, + "step": 231, + "time_per_iteration": 2.694908380508423 + }, + { + "auxiliary_loss_clip": 0.01343292, + "auxiliary_loss_mlp": 0.01049448, + "balance_loss_clip": 1.09799826, + "balance_loss_mlp": 1.03446901, + "epoch": 0.027896350628269104, + "flos": 47920347066240.0, + "grad_norm": 2.686162408830448, + "language_loss": 0.8635546, + "learning_rate": 3.9458668295891026e-06, + "loss": 0.88748193, + "num_input_tokens_seen": 4857760, + "step": 232, + "time_per_iteration": 2.880542039871216 + }, + { + "auxiliary_loss_clip": 0.01342377, + "auxiliary_loss_mlp": 0.01049406, + "balance_loss_clip": 1.09724379, + "balance_loss_mlp": 1.03454113, + "epoch": 0.028016593518908194, + "flos": 21684550734720.0, + "grad_norm": 2.1969705065643015, + "language_loss": 0.86786014, + "learning_rate": 3.948982730951712e-06, + "loss": 0.89177799, + "num_input_tokens_seen": 4875855, + "step": 233, + "time_per_iteration": 2.7622835636138916 + }, + { + "auxiliary_loss_clip": 0.01350369, + "auxiliary_loss_mlp": 0.01053804, + "balance_loss_clip": 1.10084271, + "balance_loss_mlp": 1.03778243, + "epoch": 0.028136836409547287, + "flos": 18439483305600.0, + "grad_norm": 2.2658017164542312, + "language_loss": 0.82178175, + "learning_rate": 3.9520852879421254e-06, + "loss": 0.84582353, + "num_input_tokens_seen": 4893200, + "step": 234, + "time_per_iteration": 2.6530425548553467 + }, + { + "auxiliary_loss_clip": 0.01342645, + "auxiliary_loss_mlp": 0.01046098, + "balance_loss_clip": 1.09832716, + "balance_loss_mlp": 1.03158998, + "epoch": 0.028257079300186377, + "flos": 31576934937600.0, + "grad_norm": 2.4562586152342853, + "language_loss": 0.81622946, + "learning_rate": 3.955174614372137e-06, + "loss": 0.84011686, + "num_input_tokens_seen": 4912965, + "step": 235, + "time_per_iteration": 2.7360708713531494 + }, + { + "auxiliary_loss_clip": 0.01345726, + "auxiliary_loss_mlp": 0.01044479, + "balance_loss_clip": 1.10118413, + "balance_loss_mlp": 1.02905345, + "epoch": 0.028377322190825467, + "flos": 23513337832320.0, + "grad_norm": 2.322059255886768, + "language_loss": 0.84258842, + "learning_rate": 3.9582508226037045e-06, + "loss": 0.86649048, + "num_input_tokens_seen": 4933105, + "step": 236, + "time_per_iteration": 2.737420082092285 + }, + { + "auxiliary_loss_clip": 0.01349512, + "auxiliary_loss_mlp": 0.01053434, + "balance_loss_clip": 1.09908199, + "balance_loss_mlp": 1.03765059, + "epoch": 0.02849756508146456, + "flos": 20479604071680.0, + "grad_norm": 2.490439469771005, + "language_loss": 0.94158798, + "learning_rate": 3.9613140235734636e-06, + "loss": 0.96561742, + "num_input_tokens_seen": 4950085, + "step": 237, + "time_per_iteration": 2.627868890762329 + }, + { + "auxiliary_loss_clip": 0.01342181, + "auxiliary_loss_mlp": 0.01064519, + "balance_loss_clip": 1.0982964, + "balance_loss_mlp": 1.05017185, + "epoch": 0.02861780797210365, + "flos": 14283362292480.0, + "grad_norm": 2.4509775116959553, + "language_loss": 0.8132, + "learning_rate": 3.96436432681674e-06, + "loss": 0.83726698, + "num_input_tokens_seen": 4968075, + "step": 238, + "time_per_iteration": 2.6665332317352295 + }, + { + "auxiliary_loss_clip": 0.01340043, + "auxiliary_loss_mlp": 0.01053252, + "balance_loss_clip": 1.09741926, + "balance_loss_mlp": 1.03823137, + "epoch": 0.02873805086274274, + "flos": 25808532053760.0, + "grad_norm": 3.237851825223377, + "language_loss": 0.88933849, + "learning_rate": 3.967401840491044e-06, + "loss": 0.91327143, + "num_input_tokens_seen": 4987355, + "step": 239, + "time_per_iteration": 0.018114328384399414 + }, + { + "auxiliary_loss_clip": 0.01341683, + "auxiliary_loss_mlp": 0.01051703, + "balance_loss_clip": 1.09982848, + "balance_loss_mlp": 1.03757668, + "epoch": 0.028858293753381833, + "flos": 17304238984320.0, + "grad_norm": 2.2854932885537305, + "language_loss": 0.88070738, + "learning_rate": 3.97042667139909e-06, + "loss": 0.90464127, + "num_input_tokens_seen": 5004680, + "step": 240, + "time_per_iteration": 2.681523323059082 + }, + { + "auxiliary_loss_clip": 0.01339038, + "auxiliary_loss_mlp": 0.01057862, + "balance_loss_clip": 1.09906077, + "balance_loss_mlp": 1.04243612, + "epoch": 0.028978536644020923, + "flos": 23038347358080.0, + "grad_norm": 2.390598708859977, + "language_loss": 0.87467515, + "learning_rate": 3.973438925011327e-06, + "loss": 0.89864409, + "num_input_tokens_seen": 5022965, + "step": 241, + "time_per_iteration": 2.6437644958496094 + }, + { + "auxiliary_loss_clip": 0.01341336, + "auxiliary_loss_mlp": 0.01043885, + "balance_loss_clip": 1.09697044, + "balance_loss_mlp": 1.02887642, + "epoch": 0.029098779534660012, + "flos": 28329712692480.0, + "grad_norm": 2.3622140655135415, + "language_loss": 0.91546607, + "learning_rate": 3.976438705488002e-06, + "loss": 0.9393183, + "num_input_tokens_seen": 5042625, + "step": 242, + "time_per_iteration": 2.739619493484497 + }, + { + "auxiliary_loss_clip": 0.01339141, + "auxiliary_loss_mlp": 0.01063455, + "balance_loss_clip": 1.09839153, + "balance_loss_mlp": 1.04799402, + "epoch": 0.029219022425299106, + "flos": 13881665520000.0, + "grad_norm": 3.204884689864057, + "language_loss": 0.93013251, + "learning_rate": 3.9794261157007744e-06, + "loss": 0.95415843, + "num_input_tokens_seen": 5060380, + "step": 243, + "time_per_iteration": 2.7517921924591064 + }, + { + "auxiliary_loss_clip": 0.01345059, + "auxiliary_loss_mlp": 0.01055729, + "balance_loss_clip": 1.10045397, + "balance_loss_mlp": 1.03950453, + "epoch": 0.029339265315938196, + "flos": 19422501788160.0, + "grad_norm": 2.174167068581041, + "language_loss": 0.84665525, + "learning_rate": 3.982401257253887e-06, + "loss": 0.87066311, + "num_input_tokens_seen": 5078720, + "step": 244, + "time_per_iteration": 2.714397668838501 + }, + { + "auxiliary_loss_clip": 0.01341253, + "auxiliary_loss_mlp": 0.01051076, + "balance_loss_clip": 1.09828413, + "balance_loss_mlp": 1.03561485, + "epoch": 0.029459508206577285, + "flos": 15669550005120.0, + "grad_norm": 2.0694228536415618, + "language_loss": 0.89407104, + "learning_rate": 3.985364230504893e-06, + "loss": 0.91799432, + "num_input_tokens_seen": 5096605, + "step": 245, + "time_per_iteration": 2.6182732582092285 + }, + { + "auxiliary_loss_clip": 0.01338853, + "auxiliary_loss_mlp": 0.01062094, + "balance_loss_clip": 1.09700501, + "balance_loss_mlp": 1.04637074, + "epoch": 0.02957975109721638, + "flos": 28220975245440.0, + "grad_norm": 2.086805995980599, + "language_loss": 0.841887, + "learning_rate": 3.988315134584976e-06, + "loss": 0.86589646, + "num_input_tokens_seen": 5116285, + "step": 246, + "time_per_iteration": 2.780465602874756 + }, + { + "auxiliary_loss_clip": 0.01341313, + "auxiliary_loss_mlp": 0.01055845, + "balance_loss_clip": 1.09906268, + "balance_loss_mlp": 1.04000199, + "epoch": 0.02969999398785547, + "flos": 24315869450880.0, + "grad_norm": 1.7546851826729368, + "language_loss": 0.80480319, + "learning_rate": 3.991254067418851e-06, + "loss": 0.82877469, + "num_input_tokens_seen": 5136825, + "step": 247, + "time_per_iteration": 2.6979424953460693 + }, + { + "auxiliary_loss_clip": 0.01337713, + "auxiliary_loss_mlp": 0.01055764, + "balance_loss_clip": 1.09807861, + "balance_loss_mlp": 1.04155445, + "epoch": 0.02982023687849456, + "flos": 35078584193280.0, + "grad_norm": 2.023760068913404, + "language_loss": 0.83029139, + "learning_rate": 3.994181125744254e-06, + "loss": 0.85422611, + "num_input_tokens_seen": 5158630, + "step": 248, + "time_per_iteration": 2.7561774253845215 + }, + { + "auxiliary_loss_clip": 0.01333846, + "auxiliary_loss_mlp": 0.0105445, + "balance_loss_clip": 1.09581196, + "balance_loss_mlp": 1.04055047, + "epoch": 0.02994047976913365, + "flos": 26177155378560.0, + "grad_norm": 2.1632866017062464, + "language_loss": 0.73896509, + "learning_rate": 3.99709640513106e-06, + "loss": 0.76284802, + "num_input_tokens_seen": 5179510, + "step": 249, + "time_per_iteration": 2.686190605163574 + }, + { + "auxiliary_loss_clip": 0.01339694, + "auxiliary_loss_mlp": 0.01053676, + "balance_loss_clip": 1.09666014, + "balance_loss_mlp": 1.03794074, + "epoch": 0.03006072265977274, + "flos": 25625028447360.0, + "grad_norm": 3.453565838454771, + "language_loss": 0.85480922, + "learning_rate": 4e-06, + "loss": 0.87874299, + "num_input_tokens_seen": 5199345, + "step": 250, + "time_per_iteration": 4.538034439086914 + }, + { + "auxiliary_loss_clip": 0.01336879, + "auxiliary_loss_mlp": 0.01057412, + "balance_loss_clip": 1.09775543, + "balance_loss_mlp": 1.04277921, + "epoch": 0.03018096555041183, + "flos": 22127078292480.0, + "grad_norm": 3.308302740888094, + "language_loss": 0.88360703, + "learning_rate": 3.999999848300794e-06, + "loss": 0.90754992, + "num_input_tokens_seen": 5218330, + "step": 251, + "time_per_iteration": 3.543375253677368 + }, + { + "auxiliary_loss_clip": 0.01335282, + "auxiliary_loss_mlp": 0.01049447, + "balance_loss_clip": 1.0947926, + "balance_loss_mlp": 1.03546381, + "epoch": 0.030301208441050925, + "flos": 30188197359360.0, + "grad_norm": 1.6256796727422218, + "language_loss": 0.88932729, + "learning_rate": 3.999999393203203e-06, + "loss": 0.91317463, + "num_input_tokens_seen": 5240740, + "step": 252, + "time_per_iteration": 2.727595806121826 + }, + { + "auxiliary_loss_clip": 0.01336104, + "auxiliary_loss_mlp": 0.01049659, + "balance_loss_clip": 1.09706759, + "balance_loss_mlp": 1.03461528, + "epoch": 0.030421451331690014, + "flos": 23621392920960.0, + "grad_norm": 2.0704301085418897, + "language_loss": 0.8504563, + "learning_rate": 3.999998634707293e-06, + "loss": 0.87431395, + "num_input_tokens_seen": 5260290, + "step": 253, + "time_per_iteration": 2.62968111038208 + }, + { + "auxiliary_loss_clip": 0.01339607, + "auxiliary_loss_mlp": 0.01059539, + "balance_loss_clip": 1.09919596, + "balance_loss_mlp": 1.04346967, + "epoch": 0.030541694222329104, + "flos": 27928446883200.0, + "grad_norm": 2.2778057863835155, + "language_loss": 0.96407402, + "learning_rate": 3.999997572813182e-06, + "loss": 0.98806548, + "num_input_tokens_seen": 5278100, + "step": 254, + "time_per_iteration": 2.69323992729187 + }, + { + "auxiliary_loss_clip": 0.01334615, + "auxiliary_loss_mlp": 0.01049294, + "balance_loss_clip": 1.09590483, + "balance_loss_mlp": 1.03482246, + "epoch": 0.030661937112968194, + "flos": 18588441006720.0, + "grad_norm": 1.896984991709446, + "language_loss": 0.87687641, + "learning_rate": 3.999996207521028e-06, + "loss": 0.90071547, + "num_input_tokens_seen": 5296810, + "step": 255, + "time_per_iteration": 2.6010689735412598 + }, + { + "auxiliary_loss_clip": 0.0133518, + "auxiliary_loss_mlp": 0.01057149, + "balance_loss_clip": 1.09528291, + "balance_loss_mlp": 1.04163957, + "epoch": 0.030782180003607287, + "flos": 12969139478400.0, + "grad_norm": 2.2218159050655957, + "language_loss": 0.82083851, + "learning_rate": 3.999994538831039e-06, + "loss": 0.84476179, + "num_input_tokens_seen": 5313395, + "step": 256, + "time_per_iteration": 2.62636661529541 + }, + { + "auxiliary_loss_clip": 0.0133297, + "auxiliary_loss_mlp": 0.01058779, + "balance_loss_clip": 1.09451115, + "balance_loss_mlp": 1.04428864, + "epoch": 0.030902422894246377, + "flos": 23335364920320.0, + "grad_norm": 11.67173623595197, + "language_loss": 0.85627842, + "learning_rate": 3.99999256674347e-06, + "loss": 0.88019586, + "num_input_tokens_seen": 5333545, + "step": 257, + "time_per_iteration": 2.6845614910125732 + }, + { + "auxiliary_loss_clip": 0.01267139, + "auxiliary_loss_mlp": 0.01004587, + "balance_loss_clip": 1.11199784, + "balance_loss_mlp": 0.99862689, + "epoch": 0.031022665784885467, + "flos": 55094151438720.0, + "grad_norm": 1.0150008814587979, + "language_loss": 0.53496408, + "learning_rate": 3.999990291258618e-06, + "loss": 0.55768144, + "num_input_tokens_seen": 5392235, + "step": 258, + "time_per_iteration": 3.184180498123169 + }, + { + "auxiliary_loss_clip": 0.01338253, + "auxiliary_loss_mlp": 0.01060743, + "balance_loss_clip": 1.10000277, + "balance_loss_mlp": 1.04562759, + "epoch": 0.03114290867552456, + "flos": 19317786664320.0, + "grad_norm": 2.3348444859032202, + "language_loss": 0.86402541, + "learning_rate": 3.999987712376829e-06, + "loss": 0.88801539, + "num_input_tokens_seen": 5410555, + "step": 259, + "time_per_iteration": 2.678488254547119 + }, + { + "auxiliary_loss_clip": 0.01332569, + "auxiliary_loss_mlp": 0.01065317, + "balance_loss_clip": 1.09587288, + "balance_loss_mlp": 1.05149484, + "epoch": 0.031263151566163654, + "flos": 20959442881920.0, + "grad_norm": 1.9202520456930272, + "language_loss": 0.82101572, + "learning_rate": 3.999984830098494e-06, + "loss": 0.84499454, + "num_input_tokens_seen": 5430135, + "step": 260, + "time_per_iteration": 2.6474626064300537 + }, + { + "auxiliary_loss_clip": 0.01329711, + "auxiliary_loss_mlp": 0.01055303, + "balance_loss_clip": 1.09232235, + "balance_loss_mlp": 1.04137909, + "epoch": 0.03138339445680274, + "flos": 14793006412800.0, + "grad_norm": 3.180392566266878, + "language_loss": 0.97718728, + "learning_rate": 3.999981644424051e-06, + "loss": 1.00103748, + "num_input_tokens_seen": 5444935, + "step": 261, + "time_per_iteration": 2.659151315689087 + }, + { + "auxiliary_loss_clip": 0.01334359, + "auxiliary_loss_mlp": 0.01051564, + "balance_loss_clip": 1.09691453, + "balance_loss_mlp": 1.03700888, + "epoch": 0.03150363734744183, + "flos": 11655599022720.0, + "grad_norm": 2.5985951166025334, + "language_loss": 0.86190969, + "learning_rate": 3.999978155353982e-06, + "loss": 0.88576889, + "num_input_tokens_seen": 5462080, + "step": 262, + "time_per_iteration": 2.7226550579071045 + }, + { + "auxiliary_loss_clip": 0.01329129, + "auxiliary_loss_mlp": 0.01058476, + "balance_loss_clip": 1.09097934, + "balance_loss_mlp": 1.04377127, + "epoch": 0.03162388023808092, + "flos": 33727732485120.0, + "grad_norm": 5.478802941736048, + "language_loss": 0.80147207, + "learning_rate": 3.9999743628888186e-06, + "loss": 0.82534814, + "num_input_tokens_seen": 5483870, + "step": 263, + "time_per_iteration": 2.732710361480713 + }, + { + "auxiliary_loss_clip": 0.0132891, + "auxiliary_loss_mlp": 0.01059359, + "balance_loss_clip": 1.09167361, + "balance_loss_mlp": 1.04388559, + "epoch": 0.03174412312872001, + "flos": 20810952057600.0, + "grad_norm": 2.212064564021404, + "language_loss": 0.89475316, + "learning_rate": 3.999970267029133e-06, + "loss": 0.91863585, + "num_input_tokens_seen": 5502830, + "step": 264, + "time_per_iteration": 2.647722005844116 + }, + { + "auxiliary_loss_clip": 0.01327458, + "auxiliary_loss_mlp": 0.01054764, + "balance_loss_clip": 1.09328365, + "balance_loss_mlp": 1.04095364, + "epoch": 0.0318643660193591, + "flos": 23727939638400.0, + "grad_norm": 1.923797836887217, + "language_loss": 0.80032778, + "learning_rate": 3.999965867775548e-06, + "loss": 0.82414997, + "num_input_tokens_seen": 5523225, + "step": 265, + "time_per_iteration": 2.700697660446167 + }, + { + "auxiliary_loss_clip": 0.01324071, + "auxiliary_loss_mlp": 0.0105931, + "balance_loss_clip": 1.09037375, + "balance_loss_mlp": 1.04474831, + "epoch": 0.0319846089099982, + "flos": 13917863450880.0, + "grad_norm": 2.682816862472668, + "language_loss": 0.86927426, + "learning_rate": 3.9999611651287315e-06, + "loss": 0.89310807, + "num_input_tokens_seen": 5541380, + "step": 266, + "time_per_iteration": 2.612830877304077 + }, + { + "auxiliary_loss_clip": 0.01329046, + "auxiliary_loss_mlp": 0.01055633, + "balance_loss_clip": 1.09299922, + "balance_loss_mlp": 1.04082727, + "epoch": 0.03210485180063729, + "flos": 14753253035520.0, + "grad_norm": 2.512670108770371, + "language_loss": 0.78644943, + "learning_rate": 3.999956159089396e-06, + "loss": 0.81029624, + "num_input_tokens_seen": 5558830, + "step": 267, + "time_per_iteration": 2.6167824268341064 + }, + { + "auxiliary_loss_clip": 0.01329591, + "auxiliary_loss_mlp": 0.01066596, + "balance_loss_clip": 1.09551108, + "balance_loss_mlp": 1.05175495, + "epoch": 0.03222509469127638, + "flos": 28913153304960.0, + "grad_norm": 2.5095861065164073, + "language_loss": 0.79645407, + "learning_rate": 3.999950849658302e-06, + "loss": 0.82041597, + "num_input_tokens_seen": 5577750, + "step": 268, + "time_per_iteration": 2.7161495685577393 + }, + { + "auxiliary_loss_clip": 0.01330178, + "auxiliary_loss_mlp": 0.01058056, + "balance_loss_clip": 1.09329104, + "balance_loss_mlp": 1.04403663, + "epoch": 0.03234533758191547, + "flos": 16946389739520.0, + "grad_norm": 3.0530387595615243, + "language_loss": 0.84054255, + "learning_rate": 3.999945236836254e-06, + "loss": 0.86442482, + "num_input_tokens_seen": 5596715, + "step": 269, + "time_per_iteration": 2.599591016769409 + }, + { + "auxiliary_loss_clip": 0.01332774, + "auxiliary_loss_mlp": 0.01053487, + "balance_loss_clip": 1.0954591, + "balance_loss_mlp": 1.0385735, + "epoch": 0.03246558047255456, + "flos": 18989096284800.0, + "grad_norm": 5.354781927929036, + "language_loss": 0.94995075, + "learning_rate": 3.999939320624103e-06, + "loss": 0.97381341, + "num_input_tokens_seen": 5611865, + "step": 270, + "time_per_iteration": 2.585845708847046 + }, + { + "auxiliary_loss_clip": 0.01322728, + "auxiliary_loss_mlp": 0.01065049, + "balance_loss_clip": 1.09244967, + "balance_loss_mlp": 1.04982615, + "epoch": 0.03258582336319365, + "flos": 23728334688000.0, + "grad_norm": 3.854429789775514, + "language_loss": 0.89705098, + "learning_rate": 3.999933101022749e-06, + "loss": 0.92092878, + "num_input_tokens_seen": 5632270, + "step": 271, + "time_per_iteration": 2.7589218616485596 + }, + { + "auxiliary_loss_clip": 0.01332448, + "auxiliary_loss_mlp": 0.01052869, + "balance_loss_clip": 1.09685087, + "balance_loss_mlp": 1.03840327, + "epoch": 0.032706066253832745, + "flos": 27670823562240.0, + "grad_norm": 1.9742924802911415, + "language_loss": 0.86924839, + "learning_rate": 3.999926578033132e-06, + "loss": 0.89310157, + "num_input_tokens_seen": 5652085, + "step": 272, + "time_per_iteration": 2.6617178916931152 + }, + { + "auxiliary_loss_clip": 0.01327117, + "auxiliary_loss_mlp": 0.01055436, + "balance_loss_clip": 1.09121978, + "balance_loss_mlp": 1.03924739, + "epoch": 0.032826309144471835, + "flos": 45624685968000.0, + "grad_norm": 2.2195503889936568, + "language_loss": 0.63004243, + "learning_rate": 3.999919751656244e-06, + "loss": 0.65386802, + "num_input_tokens_seen": 5678985, + "step": 273, + "time_per_iteration": 2.868218183517456 + }, + { + "auxiliary_loss_clip": 0.01324086, + "auxiliary_loss_mlp": 0.01058573, + "balance_loss_clip": 1.09288442, + "balance_loss_mlp": 1.04255104, + "epoch": 0.032946552035110925, + "flos": 25812374808960.0, + "grad_norm": 2.529698726771498, + "language_loss": 0.75764, + "learning_rate": 3.9999126218931195e-06, + "loss": 0.7814666, + "num_input_tokens_seen": 5697020, + "step": 274, + "time_per_iteration": 2.7508387565612793 + }, + { + "auxiliary_loss_clip": 0.01328244, + "auxiliary_loss_mlp": 0.01060912, + "balance_loss_clip": 1.09545255, + "balance_loss_mlp": 1.04713142, + "epoch": 0.033066794925750015, + "flos": 15121984101120.0, + "grad_norm": 2.776467699451331, + "language_loss": 0.89650035, + "learning_rate": 3.99990518874484e-06, + "loss": 0.92039192, + "num_input_tokens_seen": 5713460, + "step": 275, + "time_per_iteration": 2.6673946380615234 + }, + { + "auxiliary_loss_clip": 0.01327935, + "auxiliary_loss_mlp": 0.01061773, + "balance_loss_clip": 1.09484816, + "balance_loss_mlp": 1.04755127, + "epoch": 0.033187037816389105, + "flos": 22776593973120.0, + "grad_norm": 3.4134569555191163, + "language_loss": 0.92681563, + "learning_rate": 3.999897452212534e-06, + "loss": 0.95071268, + "num_input_tokens_seen": 5730790, + "step": 276, + "time_per_iteration": 3.5454299449920654 + }, + { + "auxiliary_loss_clip": 0.01330176, + "auxiliary_loss_mlp": 0.01045641, + "balance_loss_clip": 1.09313691, + "balance_loss_mlp": 1.03050184, + "epoch": 0.033307280707028195, + "flos": 23331414424320.0, + "grad_norm": 2.1148381877171594, + "language_loss": 1.00128508, + "learning_rate": 3.999889412297374e-06, + "loss": 1.02504325, + "num_input_tokens_seen": 5750215, + "step": 277, + "time_per_iteration": 3.5809473991394043 + }, + { + "auxiliary_loss_clip": 0.01325412, + "auxiliary_loss_mlp": 0.01060674, + "balance_loss_clip": 1.09203768, + "balance_loss_mlp": 1.0453558, + "epoch": 0.03342752359766729, + "flos": 28840290566400.0, + "grad_norm": 3.866682466045861, + "language_loss": 0.78948897, + "learning_rate": 3.999881069000581e-06, + "loss": 0.81334984, + "num_input_tokens_seen": 5769945, + "step": 278, + "time_per_iteration": 2.6851370334625244 + }, + { + "auxiliary_loss_clip": 0.01322594, + "auxiliary_loss_mlp": 0.0105377, + "balance_loss_clip": 1.09002376, + "balance_loss_mlp": 1.03857112, + "epoch": 0.03354776648830638, + "flos": 19384544090880.0, + "grad_norm": 2.756139086218236, + "language_loss": 0.86676461, + "learning_rate": 3.99987242232342e-06, + "loss": 0.8905282, + "num_input_tokens_seen": 5784950, + "step": 279, + "time_per_iteration": 2.6347029209136963 + }, + { + "auxiliary_loss_clip": 0.01328089, + "auxiliary_loss_mlp": 0.01065647, + "balance_loss_clip": 1.09546971, + "balance_loss_mlp": 1.05068588, + "epoch": 0.03366800937894547, + "flos": 17858628472320.0, + "grad_norm": 2.4724564933496365, + "language_loss": 0.79464638, + "learning_rate": 3.9998634722672026e-06, + "loss": 0.81858373, + "num_input_tokens_seen": 5805005, + "step": 280, + "time_per_iteration": 2.7160584926605225 + }, + { + "auxiliary_loss_clip": 0.01326873, + "auxiliary_loss_mlp": 0.01055396, + "balance_loss_clip": 1.09353781, + "balance_loss_mlp": 1.04132962, + "epoch": 0.03378825226958456, + "flos": 35951033635200.0, + "grad_norm": 2.5586419192768637, + "language_loss": 0.7856822, + "learning_rate": 3.999854218833286e-06, + "loss": 0.80950493, + "num_input_tokens_seen": 5825825, + "step": 281, + "time_per_iteration": 2.7886548042297363 + }, + { + "auxiliary_loss_clip": 0.01324929, + "auxiliary_loss_mlp": 0.01055037, + "balance_loss_clip": 1.09307837, + "balance_loss_mlp": 1.04118514, + "epoch": 0.03390849516022365, + "flos": 25702488126720.0, + "grad_norm": 1.973426291927537, + "language_loss": 0.82147652, + "learning_rate": 3.999844662023075e-06, + "loss": 0.84527618, + "num_input_tokens_seen": 5845700, + "step": 282, + "time_per_iteration": 2.6806318759918213 + }, + { + "auxiliary_loss_clip": 0.01319966, + "auxiliary_loss_mlp": 0.0105536, + "balance_loss_clip": 1.09219742, + "balance_loss_mlp": 1.04138827, + "epoch": 0.03402873805086274, + "flos": 21284505987840.0, + "grad_norm": 2.4752122992211016, + "language_loss": 0.92159081, + "learning_rate": 3.999834801838018e-06, + "loss": 0.94534409, + "num_input_tokens_seen": 5864680, + "step": 283, + "time_per_iteration": 2.6267974376678467 + }, + { + "auxiliary_loss_clip": 0.01320739, + "auxiliary_loss_mlp": 0.01045006, + "balance_loss_clip": 1.09322834, + "balance_loss_mlp": 1.03049827, + "epoch": 0.03414898094150183, + "flos": 22710913954560.0, + "grad_norm": 1.8504925395315959, + "language_loss": 0.73902237, + "learning_rate": 3.9998246382796115e-06, + "loss": 0.76267976, + "num_input_tokens_seen": 5884260, + "step": 284, + "time_per_iteration": 2.6083638668060303 + }, + { + "auxiliary_loss_clip": 0.01327032, + "auxiliary_loss_mlp": 0.01065252, + "balance_loss_clip": 1.0933156, + "balance_loss_mlp": 1.0502913, + "epoch": 0.03426922383214093, + "flos": 18879927874560.0, + "grad_norm": 2.2198417640985038, + "language_loss": 0.90919793, + "learning_rate": 3.999814171349399e-06, + "loss": 0.93312073, + "num_input_tokens_seen": 5902120, + "step": 285, + "time_per_iteration": 2.66696834564209 + }, + { + "auxiliary_loss_clip": 0.01319532, + "auxiliary_loss_mlp": 0.01065958, + "balance_loss_clip": 1.08913326, + "balance_loss_mlp": 1.05149198, + "epoch": 0.03438946672278002, + "flos": 34752012716160.0, + "grad_norm": 1.6683656394211213, + "language_loss": 0.73609507, + "learning_rate": 3.9998034010489655e-06, + "loss": 0.75994998, + "num_input_tokens_seen": 5925810, + "step": 286, + "time_per_iteration": 2.7041008472442627 + }, + { + "auxiliary_loss_clip": 0.01327643, + "auxiliary_loss_mlp": 0.01049444, + "balance_loss_clip": 1.09740543, + "balance_loss_mlp": 1.03540123, + "epoch": 0.03450970961341911, + "flos": 22164102236160.0, + "grad_norm": 2.3965787733240598, + "language_loss": 0.75972819, + "learning_rate": 3.999792327379946e-06, + "loss": 0.78349912, + "num_input_tokens_seen": 5945185, + "step": 287, + "time_per_iteration": 2.616563320159912 + }, + { + "auxiliary_loss_clip": 0.01324618, + "auxiliary_loss_mlp": 0.01057538, + "balance_loss_clip": 1.09523225, + "balance_loss_mlp": 1.04303002, + "epoch": 0.034629952504058197, + "flos": 21725740656000.0, + "grad_norm": 2.380724942557271, + "language_loss": 0.96633935, + "learning_rate": 3.999780950344021e-06, + "loss": 0.99016088, + "num_input_tokens_seen": 5963375, + "step": 288, + "time_per_iteration": 2.6032893657684326 + }, + { + "auxiliary_loss_clip": 0.01326535, + "auxiliary_loss_mlp": 0.01055196, + "balance_loss_clip": 1.09305072, + "balance_loss_mlp": 1.04115915, + "epoch": 0.034750195394697286, + "flos": 20047994248320.0, + "grad_norm": 1.9105777645225372, + "language_loss": 0.82639277, + "learning_rate": 3.999769269942916e-06, + "loss": 0.85021007, + "num_input_tokens_seen": 5983415, + "step": 289, + "time_per_iteration": 2.6502907276153564 + }, + { + "auxiliary_loss_clip": 0.01325465, + "auxiliary_loss_mlp": 0.01060317, + "balance_loss_clip": 1.0947361, + "balance_loss_mlp": 1.04736471, + "epoch": 0.034870438285336376, + "flos": 27965865876480.0, + "grad_norm": 1.9625225208901669, + "language_loss": 0.80971527, + "learning_rate": 3.999757286178402e-06, + "loss": 0.83357298, + "num_input_tokens_seen": 6005850, + "step": 290, + "time_per_iteration": 2.7036664485931396 + }, + { + "auxiliary_loss_clip": 0.01319373, + "auxiliary_loss_mlp": 0.0105445, + "balance_loss_clip": 1.09029722, + "balance_loss_mlp": 1.03830957, + "epoch": 0.03499068117597547, + "flos": 22017514832640.0, + "grad_norm": 1.9307504032109595, + "language_loss": 0.9061594, + "learning_rate": 3.999744999052299e-06, + "loss": 0.92989767, + "num_input_tokens_seen": 6027240, + "step": 291, + "time_per_iteration": 2.636772394180298 + }, + { + "auxiliary_loss_clip": 0.01253732, + "auxiliary_loss_mlp": 0.01008762, + "balance_loss_clip": 1.10562325, + "balance_loss_mlp": 1.00339723, + "epoch": 0.03511092406661456, + "flos": 57242147725440.0, + "grad_norm": 0.9552839422220246, + "language_loss": 0.61154181, + "learning_rate": 3.9997324085664675e-06, + "loss": 0.63416672, + "num_input_tokens_seen": 6087470, + "step": 292, + "time_per_iteration": 3.159714937210083 + }, + { + "auxiliary_loss_clip": 0.01322663, + "auxiliary_loss_mlp": 0.01046228, + "balance_loss_clip": 1.09208989, + "balance_loss_mlp": 1.03260243, + "epoch": 0.03523116695725365, + "flos": 22928065626240.0, + "grad_norm": 2.2378609600620116, + "language_loss": 0.92041886, + "learning_rate": 3.999719514722821e-06, + "loss": 0.94410777, + "num_input_tokens_seen": 6107600, + "step": 293, + "time_per_iteration": 2.62870717048645 + }, + { + "auxiliary_loss_clip": 0.01316663, + "auxiliary_loss_mlp": 0.01049828, + "balance_loss_clip": 1.09123707, + "balance_loss_mlp": 1.0355227, + "epoch": 0.03535140984789274, + "flos": 36903241226880.0, + "grad_norm": 2.1420325021003794, + "language_loss": 0.74753416, + "learning_rate": 3.999706317523314e-06, + "loss": 0.77119899, + "num_input_tokens_seen": 6126160, + "step": 294, + "time_per_iteration": 2.748469114303589 + }, + { + "auxiliary_loss_clip": 0.01318117, + "auxiliary_loss_mlp": 0.01055426, + "balance_loss_clip": 1.09242582, + "balance_loss_mlp": 1.04066801, + "epoch": 0.03547165273853183, + "flos": 20449152316800.0, + "grad_norm": 2.353250915138766, + "language_loss": 0.86147094, + "learning_rate": 3.999692816969948e-06, + "loss": 0.88520634, + "num_input_tokens_seen": 6145695, + "step": 295, + "time_per_iteration": 2.625680923461914 + }, + { + "auxiliary_loss_clip": 0.01247087, + "auxiliary_loss_mlp": 0.01005397, + "balance_loss_clip": 1.10034156, + "balance_loss_mlp": 1.00017571, + "epoch": 0.03559189562917092, + "flos": 69850564871040.0, + "grad_norm": 1.000430273333583, + "language_loss": 0.69349778, + "learning_rate": 3.999679013064772e-06, + "loss": 0.71602267, + "num_input_tokens_seen": 6212440, + "step": 296, + "time_per_iteration": 3.2765519618988037 + }, + { + "auxiliary_loss_clip": 0.01316511, + "auxiliary_loss_mlp": 0.01055579, + "balance_loss_clip": 1.08961487, + "balance_loss_mlp": 1.04099989, + "epoch": 0.03571213851981002, + "flos": 21651944163840.0, + "grad_norm": 2.6889417404900944, + "language_loss": 0.85540426, + "learning_rate": 3.99966490580988e-06, + "loss": 0.87912512, + "num_input_tokens_seen": 6229800, + "step": 297, + "time_per_iteration": 2.606851100921631 + }, + { + "auxiliary_loss_clip": 0.01321268, + "auxiliary_loss_mlp": 0.0106139, + "balance_loss_clip": 1.0918802, + "balance_loss_mlp": 1.04631579, + "epoch": 0.03583238141044911, + "flos": 43945610757120.0, + "grad_norm": 2.69156414991708, + "language_loss": 0.65827692, + "learning_rate": 3.999650495207411e-06, + "loss": 0.68210346, + "num_input_tokens_seen": 6255825, + "step": 298, + "time_per_iteration": 2.8176090717315674 + }, + { + "auxiliary_loss_clip": 0.01317263, + "auxiliary_loss_mlp": 0.01057336, + "balance_loss_clip": 1.09203339, + "balance_loss_mlp": 1.04205942, + "epoch": 0.0359526243010882, + "flos": 18910810592640.0, + "grad_norm": 2.890446148932292, + "language_loss": 0.90478301, + "learning_rate": 3.999635781259553e-06, + "loss": 0.92852902, + "num_input_tokens_seen": 6271090, + "step": 299, + "time_per_iteration": 2.6257500648498535 + }, + { + "auxiliary_loss_clip": 0.01235342, + "auxiliary_loss_mlp": 0.01005329, + "balance_loss_clip": 1.09091163, + "balance_loss_mlp": 1.00022686, + "epoch": 0.03607286719172729, + "flos": 61668892782720.0, + "grad_norm": 0.9228511378902715, + "language_loss": 0.52293402, + "learning_rate": 3.999620763968535e-06, + "loss": 0.54534072, + "num_input_tokens_seen": 6329965, + "step": 300, + "time_per_iteration": 3.0535807609558105 + }, + { + "auxiliary_loss_clip": 0.01321671, + "auxiliary_loss_mlp": 0.01053391, + "balance_loss_clip": 1.09503627, + "balance_loss_mlp": 1.03925908, + "epoch": 0.03619311008236638, + "flos": 27819062991360.0, + "grad_norm": 1.7626203508716394, + "language_loss": 0.86561698, + "learning_rate": 3.999605443336638e-06, + "loss": 0.88936758, + "num_input_tokens_seen": 6352095, + "step": 301, + "time_per_iteration": 2.6757659912109375 + }, + { + "auxiliary_loss_clip": 0.01315928, + "auxiliary_loss_mlp": 0.01060494, + "balance_loss_clip": 1.08654475, + "balance_loss_mlp": 1.04468703, + "epoch": 0.03631335297300547, + "flos": 13621133197440.0, + "grad_norm": 2.457563448606435, + "language_loss": 0.89499485, + "learning_rate": 3.999589819366185e-06, + "loss": 0.91875911, + "num_input_tokens_seen": 6365885, + "step": 302, + "time_per_iteration": 3.487703561782837 + }, + { + "auxiliary_loss_clip": 0.013165, + "auxiliary_loss_mlp": 0.01046253, + "balance_loss_clip": 1.08863652, + "balance_loss_mlp": 1.03192377, + "epoch": 0.036433595863644565, + "flos": 27631788456960.0, + "grad_norm": 2.3324321576073284, + "language_loss": 0.84978521, + "learning_rate": 3.999573892059547e-06, + "loss": 0.87341279, + "num_input_tokens_seen": 6385015, + "step": 303, + "time_per_iteration": 3.6203930377960205 + }, + { + "auxiliary_loss_clip": 0.01320956, + "auxiliary_loss_mlp": 0.01066891, + "balance_loss_clip": 1.09187448, + "balance_loss_mlp": 1.0516324, + "epoch": 0.036553838754283655, + "flos": 24572020314240.0, + "grad_norm": 2.013143178351301, + "language_loss": 0.80999774, + "learning_rate": 3.999557661419138e-06, + "loss": 0.83387625, + "num_input_tokens_seen": 6405165, + "step": 304, + "time_per_iteration": 4.461326837539673 + }, + { + "auxiliary_loss_clip": 0.01323587, + "auxiliary_loss_mlp": 0.01056355, + "balance_loss_clip": 1.09365129, + "balance_loss_mlp": 1.04237735, + "epoch": 0.036674081644922744, + "flos": 23404313076480.0, + "grad_norm": 1.8905628149585525, + "language_loss": 0.81370723, + "learning_rate": 3.9995411274474225e-06, + "loss": 0.83750671, + "num_input_tokens_seen": 6424445, + "step": 305, + "time_per_iteration": 2.6403045654296875 + }, + { + "auxiliary_loss_clip": 0.01324838, + "auxiliary_loss_mlp": 0.01056632, + "balance_loss_clip": 1.09563875, + "balance_loss_mlp": 1.04217815, + "epoch": 0.036794324535561834, + "flos": 27489690253440.0, + "grad_norm": 2.073334096880232, + "language_loss": 0.81599182, + "learning_rate": 3.999524290146908e-06, + "loss": 0.83980644, + "num_input_tokens_seen": 6444650, + "step": 306, + "time_per_iteration": 2.6481003761291504 + }, + { + "auxiliary_loss_clip": 0.01319562, + "auxiliary_loss_mlp": 0.01052027, + "balance_loss_clip": 1.0918541, + "balance_loss_mlp": 1.0375669, + "epoch": 0.036914567426200924, + "flos": 19463476227840.0, + "grad_norm": 2.8750279931375977, + "language_loss": 0.92976999, + "learning_rate": 3.9995071495201485e-06, + "loss": 0.95348585, + "num_input_tokens_seen": 6461755, + "step": 307, + "time_per_iteration": 2.8121542930603027 + }, + { + "auxiliary_loss_clip": 0.01318725, + "auxiliary_loss_mlp": 0.01054091, + "balance_loss_clip": 1.09129012, + "balance_loss_mlp": 1.04007781, + "epoch": 0.037034810316840014, + "flos": 22309324922880.0, + "grad_norm": 26.41444751656569, + "language_loss": 0.97708136, + "learning_rate": 3.999489705569744e-06, + "loss": 1.00080943, + "num_input_tokens_seen": 6479455, + "step": 308, + "time_per_iteration": 2.5880260467529297 + }, + { + "auxiliary_loss_clip": 0.01319021, + "auxiliary_loss_mlp": 0.01055573, + "balance_loss_clip": 1.09127355, + "balance_loss_mlp": 1.04141676, + "epoch": 0.03715505320747911, + "flos": 18588333265920.0, + "grad_norm": 2.215013477577618, + "language_loss": 0.86484087, + "learning_rate": 3.999471958298341e-06, + "loss": 0.88858682, + "num_input_tokens_seen": 6498365, + "step": 309, + "time_per_iteration": 2.593712568283081 + }, + { + "auxiliary_loss_clip": 0.01324717, + "auxiliary_loss_mlp": 0.01056285, + "balance_loss_clip": 1.09564435, + "balance_loss_mlp": 1.04120517, + "epoch": 0.0372752960981182, + "flos": 35955343267200.0, + "grad_norm": 1.9621088024366926, + "language_loss": 0.76244557, + "learning_rate": 3.999453907708631e-06, + "loss": 0.7862556, + "num_input_tokens_seen": 6520770, + "step": 310, + "time_per_iteration": 2.7113888263702393 + }, + { + "auxiliary_loss_clip": 0.01315025, + "auxiliary_loss_mlp": 0.01060915, + "balance_loss_clip": 1.09079385, + "balance_loss_mlp": 1.04756927, + "epoch": 0.03739553898875729, + "flos": 20814040627200.0, + "grad_norm": 1.8680262397763308, + "language_loss": 0.81465566, + "learning_rate": 3.999435553803353e-06, + "loss": 0.83841509, + "num_input_tokens_seen": 6540170, + "step": 311, + "time_per_iteration": 2.611269235610962 + }, + { + "auxiliary_loss_clip": 0.01314084, + "auxiliary_loss_mlp": 0.01061623, + "balance_loss_clip": 1.08779478, + "balance_loss_mlp": 1.0465548, + "epoch": 0.03751578187939638, + "flos": 20264140339200.0, + "grad_norm": 2.943131768626363, + "language_loss": 0.83418417, + "learning_rate": 3.999416896585292e-06, + "loss": 0.85794121, + "num_input_tokens_seen": 6557200, + "step": 312, + "time_per_iteration": 2.6067564487457275 + }, + { + "auxiliary_loss_clip": 0.0131759, + "auxiliary_loss_mlp": 0.01056039, + "balance_loss_clip": 1.09182167, + "balance_loss_mlp": 1.04238391, + "epoch": 0.03763602477003547, + "flos": 20668063754880.0, + "grad_norm": 3.709514220955351, + "language_loss": 0.85834336, + "learning_rate": 3.9993979360572775e-06, + "loss": 0.8820796, + "num_input_tokens_seen": 6577340, + "step": 313, + "time_per_iteration": 2.6433498859405518 + }, + { + "auxiliary_loss_clip": 0.01322585, + "auxiliary_loss_mlp": 0.01058863, + "balance_loss_clip": 1.09326661, + "balance_loss_mlp": 1.04297245, + "epoch": 0.03775626766067456, + "flos": 16691352197760.0, + "grad_norm": 2.7189865648530875, + "language_loss": 0.82835263, + "learning_rate": 3.999378672222185e-06, + "loss": 0.85216713, + "num_input_tokens_seen": 6595125, + "step": 314, + "time_per_iteration": 2.5843801498413086 + }, + { + "auxiliary_loss_clip": 0.01310222, + "auxiliary_loss_mlp": 0.0105923, + "balance_loss_clip": 1.08859015, + "balance_loss_mlp": 1.04535389, + "epoch": 0.03787651055131366, + "flos": 21141797253120.0, + "grad_norm": 1.9014341607425074, + "language_loss": 0.82697517, + "learning_rate": 3.9993591050829385e-06, + "loss": 0.85066968, + "num_input_tokens_seen": 6612990, + "step": 315, + "time_per_iteration": 2.6221089363098145 + }, + { + "auxiliary_loss_clip": 0.01318443, + "auxiliary_loss_mlp": 0.01050505, + "balance_loss_clip": 1.09360957, + "balance_loss_mlp": 1.03616428, + "epoch": 0.037996753441952746, + "flos": 22018089450240.0, + "grad_norm": 1.93256041473364, + "language_loss": 0.79469073, + "learning_rate": 3.999339234642506e-06, + "loss": 0.81838024, + "num_input_tokens_seen": 6632740, + "step": 316, + "time_per_iteration": 2.6091904640197754 + }, + { + "auxiliary_loss_clip": 0.0131664, + "auxiliary_loss_mlp": 0.01053302, + "balance_loss_clip": 1.09153688, + "balance_loss_mlp": 1.03834128, + "epoch": 0.038116996332591836, + "flos": 27709391790720.0, + "grad_norm": 3.060179142511417, + "language_loss": 0.83780169, + "learning_rate": 3.9993190609038994e-06, + "loss": 0.8615011, + "num_input_tokens_seen": 6651505, + "step": 317, + "time_per_iteration": 2.6601297855377197 + }, + { + "auxiliary_loss_clip": 0.01319361, + "auxiliary_loss_mlp": 0.01062773, + "balance_loss_clip": 1.09254265, + "balance_loss_mlp": 1.04840839, + "epoch": 0.038237239223230926, + "flos": 21178067011200.0, + "grad_norm": 2.090386900369453, + "language_loss": 0.82924205, + "learning_rate": 3.999298583870182e-06, + "loss": 0.8530634, + "num_input_tokens_seen": 6671090, + "step": 318, + "time_per_iteration": 2.6636996269226074 + }, + { + "auxiliary_loss_clip": 0.01314205, + "auxiliary_loss_mlp": 0.01054702, + "balance_loss_clip": 1.08953464, + "balance_loss_mlp": 1.04106414, + "epoch": 0.038357482113870016, + "flos": 25556618995200.0, + "grad_norm": 1.8484473538681887, + "language_loss": 0.77783185, + "learning_rate": 3.999277803544458e-06, + "loss": 0.80152094, + "num_input_tokens_seen": 6691245, + "step": 319, + "time_per_iteration": 2.686126470565796 + }, + { + "auxiliary_loss_clip": 0.0120876, + "auxiliary_loss_mlp": 0.01010572, + "balance_loss_clip": 1.07097876, + "balance_loss_mlp": 1.00573242, + "epoch": 0.038477725004509106, + "flos": 59227578034560.0, + "grad_norm": 0.9550138452786125, + "language_loss": 0.62428832, + "learning_rate": 3.999256719929882e-06, + "loss": 0.64648163, + "num_input_tokens_seen": 6752520, + "step": 320, + "time_per_iteration": 3.1476049423217773 + }, + { + "auxiliary_loss_clip": 0.01207313, + "auxiliary_loss_mlp": 0.01006673, + "balance_loss_clip": 1.06990457, + "balance_loss_mlp": 1.00185728, + "epoch": 0.0385979678951482, + "flos": 67317676398720.0, + "grad_norm": 1.2030289880227205, + "language_loss": 0.67108512, + "learning_rate": 3.999235333029651e-06, + "loss": 0.69322503, + "num_input_tokens_seen": 6806460, + "step": 321, + "time_per_iteration": 3.0663397312164307 + }, + { + "auxiliary_loss_clip": 0.01319369, + "auxiliary_loss_mlp": 0.01056937, + "balance_loss_clip": 1.09542871, + "balance_loss_mlp": 1.04200017, + "epoch": 0.03871821078578729, + "flos": 22746752749440.0, + "grad_norm": 2.1273180013967834, + "language_loss": 0.82035661, + "learning_rate": 3.999213642847009e-06, + "loss": 0.84411967, + "num_input_tokens_seen": 6827045, + "step": 322, + "time_per_iteration": 2.669813871383667 + }, + { + "auxiliary_loss_clip": 0.01310026, + "auxiliary_loss_mlp": 0.01059104, + "balance_loss_clip": 1.08778286, + "balance_loss_mlp": 1.04468012, + "epoch": 0.03883845367642638, + "flos": 26280613526400.0, + "grad_norm": 1.7598425966338869, + "language_loss": 0.91308618, + "learning_rate": 3.999191649385247e-06, + "loss": 0.93677747, + "num_input_tokens_seen": 6848220, + "step": 323, + "time_per_iteration": 2.662250518798828 + }, + { + "auxiliary_loss_clip": 0.01202396, + "auxiliary_loss_mlp": 0.01003276, + "balance_loss_clip": 1.06624007, + "balance_loss_mlp": 0.99865019, + "epoch": 0.03895869656706547, + "flos": 56962835568000.0, + "grad_norm": 0.9044010503991111, + "language_loss": 0.5982824, + "learning_rate": 3.999169352647702e-06, + "loss": 0.62033916, + "num_input_tokens_seen": 6909400, + "step": 324, + "time_per_iteration": 3.1315605640411377 + }, + { + "auxiliary_loss_clip": 0.01318533, + "auxiliary_loss_mlp": 0.01065432, + "balance_loss_clip": 1.09132993, + "balance_loss_mlp": 1.05035794, + "epoch": 0.03907893945770456, + "flos": 24863363527680.0, + "grad_norm": 2.357880815631523, + "language_loss": 0.83029407, + "learning_rate": 3.999146752637755e-06, + "loss": 0.85413373, + "num_input_tokens_seen": 6930445, + "step": 325, + "time_per_iteration": 2.650099515914917 + }, + { + "auxiliary_loss_clip": 0.01309269, + "auxiliary_loss_mlp": 0.01058158, + "balance_loss_clip": 1.08716941, + "balance_loss_mlp": 1.0434947, + "epoch": 0.03919918234834365, + "flos": 18368595815040.0, + "grad_norm": 2.837603890668335, + "language_loss": 0.89649308, + "learning_rate": 3.999123849358836e-06, + "loss": 0.92016739, + "num_input_tokens_seen": 6948110, + "step": 326, + "time_per_iteration": 2.6258556842803955 + }, + { + "auxiliary_loss_clip": 0.01308693, + "auxiliary_loss_mlp": 0.01048576, + "balance_loss_clip": 1.08620095, + "balance_loss_mlp": 1.03486681, + "epoch": 0.03931942523898275, + "flos": 25225414663680.0, + "grad_norm": 2.3279215652349445, + "language_loss": 0.74711829, + "learning_rate": 3.999100642814418e-06, + "loss": 0.77069104, + "num_input_tokens_seen": 6968550, + "step": 327, + "time_per_iteration": 2.7076363563537598 + }, + { + "auxiliary_loss_clip": 0.01316621, + "auxiliary_loss_mlp": 0.01063706, + "balance_loss_clip": 1.09325886, + "balance_loss_mlp": 1.04933524, + "epoch": 0.03943966812962184, + "flos": 23257905240960.0, + "grad_norm": 2.183568470199827, + "language_loss": 0.88149959, + "learning_rate": 3.999077133008022e-06, + "loss": 0.90530288, + "num_input_tokens_seen": 6987135, + "step": 328, + "time_per_iteration": 2.6436216831207275 + }, + { + "auxiliary_loss_clip": 0.01315415, + "auxiliary_loss_mlp": 0.01055324, + "balance_loss_clip": 1.09243834, + "balance_loss_mlp": 1.04132295, + "epoch": 0.03955991102026093, + "flos": 29168837291520.0, + "grad_norm": 2.1145441282087045, + "language_loss": 0.90474641, + "learning_rate": 3.9990533199432145e-06, + "loss": 0.9284538, + "num_input_tokens_seen": 7008630, + "step": 329, + "time_per_iteration": 4.569812059402466 + }, + { + "auxiliary_loss_clip": 0.01314199, + "auxiliary_loss_mlp": 0.01051919, + "balance_loss_clip": 1.09009469, + "balance_loss_mlp": 1.03752446, + "epoch": 0.03968015391090002, + "flos": 17602441695360.0, + "grad_norm": 2.477621102563168, + "language_loss": 0.75760794, + "learning_rate": 3.999029203623608e-06, + "loss": 0.78126907, + "num_input_tokens_seen": 7026350, + "step": 330, + "time_per_iteration": 3.5798158645629883 + }, + { + "auxiliary_loss_clip": 0.0131146, + "auxiliary_loss_mlp": 0.01050654, + "balance_loss_clip": 1.09039438, + "balance_loss_mlp": 1.03675389, + "epoch": 0.03980039680153911, + "flos": 21799285752960.0, + "grad_norm": 1.9796100175006286, + "language_loss": 0.86890543, + "learning_rate": 3.99900478405286e-06, + "loss": 0.89252657, + "num_input_tokens_seen": 7045660, + "step": 331, + "time_per_iteration": 3.587165594100952 + }, + { + "auxiliary_loss_clip": 0.0131707, + "auxiliary_loss_mlp": 0.01055732, + "balance_loss_clip": 1.09573567, + "balance_loss_mlp": 1.04143262, + "epoch": 0.0399206396921782, + "flos": 15195134148480.0, + "grad_norm": 2.9580602096524813, + "language_loss": 0.82770026, + "learning_rate": 3.998980061234676e-06, + "loss": 0.85142827, + "num_input_tokens_seen": 7063575, + "step": 332, + "time_per_iteration": 2.616187572479248 + }, + { + "auxiliary_loss_clip": 0.01313096, + "auxiliary_loss_mlp": 0.01063387, + "balance_loss_clip": 1.0900898, + "balance_loss_mlp": 1.04885554, + "epoch": 0.040040882582817294, + "flos": 14422910630400.0, + "grad_norm": 2.6250396573838906, + "language_loss": 0.75703758, + "learning_rate": 3.9989550351728055e-06, + "loss": 0.78080237, + "num_input_tokens_seen": 7080505, + "step": 333, + "time_per_iteration": 2.5663530826568604 + }, + { + "auxiliary_loss_clip": 0.01311341, + "auxiliary_loss_mlp": 0.01052819, + "balance_loss_clip": 1.09086919, + "balance_loss_mlp": 1.03905594, + "epoch": 0.040161125473456384, + "flos": 19280906375040.0, + "grad_norm": 2.377694539633658, + "language_loss": 0.8467257, + "learning_rate": 3.998929705871046e-06, + "loss": 0.87036729, + "num_input_tokens_seen": 7097860, + "step": 334, + "time_per_iteration": 2.594193696975708 + }, + { + "auxiliary_loss_clip": 0.01309325, + "auxiliary_loss_mlp": 0.01057013, + "balance_loss_clip": 1.0896827, + "balance_loss_mlp": 1.04326844, + "epoch": 0.040281368364095474, + "flos": 17821101738240.0, + "grad_norm": 2.845736804498238, + "language_loss": 0.8909018, + "learning_rate": 3.99890407333324e-06, + "loss": 0.91456521, + "num_input_tokens_seen": 7116390, + "step": 335, + "time_per_iteration": 2.5981552600860596 + }, + { + "auxiliary_loss_clip": 0.01307567, + "auxiliary_loss_mlp": 0.01055131, + "balance_loss_clip": 1.08633053, + "balance_loss_mlp": 1.04071844, + "epoch": 0.040401611254734564, + "flos": 19573757959680.0, + "grad_norm": 2.3811017563707995, + "language_loss": 0.87204266, + "learning_rate": 3.998878137563275e-06, + "loss": 0.89566958, + "num_input_tokens_seen": 7135940, + "step": 336, + "time_per_iteration": 2.6289615631103516 + }, + { + "auxiliary_loss_clip": 0.01305633, + "auxiliary_loss_mlp": 0.01046596, + "balance_loss_clip": 1.08697772, + "balance_loss_mlp": 1.03151584, + "epoch": 0.040521854145373654, + "flos": 22054466949120.0, + "grad_norm": 2.2324109039013895, + "language_loss": 0.84983945, + "learning_rate": 3.998851898565085e-06, + "loss": 0.87336177, + "num_input_tokens_seen": 7155745, + "step": 337, + "time_per_iteration": 2.620145320892334 + }, + { + "auxiliary_loss_clip": 0.01308998, + "auxiliary_loss_mlp": 0.01049342, + "balance_loss_clip": 1.08789515, + "balance_loss_mlp": 1.03526974, + "epoch": 0.04064209703601274, + "flos": 22674644196480.0, + "grad_norm": 2.214350414984848, + "language_loss": 0.82811528, + "learning_rate": 3.998825356342653e-06, + "loss": 0.8516987, + "num_input_tokens_seen": 7175920, + "step": 338, + "time_per_iteration": 2.640328884124756 + }, + { + "auxiliary_loss_clip": 0.01309366, + "auxiliary_loss_mlp": 0.01049505, + "balance_loss_clip": 1.08721924, + "balance_loss_mlp": 1.0351162, + "epoch": 0.04076233992665183, + "flos": 38582172783360.0, + "grad_norm": 3.69406433642311, + "language_loss": 0.73286611, + "learning_rate": 3.998798510900003e-06, + "loss": 0.75645483, + "num_input_tokens_seen": 7198720, + "step": 339, + "time_per_iteration": 2.754089593887329 + }, + { + "auxiliary_loss_clip": 0.0130365, + "auxiliary_loss_mlp": 0.01048285, + "balance_loss_clip": 1.08238518, + "balance_loss_mlp": 1.034361, + "epoch": 0.04088258281729093, + "flos": 25885309374720.0, + "grad_norm": 2.531568276724195, + "language_loss": 0.83789796, + "learning_rate": 3.998771362241207e-06, + "loss": 0.86141729, + "num_input_tokens_seen": 7219125, + "step": 340, + "time_per_iteration": 2.6324732303619385 + }, + { + "auxiliary_loss_clip": 0.01307926, + "auxiliary_loss_mlp": 0.01043521, + "balance_loss_clip": 1.08821273, + "balance_loss_mlp": 1.03027713, + "epoch": 0.04100282570793002, + "flos": 19789832223360.0, + "grad_norm": 1.8171591746293552, + "language_loss": 0.88082856, + "learning_rate": 3.998743910370385e-06, + "loss": 0.90434301, + "num_input_tokens_seen": 7237985, + "step": 341, + "time_per_iteration": 2.6540615558624268 + }, + { + "auxiliary_loss_clip": 0.01314809, + "auxiliary_loss_mlp": 0.01054133, + "balance_loss_clip": 1.09515572, + "balance_loss_mlp": 1.04081726, + "epoch": 0.04112306859856911, + "flos": 22565152563840.0, + "grad_norm": 2.3649564611972247, + "language_loss": 0.73410296, + "learning_rate": 3.998716155291702e-06, + "loss": 0.75779241, + "num_input_tokens_seen": 7255825, + "step": 342, + "time_per_iteration": 2.600909471511841 + }, + { + "auxiliary_loss_clip": 0.01311224, + "auxiliary_loss_mlp": 0.01051765, + "balance_loss_clip": 1.09220028, + "balance_loss_mlp": 1.03876543, + "epoch": 0.0412433114892082, + "flos": 25040654081280.0, + "grad_norm": 2.5878285005949486, + "language_loss": 0.90464705, + "learning_rate": 3.998688097009366e-06, + "loss": 0.92827702, + "num_input_tokens_seen": 7276590, + "step": 343, + "time_per_iteration": 2.6670353412628174 + }, + { + "auxiliary_loss_clip": 0.01306987, + "auxiliary_loss_mlp": 0.01048005, + "balance_loss_clip": 1.08880818, + "balance_loss_mlp": 1.03477263, + "epoch": 0.04136355437984729, + "flos": 25191371548800.0, + "grad_norm": 3.039270178495479, + "language_loss": 0.80082053, + "learning_rate": 3.998659735527636e-06, + "loss": 0.82437044, + "num_input_tokens_seen": 7295680, + "step": 344, + "time_per_iteration": 2.6320793628692627 + }, + { + "auxiliary_loss_clip": 0.0130807, + "auxiliary_loss_mlp": 0.01051415, + "balance_loss_clip": 1.08832955, + "balance_loss_mlp": 1.03769398, + "epoch": 0.04148379727048638, + "flos": 22966777509120.0, + "grad_norm": 2.1211479314687285, + "language_loss": 0.77879167, + "learning_rate": 3.998631070850813e-06, + "loss": 0.80238652, + "num_input_tokens_seen": 7316300, + "step": 345, + "time_per_iteration": 2.7204134464263916 + }, + { + "auxiliary_loss_clip": 0.01308283, + "auxiliary_loss_mlp": 0.01058604, + "balance_loss_clip": 1.09113646, + "balance_loss_mlp": 1.04537201, + "epoch": 0.041604040161125476, + "flos": 14063481187200.0, + "grad_norm": 2.6473637431708683, + "language_loss": 0.83384061, + "learning_rate": 3.9986021029832455e-06, + "loss": 0.85750949, + "num_input_tokens_seen": 7333615, + "step": 346, + "time_per_iteration": 2.6298117637634277 + }, + { + "auxiliary_loss_clip": 0.01306337, + "auxiliary_loss_mlp": 0.01058518, + "balance_loss_clip": 1.0859462, + "balance_loss_mlp": 1.04386663, + "epoch": 0.041724283051764566, + "flos": 12091877614080.0, + "grad_norm": 5.068928889622225, + "language_loss": 0.91789019, + "learning_rate": 3.9985728319293285e-06, + "loss": 0.94153869, + "num_input_tokens_seen": 7347590, + "step": 347, + "time_per_iteration": 2.622990369796753 + }, + { + "auxiliary_loss_clip": 0.01309711, + "auxiliary_loss_mlp": 0.01053537, + "balance_loss_clip": 1.0872283, + "balance_loss_mlp": 1.03855252, + "epoch": 0.041844525942403656, + "flos": 12385303816320.0, + "grad_norm": 2.543618564914954, + "language_loss": 0.85718197, + "learning_rate": 3.998543257693501e-06, + "loss": 0.88081443, + "num_input_tokens_seen": 7364345, + "step": 348, + "time_per_iteration": 2.6099817752838135 + }, + { + "auxiliary_loss_clip": 0.01313406, + "auxiliary_loss_mlp": 0.010536, + "balance_loss_clip": 1.09271121, + "balance_loss_mlp": 1.04099321, + "epoch": 0.041964768833042745, + "flos": 23769345041280.0, + "grad_norm": 1.9353825484519982, + "language_loss": 0.87794471, + "learning_rate": 3.998513380280251e-06, + "loss": 0.90161479, + "num_input_tokens_seen": 7384625, + "step": 349, + "time_per_iteration": 2.669072151184082 + }, + { + "auxiliary_loss_clip": 0.01312311, + "auxiliary_loss_mlp": 0.01053917, + "balance_loss_clip": 1.09017849, + "balance_loss_mlp": 1.04020238, + "epoch": 0.042085011723681835, + "flos": 11875336473600.0, + "grad_norm": 2.1915583198588107, + "language_loss": 0.94667542, + "learning_rate": 3.99848319969411e-06, + "loss": 0.97033775, + "num_input_tokens_seen": 7402225, + "step": 350, + "time_per_iteration": 2.5798609256744385 + }, + { + "auxiliary_loss_clip": 0.01316216, + "auxiliary_loss_mlp": 0.0105596, + "balance_loss_clip": 1.0943079, + "balance_loss_mlp": 1.04161906, + "epoch": 0.042205254614320925, + "flos": 16873957964160.0, + "grad_norm": 2.1895912133191104, + "language_loss": 0.79203308, + "learning_rate": 3.9984527159396564e-06, + "loss": 0.81575483, + "num_input_tokens_seen": 7420865, + "step": 351, + "time_per_iteration": 2.6342029571533203 + }, + { + "auxiliary_loss_clip": 0.01300953, + "auxiliary_loss_mlp": 0.01046956, + "balance_loss_clip": 1.08174562, + "balance_loss_mlp": 1.03200698, + "epoch": 0.04232549750496002, + "flos": 25118508810240.0, + "grad_norm": 2.568861046791861, + "language_loss": 0.84188473, + "learning_rate": 3.9984219290215154e-06, + "loss": 0.86536384, + "num_input_tokens_seen": 7441040, + "step": 352, + "time_per_iteration": 2.693084239959717 + }, + { + "auxiliary_loss_clip": 0.01309131, + "auxiliary_loss_mlp": 0.01051674, + "balance_loss_clip": 1.08850837, + "balance_loss_mlp": 1.037184, + "epoch": 0.04244574039559911, + "flos": 26724541714560.0, + "grad_norm": 1.6188780996509067, + "language_loss": 0.89199901, + "learning_rate": 3.998390838944356e-06, + "loss": 0.91560704, + "num_input_tokens_seen": 7462545, + "step": 353, + "time_per_iteration": 2.792365312576294 + }, + { + "auxiliary_loss_clip": 0.01308982, + "auxiliary_loss_mlp": 0.01056297, + "balance_loss_clip": 1.08837152, + "balance_loss_mlp": 1.04255223, + "epoch": 0.0425659832862382, + "flos": 20923244951040.0, + "grad_norm": 3.2382416460170567, + "language_loss": 0.90409052, + "learning_rate": 3.998359445712895e-06, + "loss": 0.92774332, + "num_input_tokens_seen": 7481650, + "step": 354, + "time_per_iteration": 2.7807109355926514 + }, + { + "auxiliary_loss_clip": 0.01306724, + "auxiliary_loss_mlp": 0.01051058, + "balance_loss_clip": 1.08637822, + "balance_loss_mlp": 1.03688347, + "epoch": 0.04268622617687729, + "flos": 23331127115520.0, + "grad_norm": 2.614984203504092, + "language_loss": 0.81009555, + "learning_rate": 3.9983277493318955e-06, + "loss": 0.83367336, + "num_input_tokens_seen": 7500945, + "step": 355, + "time_per_iteration": 4.474182844161987 + }, + { + "auxiliary_loss_clip": 0.01306123, + "auxiliary_loss_mlp": 0.01050972, + "balance_loss_clip": 1.08537221, + "balance_loss_mlp": 1.03638053, + "epoch": 0.04280646906751638, + "flos": 25994010908160.0, + "grad_norm": 1.6026662469556396, + "language_loss": 0.81184798, + "learning_rate": 3.998295749806165e-06, + "loss": 0.83541894, + "num_input_tokens_seen": 7522170, + "step": 356, + "time_per_iteration": 3.5499775409698486 + }, + { + "auxiliary_loss_clip": 0.01305359, + "auxiliary_loss_mlp": 0.01056713, + "balance_loss_clip": 1.08752346, + "balance_loss_mlp": 1.04310501, + "epoch": 0.04292671195815547, + "flos": 26906824258560.0, + "grad_norm": 2.363367110446849, + "language_loss": 0.8309198, + "learning_rate": 3.998263447140558e-06, + "loss": 0.85454053, + "num_input_tokens_seen": 7542370, + "step": 357, + "time_per_iteration": 3.534044027328491 + }, + { + "auxiliary_loss_clip": 0.01304056, + "auxiliary_loss_mlp": 0.01053822, + "balance_loss_clip": 1.08481097, + "balance_loss_mlp": 1.04032719, + "epoch": 0.04304695484879457, + "flos": 39457315745280.0, + "grad_norm": 1.9398768946309888, + "language_loss": 0.8171953, + "learning_rate": 3.998230841339976e-06, + "loss": 0.84077406, + "num_input_tokens_seen": 7564380, + "step": 358, + "time_per_iteration": 2.752437114715576 + }, + { + "auxiliary_loss_clip": 0.01304387, + "auxiliary_loss_mlp": 0.0106937, + "balance_loss_clip": 1.08835304, + "balance_loss_mlp": 1.05569029, + "epoch": 0.04316719773943366, + "flos": 19646297475840.0, + "grad_norm": 2.1792611961934627, + "language_loss": 0.85042834, + "learning_rate": 3.998197932409363e-06, + "loss": 0.87416589, + "num_input_tokens_seen": 7582390, + "step": 359, + "time_per_iteration": 2.602876901626587 + }, + { + "auxiliary_loss_clip": 0.01301058, + "auxiliary_loss_mlp": 0.01051727, + "balance_loss_clip": 1.08521342, + "balance_loss_mlp": 1.03859019, + "epoch": 0.04328744063007275, + "flos": 22452320966400.0, + "grad_norm": 3.185368029813027, + "language_loss": 0.85971242, + "learning_rate": 3.9981647203537125e-06, + "loss": 0.88324028, + "num_input_tokens_seen": 7599890, + "step": 360, + "time_per_iteration": 2.583787679672241 + }, + { + "auxiliary_loss_clip": 0.01305604, + "auxiliary_loss_mlp": 0.01055699, + "balance_loss_clip": 1.08653164, + "balance_loss_mlp": 1.04246044, + "epoch": 0.04340768352071184, + "flos": 21283033530240.0, + "grad_norm": 1.9792481402973836, + "language_loss": 0.95948601, + "learning_rate": 3.998131205178063e-06, + "loss": 0.9830991, + "num_input_tokens_seen": 7618360, + "step": 361, + "time_per_iteration": 2.6178455352783203 + }, + { + "auxiliary_loss_clip": 0.01303372, + "auxiliary_loss_mlp": 0.01050853, + "balance_loss_clip": 1.08522749, + "balance_loss_mlp": 1.03539193, + "epoch": 0.04352792641135093, + "flos": 11583705951360.0, + "grad_norm": 3.2614151861735965, + "language_loss": 0.76711673, + "learning_rate": 3.998097386887498e-06, + "loss": 0.79065895, + "num_input_tokens_seen": 7635435, + "step": 362, + "time_per_iteration": 2.574432849884033 + }, + { + "auxiliary_loss_clip": 0.01303618, + "auxiliary_loss_mlp": 0.01047155, + "balance_loss_clip": 1.08762252, + "balance_loss_mlp": 1.03416157, + "epoch": 0.04364816930199002, + "flos": 23623547736960.0, + "grad_norm": 1.7510516813012547, + "language_loss": 0.85092843, + "learning_rate": 3.998063265487148e-06, + "loss": 0.8744362, + "num_input_tokens_seen": 7656485, + "step": 363, + "time_per_iteration": 2.661924123764038 + }, + { + "auxiliary_loss_clip": 0.0130237, + "auxiliary_loss_mlp": 0.01048702, + "balance_loss_clip": 1.08658648, + "balance_loss_mlp": 1.03555369, + "epoch": 0.043768412192629114, + "flos": 14429734214400.0, + "grad_norm": 1.8879752967229084, + "language_loss": 0.80998111, + "learning_rate": 3.99802884098219e-06, + "loss": 0.83349192, + "num_input_tokens_seen": 7674595, + "step": 364, + "time_per_iteration": 2.6175994873046875 + }, + { + "auxiliary_loss_clip": 0.01299612, + "auxiliary_loss_mlp": 0.010425, + "balance_loss_clip": 1.08116603, + "balance_loss_mlp": 1.02817726, + "epoch": 0.043888655083268203, + "flos": 26468893641600.0, + "grad_norm": 2.6969064074211055, + "language_loss": 0.82289845, + "learning_rate": 3.997994113377845e-06, + "loss": 0.84631956, + "num_input_tokens_seen": 7693495, + "step": 365, + "time_per_iteration": 2.621397018432617 + }, + { + "auxiliary_loss_clip": 0.01303556, + "auxiliary_loss_mlp": 0.0104855, + "balance_loss_clip": 1.08668065, + "balance_loss_mlp": 1.03499579, + "epoch": 0.04400889797390729, + "flos": 27235263242880.0, + "grad_norm": 2.3778560846013455, + "language_loss": 0.83328032, + "learning_rate": 3.9979590826793815e-06, + "loss": 0.85680139, + "num_input_tokens_seen": 7714685, + "step": 366, + "time_per_iteration": 2.6679019927978516 + }, + { + "auxiliary_loss_clip": 0.01303551, + "auxiliary_loss_mlp": 0.01057058, + "balance_loss_clip": 1.08564651, + "balance_loss_mlp": 1.04266918, + "epoch": 0.04412914086454638, + "flos": 20119528183680.0, + "grad_norm": 2.0232462251891365, + "language_loss": 0.80927324, + "learning_rate": 3.997923748892113e-06, + "loss": 0.83287936, + "num_input_tokens_seen": 7734005, + "step": 367, + "time_per_iteration": 2.586362600326538 + }, + { + "auxiliary_loss_clip": 0.01304617, + "auxiliary_loss_mlp": 0.01059457, + "balance_loss_clip": 1.08766675, + "balance_loss_mlp": 1.04597497, + "epoch": 0.04424938375518547, + "flos": 22604618632320.0, + "grad_norm": 1.7762986930242897, + "language_loss": 0.88537717, + "learning_rate": 3.9978881120214015e-06, + "loss": 0.90901792, + "num_input_tokens_seen": 7755525, + "step": 368, + "time_per_iteration": 2.6624162197113037 + }, + { + "auxiliary_loss_clip": 0.01298059, + "auxiliary_loss_mlp": 0.01059918, + "balance_loss_clip": 1.08087933, + "balance_loss_mlp": 1.04650712, + "epoch": 0.04436962664582456, + "flos": 24132365844480.0, + "grad_norm": 2.25329599163098, + "language_loss": 0.79330897, + "learning_rate": 3.997852172072652e-06, + "loss": 0.81688869, + "num_input_tokens_seen": 7776740, + "step": 369, + "time_per_iteration": 2.612854242324829 + }, + { + "auxiliary_loss_clip": 0.013002, + "auxiliary_loss_mlp": 0.01046262, + "balance_loss_clip": 1.08435297, + "balance_loss_mlp": 1.03208804, + "epoch": 0.04448986953646366, + "flos": 18222906251520.0, + "grad_norm": 2.7196320014946402, + "language_loss": 0.8922081, + "learning_rate": 3.9978159290513155e-06, + "loss": 0.91567272, + "num_input_tokens_seen": 7794820, + "step": 370, + "time_per_iteration": 2.5863258838653564 + }, + { + "auxiliary_loss_clip": 0.01301708, + "auxiliary_loss_mlp": 0.01046621, + "balance_loss_clip": 1.08448958, + "balance_loss_mlp": 1.03353798, + "epoch": 0.04461011242710275, + "flos": 30117920400000.0, + "grad_norm": 1.6680496226121366, + "language_loss": 0.80228978, + "learning_rate": 3.997779382962892e-06, + "loss": 0.82577312, + "num_input_tokens_seen": 7817705, + "step": 371, + "time_per_iteration": 2.712160348892212 + }, + { + "auxiliary_loss_clip": 0.0129867, + "auxiliary_loss_mlp": 0.01055214, + "balance_loss_clip": 1.08248854, + "balance_loss_mlp": 1.04176152, + "epoch": 0.04473035531774184, + "flos": 29752529299200.0, + "grad_norm": 2.3302537462959574, + "language_loss": 0.73756182, + "learning_rate": 3.997742533812924e-06, + "loss": 0.76110065, + "num_input_tokens_seen": 7840970, + "step": 372, + "time_per_iteration": 2.7309818267822266 + }, + { + "auxiliary_loss_clip": 0.01303978, + "auxiliary_loss_mlp": 0.01049706, + "balance_loss_clip": 1.08784819, + "balance_loss_mlp": 1.03577018, + "epoch": 0.04485059820838093, + "flos": 13151565676800.0, + "grad_norm": 3.2705447918352237, + "language_loss": 0.9271276, + "learning_rate": 3.997705381607001e-06, + "loss": 0.95066446, + "num_input_tokens_seen": 7857785, + "step": 373, + "time_per_iteration": 2.6399617195129395 + }, + { + "auxiliary_loss_clip": 0.01198749, + "auxiliary_loss_mlp": 0.01011213, + "balance_loss_clip": 1.0660615, + "balance_loss_mlp": 1.00682628, + "epoch": 0.04497084109902002, + "flos": 68094209548800.0, + "grad_norm": 0.9836124747511189, + "language_loss": 0.60252666, + "learning_rate": 3.997667926350761e-06, + "loss": 0.62462628, + "num_input_tokens_seen": 7916115, + "step": 374, + "time_per_iteration": 3.10239577293396 + }, + { + "auxiliary_loss_clip": 0.01196648, + "auxiliary_loss_mlp": 0.01007312, + "balance_loss_clip": 1.06474435, + "balance_loss_mlp": 1.00280559, + "epoch": 0.04509108398965911, + "flos": 64342263346560.0, + "grad_norm": 0.905668008621706, + "language_loss": 0.57798165, + "learning_rate": 3.997630168049886e-06, + "loss": 0.60002124, + "num_input_tokens_seen": 7974480, + "step": 375, + "time_per_iteration": 3.217982292175293 + }, + { + "auxiliary_loss_clip": 0.01306359, + "auxiliary_loss_mlp": 0.0104424, + "balance_loss_clip": 1.08775926, + "balance_loss_mlp": 1.03083456, + "epoch": 0.045211326880298205, + "flos": 22271115830400.0, + "grad_norm": 2.116741185652129, + "language_loss": 0.77618718, + "learning_rate": 3.997592106710101e-06, + "loss": 0.79969323, + "num_input_tokens_seen": 7993940, + "step": 376, + "time_per_iteration": 2.615664482116699 + }, + { + "auxiliary_loss_clip": 0.01300245, + "auxiliary_loss_mlp": 0.01053235, + "balance_loss_clip": 1.084162, + "balance_loss_mlp": 1.03960347, + "epoch": 0.045331569770937295, + "flos": 32159441796480.0, + "grad_norm": 3.137266415377361, + "language_loss": 0.66134745, + "learning_rate": 3.997553742337182e-06, + "loss": 0.68488228, + "num_input_tokens_seen": 8013365, + "step": 377, + "time_per_iteration": 2.7381834983825684 + }, + { + "auxiliary_loss_clip": 0.01301411, + "auxiliary_loss_mlp": 0.01053363, + "balance_loss_clip": 1.08613181, + "balance_loss_mlp": 1.04055953, + "epoch": 0.045451812661576385, + "flos": 22163455791360.0, + "grad_norm": 1.9492665660125146, + "language_loss": 0.91387641, + "learning_rate": 3.997515074936949e-06, + "loss": 0.93742406, + "num_input_tokens_seen": 8034240, + "step": 378, + "time_per_iteration": 2.6207218170166016 + }, + { + "auxiliary_loss_clip": 0.01297385, + "auxiliary_loss_mlp": 0.01045638, + "balance_loss_clip": 1.08355474, + "balance_loss_mlp": 1.03227496, + "epoch": 0.045572055552215475, + "flos": 16581968305920.0, + "grad_norm": 3.5398669458010072, + "language_loss": 0.86683911, + "learning_rate": 3.997476104515268e-06, + "loss": 0.89026928, + "num_input_tokens_seen": 8052430, + "step": 379, + "time_per_iteration": 2.6588921546936035 + }, + { + "auxiliary_loss_clip": 0.0129526, + "auxiliary_loss_mlp": 0.01044619, + "balance_loss_clip": 1.08435404, + "balance_loss_mlp": 1.03236401, + "epoch": 0.045692298442854565, + "flos": 17603375448960.0, + "grad_norm": 2.5194703361921604, + "language_loss": 0.77494735, + "learning_rate": 3.9974368310780485e-06, + "loss": 0.79834616, + "num_input_tokens_seen": 8069605, + "step": 380, + "time_per_iteration": 2.5814156532287598 + }, + { + "auxiliary_loss_clip": 0.0130738, + "auxiliary_loss_mlp": 0.01061646, + "balance_loss_clip": 1.08682084, + "balance_loss_mlp": 1.047436, + "epoch": 0.045812541333493655, + "flos": 26761098781440.0, + "grad_norm": 2.762949458350533, + "language_loss": 0.74577075, + "learning_rate": 3.997397254631251e-06, + "loss": 0.76946092, + "num_input_tokens_seen": 8090225, + "step": 381, + "time_per_iteration": 3.619298219680786 + }, + { + "auxiliary_loss_clip": 0.01190175, + "auxiliary_loss_mlp": 0.01026415, + "balance_loss_clip": 1.05861235, + "balance_loss_mlp": 1.02178943, + "epoch": 0.04593278422413275, + "flos": 60250349894400.0, + "grad_norm": 0.8273765413568055, + "language_loss": 0.60081244, + "learning_rate": 3.997357375180878e-06, + "loss": 0.62297833, + "num_input_tokens_seen": 8154505, + "step": 382, + "time_per_iteration": 3.2793266773223877 + }, + { + "auxiliary_loss_clip": 0.01295848, + "auxiliary_loss_mlp": 0.01051379, + "balance_loss_clip": 1.08219826, + "balance_loss_mlp": 1.03851628, + "epoch": 0.04605302711477184, + "flos": 21799249839360.0, + "grad_norm": 2.0993621747149613, + "language_loss": 0.75170445, + "learning_rate": 3.997317192732979e-06, + "loss": 0.77517676, + "num_input_tokens_seen": 8173285, + "step": 383, + "time_per_iteration": 3.5043282508850098 + }, + { + "auxiliary_loss_clip": 0.01298142, + "auxiliary_loss_mlp": 0.01053932, + "balance_loss_clip": 1.08233356, + "balance_loss_mlp": 1.03954375, + "epoch": 0.04617327000541093, + "flos": 19459705299840.0, + "grad_norm": 1.810003540733061, + "language_loss": 0.82497561, + "learning_rate": 3.99727670729365e-06, + "loss": 0.84849632, + "num_input_tokens_seen": 8191845, + "step": 384, + "time_per_iteration": 3.465014696121216 + }, + { + "auxiliary_loss_clip": 0.01298231, + "auxiliary_loss_mlp": 0.0105191, + "balance_loss_clip": 1.08554554, + "balance_loss_mlp": 1.03938079, + "epoch": 0.04629351289605002, + "flos": 25411468135680.0, + "grad_norm": 1.9416919996229298, + "language_loss": 0.77745366, + "learning_rate": 3.997235918869033e-06, + "loss": 0.80095506, + "num_input_tokens_seen": 8212880, + "step": 385, + "time_per_iteration": 2.757336139678955 + }, + { + "auxiliary_loss_clip": 0.01299321, + "auxiliary_loss_mlp": 0.01053274, + "balance_loss_clip": 1.08487654, + "balance_loss_mlp": 1.04023302, + "epoch": 0.04641375578668911, + "flos": 20558284813440.0, + "grad_norm": 2.183075926779125, + "language_loss": 0.82317382, + "learning_rate": 3.997194827465315e-06, + "loss": 0.84669983, + "num_input_tokens_seen": 8231475, + "step": 386, + "time_per_iteration": 2.621617555618286 + }, + { + "auxiliary_loss_clip": 0.01297931, + "auxiliary_loss_mlp": 0.01048016, + "balance_loss_clip": 1.08308935, + "balance_loss_mlp": 1.03437853, + "epoch": 0.0465339986773282, + "flos": 13188661447680.0, + "grad_norm": 2.915285899334016, + "language_loss": 0.91379786, + "learning_rate": 3.997153433088728e-06, + "loss": 0.93725729, + "num_input_tokens_seen": 8248600, + "step": 387, + "time_per_iteration": 2.589742422103882 + }, + { + "auxiliary_loss_clip": 0.01298954, + "auxiliary_loss_mlp": 0.01050822, + "balance_loss_clip": 1.08403444, + "balance_loss_mlp": 1.03707719, + "epoch": 0.0466542415679673, + "flos": 25556547168000.0, + "grad_norm": 1.809849255364686, + "language_loss": 0.81295216, + "learning_rate": 3.997111735745554e-06, + "loss": 0.83644986, + "num_input_tokens_seen": 8271570, + "step": 388, + "time_per_iteration": 2.6413590908050537 + }, + { + "auxiliary_loss_clip": 0.0129533, + "auxiliary_loss_mlp": 0.01045376, + "balance_loss_clip": 1.08322394, + "balance_loss_mlp": 1.03288269, + "epoch": 0.04677448445860639, + "flos": 22236749493120.0, + "grad_norm": 2.2429408254507126, + "language_loss": 0.82765275, + "learning_rate": 3.997069735442118e-06, + "loss": 0.85105979, + "num_input_tokens_seen": 8291265, + "step": 389, + "time_per_iteration": 2.5964834690093994 + }, + { + "auxiliary_loss_clip": 0.01295169, + "auxiliary_loss_mlp": 0.01052795, + "balance_loss_clip": 1.08252358, + "balance_loss_mlp": 1.0395273, + "epoch": 0.04689472734924548, + "flos": 28147825198080.0, + "grad_norm": 1.6056739006725278, + "language_loss": 0.80395365, + "learning_rate": 3.997027432184792e-06, + "loss": 0.82743323, + "num_input_tokens_seen": 8315925, + "step": 390, + "time_per_iteration": 2.6949026584625244 + }, + { + "auxiliary_loss_clip": 0.01293934, + "auxiliary_loss_mlp": 0.01044426, + "balance_loss_clip": 1.08066821, + "balance_loss_mlp": 1.03121781, + "epoch": 0.04701497023988457, + "flos": 23148952312320.0, + "grad_norm": 1.8737046216379776, + "language_loss": 0.89227974, + "learning_rate": 3.99698482597999e-06, + "loss": 0.91566336, + "num_input_tokens_seen": 8333605, + "step": 391, + "time_per_iteration": 2.6441550254821777 + }, + { + "auxiliary_loss_clip": 0.01180948, + "auxiliary_loss_mlp": 0.01001955, + "balance_loss_clip": 1.05214572, + "balance_loss_mlp": 0.99763966, + "epoch": 0.04713521313052366, + "flos": 64827668764800.0, + "grad_norm": 0.8700629774026185, + "language_loss": 0.63913035, + "learning_rate": 3.99694191683418e-06, + "loss": 0.66095936, + "num_input_tokens_seen": 8394405, + "step": 392, + "time_per_iteration": 3.258383274078369 + }, + { + "auxiliary_loss_clip": 0.01301264, + "auxiliary_loss_mlp": 0.01055715, + "balance_loss_clip": 1.08713627, + "balance_loss_mlp": 1.04109991, + "epoch": 0.047255456021162746, + "flos": 18771585477120.0, + "grad_norm": 2.5070191776787576, + "language_loss": 0.81781697, + "learning_rate": 3.996898704753867e-06, + "loss": 0.8413868, + "num_input_tokens_seen": 8412355, + "step": 393, + "time_per_iteration": 2.6314265727996826 + }, + { + "auxiliary_loss_clip": 0.01298251, + "auxiliary_loss_mlp": 0.01048645, + "balance_loss_clip": 1.08171844, + "balance_loss_mlp": 1.03490639, + "epoch": 0.04737569891180184, + "flos": 22053820504320.0, + "grad_norm": 2.7647938008001307, + "language_loss": 0.87784374, + "learning_rate": 3.996855189745609e-06, + "loss": 0.90131271, + "num_input_tokens_seen": 8431620, + "step": 394, + "time_per_iteration": 2.6061179637908936 + }, + { + "auxiliary_loss_clip": 0.01296107, + "auxiliary_loss_mlp": 0.01065119, + "balance_loss_clip": 1.08297896, + "balance_loss_mlp": 1.05120707, + "epoch": 0.04749594180244093, + "flos": 29057370410880.0, + "grad_norm": 1.918380578137693, + "language_loss": 0.92398453, + "learning_rate": 3.996811371816007e-06, + "loss": 0.94759679, + "num_input_tokens_seen": 8454045, + "step": 395, + "time_per_iteration": 2.7444663047790527 + }, + { + "auxiliary_loss_clip": 0.01295112, + "auxiliary_loss_mlp": 0.01056932, + "balance_loss_clip": 1.08324122, + "balance_loss_mlp": 1.04348493, + "epoch": 0.04761618469308002, + "flos": 35112268172160.0, + "grad_norm": 1.8928836966179192, + "language_loss": 0.77733541, + "learning_rate": 3.996767250971707e-06, + "loss": 0.80085588, + "num_input_tokens_seen": 8476785, + "step": 396, + "time_per_iteration": 2.846000909805298 + }, + { + "auxiliary_loss_clip": 0.01298315, + "auxiliary_loss_mlp": 0.0105093, + "balance_loss_clip": 1.0849421, + "balance_loss_mlp": 1.03874731, + "epoch": 0.04773642758371911, + "flos": 25630702796160.0, + "grad_norm": 4.021906739245687, + "language_loss": 0.8698988, + "learning_rate": 3.996722827219403e-06, + "loss": 0.89339125, + "num_input_tokens_seen": 8498400, + "step": 397, + "time_per_iteration": 2.7793469429016113 + }, + { + "auxiliary_loss_clip": 0.01296966, + "auxiliary_loss_mlp": 0.01049665, + "balance_loss_clip": 1.08337355, + "balance_loss_mlp": 1.03531194, + "epoch": 0.0478566704743582, + "flos": 20631506688000.0, + "grad_norm": 2.4168682817562077, + "language_loss": 0.8250556, + "learning_rate": 3.996678100565833e-06, + "loss": 0.84852189, + "num_input_tokens_seen": 8517455, + "step": 398, + "time_per_iteration": 2.6806962490081787 + }, + { + "auxiliary_loss_clip": 0.01294167, + "auxiliary_loss_mlp": 0.01052662, + "balance_loss_clip": 1.08231819, + "balance_loss_mlp": 1.03988862, + "epoch": 0.04797691336499729, + "flos": 18835721210880.0, + "grad_norm": 5.309619745661796, + "language_loss": 0.8835752, + "learning_rate": 3.996633071017783e-06, + "loss": 0.90704352, + "num_input_tokens_seen": 8534085, + "step": 399, + "time_per_iteration": 2.6742827892303467 + }, + { + "auxiliary_loss_clip": 0.01295851, + "auxiliary_loss_mlp": 0.01058969, + "balance_loss_clip": 1.08484197, + "balance_loss_mlp": 1.04527175, + "epoch": 0.04809715625563638, + "flos": 21099673578240.0, + "grad_norm": 2.3475826943489566, + "language_loss": 0.81540626, + "learning_rate": 3.996587738582084e-06, + "loss": 0.83895445, + "num_input_tokens_seen": 8550885, + "step": 400, + "time_per_iteration": 2.706061601638794 + }, + { + "auxiliary_loss_clip": 0.01292976, + "auxiliary_loss_mlp": 0.01050743, + "balance_loss_clip": 1.08157873, + "balance_loss_mlp": 1.03826141, + "epoch": 0.04821739914627548, + "flos": 23805650712960.0, + "grad_norm": 2.9839395616629205, + "language_loss": 0.86175275, + "learning_rate": 3.9965421032656115e-06, + "loss": 0.88519001, + "num_input_tokens_seen": 8570815, + "step": 401, + "time_per_iteration": 2.7361044883728027 + }, + { + "auxiliary_loss_clip": 0.01291077, + "auxiliary_loss_mlp": 0.01049124, + "balance_loss_clip": 1.08114362, + "balance_loss_mlp": 1.03664255, + "epoch": 0.04833764203691457, + "flos": 22200587475840.0, + "grad_norm": 2.3679509064145927, + "language_loss": 0.94306862, + "learning_rate": 3.99649616507529e-06, + "loss": 0.96647066, + "num_input_tokens_seen": 8589910, + "step": 402, + "time_per_iteration": 2.6628611087799072 + }, + { + "auxiliary_loss_clip": 0.01175973, + "auxiliary_loss_mlp": 0.01004453, + "balance_loss_clip": 1.04868567, + "balance_loss_mlp": 1.00042343, + "epoch": 0.04845788492755366, + "flos": 65904376896000.0, + "grad_norm": 0.9862401558800402, + "language_loss": 0.63082868, + "learning_rate": 3.996449924018088e-06, + "loss": 0.65263295, + "num_input_tokens_seen": 8650370, + "step": 403, + "time_per_iteration": 3.1459217071533203 + }, + { + "auxiliary_loss_clip": 0.0129566, + "auxiliary_loss_mlp": 0.0104715, + "balance_loss_clip": 1.08507454, + "balance_loss_mlp": 1.03414989, + "epoch": 0.04857812781819275, + "flos": 19281301424640.0, + "grad_norm": 1.8518587466222352, + "language_loss": 0.79446232, + "learning_rate": 3.99640338010102e-06, + "loss": 0.81789041, + "num_input_tokens_seen": 8669475, + "step": 404, + "time_per_iteration": 2.6388165950775146 + }, + { + "auxiliary_loss_clip": 0.01292825, + "auxiliary_loss_mlp": 0.01043025, + "balance_loss_clip": 1.08116484, + "balance_loss_mlp": 1.02988863, + "epoch": 0.04869837070883184, + "flos": 24062376193920.0, + "grad_norm": 1.9626407057949127, + "language_loss": 0.78810692, + "learning_rate": 3.996356533331146e-06, + "loss": 0.81146538, + "num_input_tokens_seen": 8691345, + "step": 405, + "time_per_iteration": 2.6542952060699463 + }, + { + "auxiliary_loss_clip": 0.01296827, + "auxiliary_loss_mlp": 0.01053931, + "balance_loss_clip": 1.08010101, + "balance_loss_mlp": 1.03948319, + "epoch": 0.04881861359947093, + "flos": 25187169657600.0, + "grad_norm": 2.085417777478674, + "language_loss": 0.61827564, + "learning_rate": 3.996309383715573e-06, + "loss": 0.64178324, + "num_input_tokens_seen": 8710125, + "step": 406, + "time_per_iteration": 2.789222240447998 + }, + { + "auxiliary_loss_clip": 0.01298533, + "auxiliary_loss_mlp": 0.01056671, + "balance_loss_clip": 1.08519697, + "balance_loss_mlp": 1.04268229, + "epoch": 0.048938856490110025, + "flos": 16362913213440.0, + "grad_norm": 5.835409776644763, + "language_loss": 0.73598611, + "learning_rate": 3.996261931261454e-06, + "loss": 0.75953817, + "num_input_tokens_seen": 8728705, + "step": 407, + "time_per_iteration": 3.5638177394866943 + }, + { + "auxiliary_loss_clip": 0.01297322, + "auxiliary_loss_mlp": 0.01051731, + "balance_loss_clip": 1.08500671, + "balance_loss_mlp": 1.0392797, + "epoch": 0.049059099380749115, + "flos": 29895094379520.0, + "grad_norm": 1.5734532798930945, + "language_loss": 0.86288208, + "learning_rate": 3.996214175975987e-06, + "loss": 0.88637263, + "num_input_tokens_seen": 8749225, + "step": 408, + "time_per_iteration": 3.554943799972534 + }, + { + "auxiliary_loss_clip": 0.01298593, + "auxiliary_loss_mlp": 0.01051723, + "balance_loss_clip": 1.08709502, + "balance_loss_mlp": 1.03779936, + "epoch": 0.049179342271388204, + "flos": 35918858027520.0, + "grad_norm": 2.163340387866288, + "language_loss": 0.79053247, + "learning_rate": 3.996166117866417e-06, + "loss": 0.81403565, + "num_input_tokens_seen": 8771160, + "step": 409, + "time_per_iteration": 2.740786075592041 + }, + { + "auxiliary_loss_clip": 0.0129051, + "auxiliary_loss_mlp": 0.01044582, + "balance_loss_clip": 1.08019876, + "balance_loss_mlp": 1.03227937, + "epoch": 0.049299585162027294, + "flos": 14611226659200.0, + "grad_norm": 2.118581201925525, + "language_loss": 0.86588931, + "learning_rate": 3.996117756940035e-06, + "loss": 0.88924015, + "num_input_tokens_seen": 8787845, + "step": 410, + "time_per_iteration": 4.35185170173645 + }, + { + "auxiliary_loss_clip": 0.01293029, + "auxiliary_loss_mlp": 0.01050356, + "balance_loss_clip": 1.08238447, + "balance_loss_mlp": 1.03639102, + "epoch": 0.049419828052666384, + "flos": 19567939956480.0, + "grad_norm": 2.0198351493791638, + "language_loss": 0.97566295, + "learning_rate": 3.996069093204175e-06, + "loss": 0.99909687, + "num_input_tokens_seen": 8803805, + "step": 411, + "time_per_iteration": 2.562957763671875 + }, + { + "auxiliary_loss_clip": 0.01296566, + "auxiliary_loss_mlp": 0.01057773, + "balance_loss_clip": 1.08426404, + "balance_loss_mlp": 1.04509521, + "epoch": 0.049540070943305474, + "flos": 13659916907520.0, + "grad_norm": 2.165081711231086, + "language_loss": 0.8811996, + "learning_rate": 3.996020126666221e-06, + "loss": 0.90474296, + "num_input_tokens_seen": 8820785, + "step": 412, + "time_per_iteration": 2.6026453971862793 + }, + { + "auxiliary_loss_clip": 0.01296181, + "auxiliary_loss_mlp": 0.01045226, + "balance_loss_clip": 1.08359051, + "balance_loss_mlp": 1.03267896, + "epoch": 0.04966031383394457, + "flos": 21832035978240.0, + "grad_norm": 2.307778857358149, + "language_loss": 0.82236099, + "learning_rate": 3.995970857333601e-06, + "loss": 0.84577501, + "num_input_tokens_seen": 8841195, + "step": 413, + "time_per_iteration": 2.6226089000701904 + }, + { + "auxiliary_loss_clip": 0.01294188, + "auxiliary_loss_mlp": 0.01052988, + "balance_loss_clip": 1.08169842, + "balance_loss_mlp": 1.03930879, + "epoch": 0.04978055672458366, + "flos": 28618793349120.0, + "grad_norm": 1.8340840259648281, + "language_loss": 0.79632199, + "learning_rate": 3.995921285213789e-06, + "loss": 0.81979382, + "num_input_tokens_seen": 8861455, + "step": 414, + "time_per_iteration": 2.673548698425293 + }, + { + "auxiliary_loss_clip": 0.01289354, + "auxiliary_loss_mlp": 0.01043798, + "balance_loss_clip": 1.08044446, + "balance_loss_mlp": 1.03108454, + "epoch": 0.04990079961522275, + "flos": 19828220883840.0, + "grad_norm": 2.7596917216462518, + "language_loss": 0.80496013, + "learning_rate": 3.995871410314305e-06, + "loss": 0.82829165, + "num_input_tokens_seen": 8880015, + "step": 415, + "time_per_iteration": 2.566574811935425 + }, + { + "auxiliary_loss_clip": 0.01179791, + "auxiliary_loss_mlp": 0.0100478, + "balance_loss_clip": 1.04446602, + "balance_loss_mlp": 1.00056028, + "epoch": 0.05002104250586184, + "flos": 62735045293440.0, + "grad_norm": 0.9151621284623224, + "language_loss": 0.59673333, + "learning_rate": 3.995821232642714e-06, + "loss": 0.61857903, + "num_input_tokens_seen": 8938420, + "step": 416, + "time_per_iteration": 3.311159372329712 + }, + { + "auxiliary_loss_clip": 0.01296677, + "auxiliary_loss_mlp": 0.01058982, + "balance_loss_clip": 1.08156085, + "balance_loss_mlp": 1.04621446, + "epoch": 0.05014128539650093, + "flos": 27928518710400.0, + "grad_norm": 2.027754073416025, + "language_loss": 0.82164413, + "learning_rate": 3.995770752206629e-06, + "loss": 0.84520072, + "num_input_tokens_seen": 8959495, + "step": 417, + "time_per_iteration": 2.7293853759765625 + }, + { + "auxiliary_loss_clip": 0.0129366, + "auxiliary_loss_mlp": 0.01050829, + "balance_loss_clip": 1.0824883, + "balance_loss_mlp": 1.03808594, + "epoch": 0.05026152828714002, + "flos": 17705576620800.0, + "grad_norm": 3.1120215200027945, + "language_loss": 0.96864808, + "learning_rate": 3.995719969013709e-06, + "loss": 0.99209291, + "num_input_tokens_seen": 8976675, + "step": 418, + "time_per_iteration": 2.6188268661499023 + }, + { + "auxiliary_loss_clip": 0.01300692, + "auxiliary_loss_mlp": 0.01047152, + "balance_loss_clip": 1.07728314, + "balance_loss_mlp": 1.03434861, + "epoch": 0.05038177117777912, + "flos": 19133277477120.0, + "grad_norm": 2.5423204191126225, + "language_loss": 0.85683626, + "learning_rate": 3.995668883071655e-06, + "loss": 0.88031471, + "num_input_tokens_seen": 8992900, + "step": 419, + "time_per_iteration": 2.751718759536743 + }, + { + "auxiliary_loss_clip": 0.01292763, + "auxiliary_loss_mlp": 0.01055084, + "balance_loss_clip": 1.08141458, + "balance_loss_mlp": 1.04128528, + "epoch": 0.050502014068418206, + "flos": 20667704618880.0, + "grad_norm": 2.3589061506459164, + "language_loss": 0.91028261, + "learning_rate": 3.995617494388219e-06, + "loss": 0.93376106, + "num_input_tokens_seen": 9011020, + "step": 420, + "time_per_iteration": 2.592637300491333 + }, + { + "auxiliary_loss_clip": 0.0129924, + "auxiliary_loss_mlp": 0.0105507, + "balance_loss_clip": 1.07578135, + "balance_loss_mlp": 1.04181433, + "epoch": 0.050622256959057296, + "flos": 21361103740800.0, + "grad_norm": 1.9301003860847699, + "language_loss": 0.80528933, + "learning_rate": 3.995565802971196e-06, + "loss": 0.82883239, + "num_input_tokens_seen": 9030995, + "step": 421, + "time_per_iteration": 2.708003520965576 + }, + { + "auxiliary_loss_clip": 0.0129783, + "auxiliary_loss_mlp": 0.01048795, + "balance_loss_clip": 1.07657552, + "balance_loss_mlp": 1.03642702, + "epoch": 0.050742499849696386, + "flos": 27673588909440.0, + "grad_norm": 2.0247253087986397, + "language_loss": 0.67457026, + "learning_rate": 3.995513808828427e-06, + "loss": 0.69803649, + "num_input_tokens_seen": 9053790, + "step": 422, + "time_per_iteration": 2.730257511138916 + }, + { + "auxiliary_loss_clip": 0.01299252, + "auxiliary_loss_mlp": 0.01052927, + "balance_loss_clip": 1.07550287, + "balance_loss_mlp": 1.03976071, + "epoch": 0.050862742740335476, + "flos": 19865999013120.0, + "grad_norm": 1.865309515794368, + "language_loss": 0.76610333, + "learning_rate": 3.9954615119678e-06, + "loss": 0.78962517, + "num_input_tokens_seen": 9072345, + "step": 423, + "time_per_iteration": 2.6885459423065186 + }, + { + "auxiliary_loss_clip": 0.01284811, + "auxiliary_loss_mlp": 0.01044652, + "balance_loss_clip": 1.07610202, + "balance_loss_mlp": 1.03169978, + "epoch": 0.050982985630974566, + "flos": 22085098272000.0, + "grad_norm": 2.0178706348738507, + "language_loss": 0.80842316, + "learning_rate": 3.995408912397248e-06, + "loss": 0.83171779, + "num_input_tokens_seen": 9090240, + "step": 424, + "time_per_iteration": 2.6687495708465576 + }, + { + "auxiliary_loss_clip": 0.01307589, + "auxiliary_loss_mlp": 0.01068049, + "balance_loss_clip": 1.07992911, + "balance_loss_mlp": 1.05480456, + "epoch": 0.05110322852161366, + "flos": 20740962407040.0, + "grad_norm": 2.4928021888021115, + "language_loss": 0.93297148, + "learning_rate": 3.99535601012475e-06, + "loss": 0.95672786, + "num_input_tokens_seen": 9105570, + "step": 425, + "time_per_iteration": 2.695495843887329 + }, + { + "auxiliary_loss_clip": 0.01306041, + "auxiliary_loss_mlp": 0.01383222, + "balance_loss_clip": 1.0768019, + "balance_loss_mlp": 1.00060558, + "epoch": 0.05122347141225275, + "flos": 28547295327360.0, + "grad_norm": 1.5825703217909484, + "language_loss": 0.75596654, + "learning_rate": 3.995302805158333e-06, + "loss": 0.78285909, + "num_input_tokens_seen": 9128225, + "step": 426, + "time_per_iteration": 2.792088031768799 + }, + { + "auxiliary_loss_clip": 0.01289816, + "auxiliary_loss_mlp": 0.01044532, + "balance_loss_clip": 1.07247186, + "balance_loss_mlp": 1.03013754, + "epoch": 0.05134371430289184, + "flos": 19722679747200.0, + "grad_norm": 1.8611426575640342, + "language_loss": 0.83538574, + "learning_rate": 3.9952492975060665e-06, + "loss": 0.85872924, + "num_input_tokens_seen": 9148295, + "step": 427, + "time_per_iteration": 2.73032546043396 + }, + { + "auxiliary_loss_clip": 0.01288252, + "auxiliary_loss_mlp": 0.01049052, + "balance_loss_clip": 1.07535386, + "balance_loss_mlp": 1.03520536, + "epoch": 0.05146395719353093, + "flos": 34458945649920.0, + "grad_norm": 2.2740563147544597, + "language_loss": 0.85092568, + "learning_rate": 3.995195487176067e-06, + "loss": 0.87429869, + "num_input_tokens_seen": 9168525, + "step": 428, + "time_per_iteration": 2.893331289291382 + }, + { + "auxiliary_loss_clip": 0.01288154, + "auxiliary_loss_mlp": 0.01043495, + "balance_loss_clip": 1.0800817, + "balance_loss_mlp": 1.0311991, + "epoch": 0.05158420008417002, + "flos": 21760286561280.0, + "grad_norm": 2.010881998898254, + "language_loss": 0.85752356, + "learning_rate": 3.995141374176499e-06, + "loss": 0.88084006, + "num_input_tokens_seen": 9186920, + "step": 429, + "time_per_iteration": 2.7725119590759277 + }, + { + "auxiliary_loss_clip": 0.01195737, + "auxiliary_loss_mlp": 0.01378784, + "balance_loss_clip": 1.03875947, + "balance_loss_mlp": 1.00037956, + "epoch": 0.05170444297480911, + "flos": 72553956226560.0, + "grad_norm": 0.8706990069175944, + "language_loss": 0.63080877, + "learning_rate": 3.995086958515572e-06, + "loss": 0.65655398, + "num_input_tokens_seen": 9244940, + "step": 430, + "time_per_iteration": 3.3441271781921387 + }, + { + "auxiliary_loss_clip": 0.01165786, + "auxiliary_loss_mlp": 0.01378706, + "balance_loss_clip": 1.04146886, + "balance_loss_mlp": 1.00038326, + "epoch": 0.05182468586544821, + "flos": 62416159326720.0, + "grad_norm": 0.8582693621026216, + "language_loss": 0.59943092, + "learning_rate": 3.995032240201538e-06, + "loss": 0.62487584, + "num_input_tokens_seen": 9307335, + "step": 431, + "time_per_iteration": 3.1941962242126465 + }, + { + "auxiliary_loss_clip": 0.01181757, + "auxiliary_loss_mlp": 0.01011745, + "balance_loss_clip": 1.03501725, + "balance_loss_mlp": 1.00764453, + "epoch": 0.0519449287560873, + "flos": 41225989432320.0, + "grad_norm": 1.0992074719225604, + "language_loss": 0.63164133, + "learning_rate": 3.9949772192427e-06, + "loss": 0.65357637, + "num_input_tokens_seen": 9353960, + "step": 432, + "time_per_iteration": 2.915034294128418 + }, + { + "auxiliary_loss_clip": 0.01296453, + "auxiliary_loss_mlp": 0.01051476, + "balance_loss_clip": 1.07246459, + "balance_loss_mlp": 1.0394001, + "epoch": 0.05206517164672639, + "flos": 17494530261120.0, + "grad_norm": 1.8929294182486038, + "language_loss": 0.79299378, + "learning_rate": 3.994921895647405e-06, + "loss": 0.81647313, + "num_input_tokens_seen": 9372130, + "step": 433, + "time_per_iteration": 3.5611155033111572 + }, + { + "auxiliary_loss_clip": 0.01161083, + "auxiliary_loss_mlp": 0.01005146, + "balance_loss_clip": 1.03761554, + "balance_loss_mlp": 1.00102127, + "epoch": 0.05218541453736548, + "flos": 64002762973440.0, + "grad_norm": 0.8395158463723692, + "language_loss": 0.55390692, + "learning_rate": 3.994866269424043e-06, + "loss": 0.57556915, + "num_input_tokens_seen": 9428500, + "step": 434, + "time_per_iteration": 4.039564609527588 + }, + { + "auxiliary_loss_clip": 0.01309303, + "auxiliary_loss_mlp": 0.01050506, + "balance_loss_clip": 1.05943656, + "balance_loss_mlp": 1.03768516, + "epoch": 0.05230565742800457, + "flos": 19317319787520.0, + "grad_norm": 2.182129608855124, + "language_loss": 0.78468353, + "learning_rate": 3.9948103405810545e-06, + "loss": 0.8082816, + "num_input_tokens_seen": 9447450, + "step": 435, + "time_per_iteration": 2.978563070297241 + }, + { + "auxiliary_loss_clip": 0.01281553, + "auxiliary_loss_mlp": 0.01049258, + "balance_loss_clip": 1.0661366, + "balance_loss_mlp": 1.03594244, + "epoch": 0.05242590031864366, + "flos": 25298636538240.0, + "grad_norm": 1.91187581959264, + "language_loss": 0.8614186, + "learning_rate": 3.994754109126923e-06, + "loss": 0.88472676, + "num_input_tokens_seen": 9468945, + "step": 436, + "time_per_iteration": 4.526792287826538 + }, + { + "auxiliary_loss_clip": 0.01306302, + "auxiliary_loss_mlp": 0.01047795, + "balance_loss_clip": 1.06742549, + "balance_loss_mlp": 1.03565347, + "epoch": 0.052546143209282754, + "flos": 26211629456640.0, + "grad_norm": 1.8596107071671824, + "language_loss": 0.93550575, + "learning_rate": 3.994697575070181e-06, + "loss": 0.95904666, + "num_input_tokens_seen": 9488405, + "step": 437, + "time_per_iteration": 2.8114922046661377 + }, + { + "auxiliary_loss_clip": 0.01301672, + "auxiliary_loss_mlp": 0.01043406, + "balance_loss_clip": 1.07932687, + "balance_loss_mlp": 1.03136015, + "epoch": 0.052666386099921844, + "flos": 22158140578560.0, + "grad_norm": 1.8303334457634768, + "language_loss": 0.91463923, + "learning_rate": 3.994640738419402e-06, + "loss": 0.93808997, + "num_input_tokens_seen": 9507780, + "step": 438, + "time_per_iteration": 2.7457644939422607 + }, + { + "auxiliary_loss_clip": 0.01297904, + "auxiliary_loss_mlp": 0.01042648, + "balance_loss_clip": 1.08022261, + "balance_loss_mlp": 1.03036928, + "epoch": 0.052786628990560934, + "flos": 23881817502720.0, + "grad_norm": 1.9035807112603134, + "language_loss": 0.80724287, + "learning_rate": 3.9945835991832075e-06, + "loss": 0.83064842, + "num_input_tokens_seen": 9529665, + "step": 439, + "time_per_iteration": 2.7689714431762695 + }, + { + "auxiliary_loss_clip": 0.0128817, + "auxiliary_loss_mlp": 0.01046899, + "balance_loss_clip": 1.08301592, + "balance_loss_mlp": 1.03410769, + "epoch": 0.052906871881200024, + "flos": 24605021934720.0, + "grad_norm": 2.4565858520976938, + "language_loss": 0.92963535, + "learning_rate": 3.994526157370268e-06, + "loss": 0.952986, + "num_input_tokens_seen": 9548280, + "step": 440, + "time_per_iteration": 2.647120952606201 + }, + { + "auxiliary_loss_clip": 0.01178893, + "auxiliary_loss_mlp": 0.0100423, + "balance_loss_clip": 1.03169274, + "balance_loss_mlp": 1.00024831, + "epoch": 0.053027114771839114, + "flos": 56461631143680.0, + "grad_norm": 0.8909285516005787, + "language_loss": 0.5927332, + "learning_rate": 3.994468412989296e-06, + "loss": 0.61456442, + "num_input_tokens_seen": 9609690, + "step": 441, + "time_per_iteration": 3.3138058185577393 + }, + { + "auxiliary_loss_clip": 0.01265567, + "auxiliary_loss_mlp": 0.010368, + "balance_loss_clip": 1.06838119, + "balance_loss_mlp": 1.02443767, + "epoch": 0.053147357662478203, + "flos": 17311098481920.0, + "grad_norm": 2.10418824520268, + "language_loss": 0.92669845, + "learning_rate": 3.994410366049052e-06, + "loss": 0.94972211, + "num_input_tokens_seen": 9627550, + "step": 442, + "time_per_iteration": 2.6490163803100586 + }, + { + "auxiliary_loss_clip": 0.01289961, + "auxiliary_loss_mlp": 0.01040068, + "balance_loss_clip": 1.07792139, + "balance_loss_mlp": 1.02753365, + "epoch": 0.0532676005531173, + "flos": 17164977955200.0, + "grad_norm": 2.122277037259409, + "language_loss": 0.82921338, + "learning_rate": 3.994352016558341e-06, + "loss": 0.85251367, + "num_input_tokens_seen": 9644855, + "step": 443, + "time_per_iteration": 2.6256234645843506 + }, + { + "auxiliary_loss_clip": 0.01294736, + "auxiliary_loss_mlp": 0.01046561, + "balance_loss_clip": 1.0796119, + "balance_loss_mlp": 1.03444958, + "epoch": 0.05338784344375639, + "flos": 27819960831360.0, + "grad_norm": 2.0448187466316368, + "language_loss": 0.74012256, + "learning_rate": 3.994293364526014e-06, + "loss": 0.7635355, + "num_input_tokens_seen": 9665740, + "step": 444, + "time_per_iteration": 2.7497100830078125 + }, + { + "auxiliary_loss_clip": 0.01290678, + "auxiliary_loss_mlp": 0.0104434, + "balance_loss_clip": 1.07520771, + "balance_loss_mlp": 1.03169179, + "epoch": 0.05350808633439548, + "flos": 21507691144320.0, + "grad_norm": 2.0015955122566984, + "language_loss": 0.84786248, + "learning_rate": 3.99423440996097e-06, + "loss": 0.87121266, + "num_input_tokens_seen": 9685280, + "step": 445, + "time_per_iteration": 2.715574264526367 + }, + { + "auxiliary_loss_clip": 0.01306379, + "auxiliary_loss_mlp": 0.01050381, + "balance_loss_clip": 1.07934844, + "balance_loss_mlp": 1.03818607, + "epoch": 0.05362832922503457, + "flos": 20084299920000.0, + "grad_norm": 2.401696545176789, + "language_loss": 0.81226927, + "learning_rate": 3.994175152872152e-06, + "loss": 0.83583689, + "num_input_tokens_seen": 9704365, + "step": 446, + "time_per_iteration": 2.7052063941955566 + }, + { + "auxiliary_loss_clip": 0.01291094, + "auxiliary_loss_mlp": 0.01041362, + "balance_loss_clip": 1.0748347, + "balance_loss_mlp": 1.02993619, + "epoch": 0.05374857211567366, + "flos": 26137222433280.0, + "grad_norm": 2.0855102825739857, + "language_loss": 0.78899765, + "learning_rate": 3.994115593268548e-06, + "loss": 0.81232214, + "num_input_tokens_seen": 9724145, + "step": 447, + "time_per_iteration": 2.6672239303588867 + }, + { + "auxiliary_loss_clip": 0.01288378, + "auxiliary_loss_mlp": 0.01044393, + "balance_loss_clip": 1.08241415, + "balance_loss_mlp": 1.03174496, + "epoch": 0.05386881500631275, + "flos": 27486817165440.0, + "grad_norm": 2.218067736171111, + "language_loss": 0.82269925, + "learning_rate": 3.994055731159195e-06, + "loss": 0.84602696, + "num_input_tokens_seen": 9741615, + "step": 448, + "time_per_iteration": 2.7176408767700195 + }, + { + "auxiliary_loss_clip": 0.0129664, + "auxiliary_loss_mlp": 0.01039782, + "balance_loss_clip": 1.08275557, + "balance_loss_mlp": 1.02791512, + "epoch": 0.053989057896951846, + "flos": 23585087249280.0, + "grad_norm": 2.408398786212518, + "language_loss": 0.86964178, + "learning_rate": 3.993995566553172e-06, + "loss": 0.89300597, + "num_input_tokens_seen": 9760580, + "step": 449, + "time_per_iteration": 2.6967313289642334 + }, + { + "auxiliary_loss_clip": 0.01279115, + "auxiliary_loss_mlp": 0.0103983, + "balance_loss_clip": 1.06956875, + "balance_loss_mlp": 1.02776623, + "epoch": 0.054109300787590936, + "flos": 25228862369280.0, + "grad_norm": 1.6967878699645167, + "language_loss": 0.77080327, + "learning_rate": 3.993935099459607e-06, + "loss": 0.79399276, + "num_input_tokens_seen": 9782195, + "step": 450, + "time_per_iteration": 2.950282573699951 + }, + { + "auxiliary_loss_clip": 0.01282351, + "auxiliary_loss_mlp": 0.01051093, + "balance_loss_clip": 1.07958794, + "balance_loss_mlp": 1.04045916, + "epoch": 0.054229543678230026, + "flos": 23841525421440.0, + "grad_norm": 1.8678099120926026, + "language_loss": 0.74092984, + "learning_rate": 3.993874329887673e-06, + "loss": 0.76426429, + "num_input_tokens_seen": 9800850, + "step": 451, + "time_per_iteration": 2.879185676574707 + }, + { + "auxiliary_loss_clip": 0.01293412, + "auxiliary_loss_mlp": 0.01045352, + "balance_loss_clip": 1.08015478, + "balance_loss_mlp": 1.03283477, + "epoch": 0.054349786568869116, + "flos": 16320933192960.0, + "grad_norm": 6.165448948422885, + "language_loss": 0.86430633, + "learning_rate": 3.993813257846589e-06, + "loss": 0.887694, + "num_input_tokens_seen": 9817605, + "step": 452, + "time_per_iteration": 2.7187552452087402 + }, + { + "auxiliary_loss_clip": 0.01297982, + "auxiliary_loss_mlp": 0.01049621, + "balance_loss_clip": 1.07979548, + "balance_loss_mlp": 1.03743148, + "epoch": 0.054470029459508205, + "flos": 18660729127680.0, + "grad_norm": 2.2430591305495358, + "language_loss": 0.93076777, + "learning_rate": 3.993751883345619e-06, + "loss": 0.95424378, + "num_input_tokens_seen": 9835965, + "step": 453, + "time_per_iteration": 2.6522226333618164 + }, + { + "auxiliary_loss_clip": 0.01286752, + "auxiliary_loss_mlp": 0.01051468, + "balance_loss_clip": 1.07373571, + "balance_loss_mlp": 1.0392251, + "epoch": 0.054590272350147295, + "flos": 17785298856960.0, + "grad_norm": 2.524232017686473, + "language_loss": 0.8750819, + "learning_rate": 3.993690206394073e-06, + "loss": 0.89846408, + "num_input_tokens_seen": 9852265, + "step": 454, + "time_per_iteration": 2.714414596557617 + }, + { + "auxiliary_loss_clip": 0.01305088, + "auxiliary_loss_mlp": 0.01055765, + "balance_loss_clip": 1.07644844, + "balance_loss_mlp": 1.04384446, + "epoch": 0.054710515240786385, + "flos": 17785945301760.0, + "grad_norm": 3.062835340621488, + "language_loss": 0.8787024, + "learning_rate": 3.993628227001307e-06, + "loss": 0.90231097, + "num_input_tokens_seen": 9870465, + "step": 455, + "time_per_iteration": 2.809573173522949 + }, + { + "auxiliary_loss_clip": 0.01290651, + "auxiliary_loss_mlp": 0.01045914, + "balance_loss_clip": 1.07201886, + "balance_loss_mlp": 1.03535223, + "epoch": 0.05483075813142548, + "flos": 48210900180480.0, + "grad_norm": 1.9004327826476004, + "language_loss": 0.7119776, + "learning_rate": 3.993565945176726e-06, + "loss": 0.73534334, + "num_input_tokens_seen": 9891490, + "step": 456, + "time_per_iteration": 3.0415186882019043 + }, + { + "auxiliary_loss_clip": 0.01297137, + "auxiliary_loss_mlp": 0.01045912, + "balance_loss_clip": 1.07788813, + "balance_loss_mlp": 1.03462839, + "epoch": 0.05495100102206457, + "flos": 19682244011520.0, + "grad_norm": 1.9788636482117148, + "language_loss": 0.84328628, + "learning_rate": 3.993503360929776e-06, + "loss": 0.8667168, + "num_input_tokens_seen": 9910375, + "step": 457, + "time_per_iteration": 2.7019729614257812 + }, + { + "auxiliary_loss_clip": 0.01325633, + "auxiliary_loss_mlp": 0.01048692, + "balance_loss_clip": 1.06746936, + "balance_loss_mlp": 1.03594255, + "epoch": 0.05507124391270366, + "flos": 26360048453760.0, + "grad_norm": 1.737803084952984, + "language_loss": 0.81212974, + "learning_rate": 3.99344047426995e-06, + "loss": 0.83587301, + "num_input_tokens_seen": 9931635, + "step": 458, + "time_per_iteration": 3.2282330989837646 + }, + { + "auxiliary_loss_clip": 0.01313994, + "auxiliary_loss_mlp": 0.0104827, + "balance_loss_clip": 1.07270169, + "balance_loss_mlp": 1.03576505, + "epoch": 0.05519148680334275, + "flos": 22601314581120.0, + "grad_norm": 2.0565246045121715, + "language_loss": 0.93422657, + "learning_rate": 3.993377285206789e-06, + "loss": 0.95784914, + "num_input_tokens_seen": 9951420, + "step": 459, + "time_per_iteration": 4.050993204116821 + }, + { + "auxiliary_loss_clip": 0.01269833, + "auxiliary_loss_mlp": 0.01057396, + "balance_loss_clip": 1.06689954, + "balance_loss_mlp": 1.04518342, + "epoch": 0.05531172969398184, + "flos": 40552519380480.0, + "grad_norm": 1.6965632127298806, + "language_loss": 0.86701953, + "learning_rate": 3.99331379374988e-06, + "loss": 0.89029181, + "num_input_tokens_seen": 9975025, + "step": 460, + "time_per_iteration": 4.417022466659546 + }, + { + "auxiliary_loss_clip": 0.01292357, + "auxiliary_loss_mlp": 0.01047336, + "balance_loss_clip": 1.06885874, + "balance_loss_mlp": 1.03517628, + "epoch": 0.05543197258462093, + "flos": 23477894087040.0, + "grad_norm": 2.0748245070579427, + "language_loss": 0.80171847, + "learning_rate": 3.993249999908852e-06, + "loss": 0.82511544, + "num_input_tokens_seen": 9995175, + "step": 461, + "time_per_iteration": 2.825619697570801 + }, + { + "auxiliary_loss_clip": 0.01286159, + "auxiliary_loss_mlp": 0.01064615, + "balance_loss_clip": 1.07872653, + "balance_loss_mlp": 1.05225849, + "epoch": 0.05555221547526003, + "flos": 18624603024000.0, + "grad_norm": 2.202175254693931, + "language_loss": 0.87429464, + "learning_rate": 3.993185903693384e-06, + "loss": 0.89780241, + "num_input_tokens_seen": 10011975, + "step": 462, + "time_per_iteration": 4.442130088806152 + }, + { + "auxiliary_loss_clip": 0.01298884, + "auxiliary_loss_mlp": 0.01040623, + "balance_loss_clip": 1.07508862, + "balance_loss_mlp": 1.02813578, + "epoch": 0.05567245836589912, + "flos": 23587098410880.0, + "grad_norm": 2.0063765519235024, + "language_loss": 0.82534027, + "learning_rate": 3.9931215051131995e-06, + "loss": 0.84873533, + "num_input_tokens_seen": 10032620, + "step": 463, + "time_per_iteration": 2.791595458984375 + }, + { + "auxiliary_loss_clip": 0.01300876, + "auxiliary_loss_mlp": 0.01061067, + "balance_loss_clip": 1.07091737, + "balance_loss_mlp": 1.04887199, + "epoch": 0.05579270125653821, + "flos": 27746667129600.0, + "grad_norm": 1.6977634962694703, + "language_loss": 0.80204356, + "learning_rate": 3.993056804178068e-06, + "loss": 0.82566297, + "num_input_tokens_seen": 10054165, + "step": 464, + "time_per_iteration": 2.783304452896118 + }, + { + "auxiliary_loss_clip": 0.01305758, + "auxiliary_loss_mlp": 0.01045496, + "balance_loss_clip": 1.07292747, + "balance_loss_mlp": 1.03284752, + "epoch": 0.0559129441471773, + "flos": 27014161075200.0, + "grad_norm": 2.141788637320988, + "language_loss": 0.84692723, + "learning_rate": 3.992991800897803e-06, + "loss": 0.87043977, + "num_input_tokens_seen": 10073970, + "step": 465, + "time_per_iteration": 2.8651375770568848 + }, + { + "auxiliary_loss_clip": 0.01282244, + "auxiliary_loss_mlp": 0.01047073, + "balance_loss_clip": 1.0788753, + "balance_loss_mlp": 1.03484249, + "epoch": 0.05603318703781639, + "flos": 15229787794560.0, + "grad_norm": 2.224270217594689, + "language_loss": 0.90066993, + "learning_rate": 3.9929264952822665e-06, + "loss": 0.92396307, + "num_input_tokens_seen": 10091505, + "step": 466, + "time_per_iteration": 2.673694372177124 + }, + { + "auxiliary_loss_clip": 0.01296586, + "auxiliary_loss_mlp": 0.0104938, + "balance_loss_clip": 1.07587624, + "balance_loss_mlp": 1.0365119, + "epoch": 0.05615342992845548, + "flos": 22266482976000.0, + "grad_norm": 1.913039028080947, + "language_loss": 0.88055229, + "learning_rate": 3.992860887341366e-06, + "loss": 0.90401196, + "num_input_tokens_seen": 10109675, + "step": 467, + "time_per_iteration": 2.664595603942871 + }, + { + "auxiliary_loss_clip": 0.01284193, + "auxiliary_loss_mlp": 0.01056456, + "balance_loss_clip": 1.0696274, + "balance_loss_mlp": 1.04418373, + "epoch": 0.056273672819094574, + "flos": 23584979508480.0, + "grad_norm": 2.4733171010997923, + "language_loss": 0.81123626, + "learning_rate": 3.992794977085052e-06, + "loss": 0.83464277, + "num_input_tokens_seen": 10127675, + "step": 468, + "time_per_iteration": 2.8174874782562256 + }, + { + "auxiliary_loss_clip": 0.01296297, + "auxiliary_loss_mlp": 0.01049216, + "balance_loss_clip": 1.06643927, + "balance_loss_mlp": 1.03633583, + "epoch": 0.056393915709733664, + "flos": 19858708552320.0, + "grad_norm": 1.964497549829656, + "language_loss": 0.84829569, + "learning_rate": 3.992728764523326e-06, + "loss": 0.87175077, + "num_input_tokens_seen": 10146620, + "step": 469, + "time_per_iteration": 2.899806499481201 + }, + { + "auxiliary_loss_clip": 0.01298184, + "auxiliary_loss_mlp": 0.01050403, + "balance_loss_clip": 1.07588887, + "balance_loss_mlp": 1.03800535, + "epoch": 0.05651415860037275, + "flos": 22163779013760.0, + "grad_norm": 2.9619916549191396, + "language_loss": 0.80744088, + "learning_rate": 3.99266224966623e-06, + "loss": 0.83092678, + "num_input_tokens_seen": 10167535, + "step": 470, + "time_per_iteration": 2.8497776985168457 + }, + { + "auxiliary_loss_clip": 0.01279814, + "auxiliary_loss_mlp": 0.01044937, + "balance_loss_clip": 1.07181728, + "balance_loss_mlp": 1.03323698, + "epoch": 0.05663440149101184, + "flos": 19463548055040.0, + "grad_norm": 1.9059033710477369, + "language_loss": 0.87789321, + "learning_rate": 3.992595432523855e-06, + "loss": 0.90114069, + "num_input_tokens_seen": 10184825, + "step": 471, + "time_per_iteration": 2.8111653327941895 + }, + { + "auxiliary_loss_clip": 0.01283856, + "auxiliary_loss_mlp": 0.01036463, + "balance_loss_clip": 1.06718123, + "balance_loss_mlp": 1.0249238, + "epoch": 0.05675464438165093, + "flos": 22670226823680.0, + "grad_norm": 1.8394350958931798, + "language_loss": 0.86096215, + "learning_rate": 3.992528313106338e-06, + "loss": 0.88416535, + "num_input_tokens_seen": 10203025, + "step": 472, + "time_per_iteration": 2.766505479812622 + }, + { + "auxiliary_loss_clip": 0.01284026, + "auxiliary_loss_mlp": 0.01382776, + "balance_loss_clip": 1.07978153, + "balance_loss_mlp": 1.00035501, + "epoch": 0.05687488727229002, + "flos": 16901177495040.0, + "grad_norm": 2.802929988624542, + "language_loss": 0.81868541, + "learning_rate": 3.9924608914238595e-06, + "loss": 0.84535336, + "num_input_tokens_seen": 10218020, + "step": 473, + "time_per_iteration": 2.724492311477661 + }, + { + "auxiliary_loss_clip": 0.01287298, + "auxiliary_loss_mlp": 0.01047691, + "balance_loss_clip": 1.07354665, + "balance_loss_mlp": 1.0352633, + "epoch": 0.05699513016292912, + "flos": 29168980945920.0, + "grad_norm": 4.2307201716242595, + "language_loss": 0.84011734, + "learning_rate": 3.992393167486648e-06, + "loss": 0.86346716, + "num_input_tokens_seen": 10237170, + "step": 474, + "time_per_iteration": 2.8145859241485596 + }, + { + "auxiliary_loss_clip": 0.01281241, + "auxiliary_loss_mlp": 0.01048142, + "balance_loss_clip": 1.07706857, + "balance_loss_mlp": 1.03531551, + "epoch": 0.05711537305356821, + "flos": 18916197632640.0, + "grad_norm": 2.4492354640859344, + "language_loss": 0.8082642, + "learning_rate": 3.992325141304977e-06, + "loss": 0.83155799, + "num_input_tokens_seen": 10255125, + "step": 475, + "time_per_iteration": 2.7680068016052246 + }, + { + "auxiliary_loss_clip": 0.01284525, + "auxiliary_loss_mlp": 0.01043882, + "balance_loss_clip": 1.06880796, + "balance_loss_mlp": 1.03177631, + "epoch": 0.0572356159442073, + "flos": 26758979879040.0, + "grad_norm": 3.694045151324362, + "language_loss": 0.86201477, + "learning_rate": 3.992256812889166e-06, + "loss": 0.88529885, + "num_input_tokens_seen": 10271230, + "step": 476, + "time_per_iteration": 2.870018720626831 + }, + { + "auxiliary_loss_clip": 0.01286, + "auxiliary_loss_mlp": 0.01048499, + "balance_loss_clip": 1.07976866, + "balance_loss_mlp": 1.03600526, + "epoch": 0.05735585883484639, + "flos": 35116146840960.0, + "grad_norm": 2.3293338010770333, + "language_loss": 0.76917601, + "learning_rate": 3.992188182249582e-06, + "loss": 0.79252094, + "num_input_tokens_seen": 10293125, + "step": 477, + "time_per_iteration": 2.7926712036132812 + }, + { + "auxiliary_loss_clip": 0.01294454, + "auxiliary_loss_mlp": 0.01050176, + "balance_loss_clip": 1.07447743, + "balance_loss_mlp": 1.03733671, + "epoch": 0.05747610172548548, + "flos": 18734381965440.0, + "grad_norm": 4.0732910062890735, + "language_loss": 0.90612763, + "learning_rate": 3.992119249396633e-06, + "loss": 0.92957395, + "num_input_tokens_seen": 10311810, + "step": 478, + "time_per_iteration": 2.848651647567749 + }, + { + "auxiliary_loss_clip": 0.01285961, + "auxiliary_loss_mlp": 0.01383069, + "balance_loss_clip": 1.07020581, + "balance_loss_mlp": 1.00037253, + "epoch": 0.05759634461612457, + "flos": 27964752554880.0, + "grad_norm": 1.8107920102306305, + "language_loss": 0.82009649, + "learning_rate": 3.992050014340778e-06, + "loss": 0.84678674, + "num_input_tokens_seen": 10332165, + "step": 479, + "time_per_iteration": 2.931861162185669 + }, + { + "auxiliary_loss_clip": 0.01163174, + "auxiliary_loss_mlp": 0.01009853, + "balance_loss_clip": 1.03705788, + "balance_loss_mlp": 1.00589561, + "epoch": 0.057716587506763666, + "flos": 69292009405440.0, + "grad_norm": 0.8414860853206028, + "language_loss": 0.54998374, + "learning_rate": 3.99198047709252e-06, + "loss": 0.57171392, + "num_input_tokens_seen": 10393685, + "step": 480, + "time_per_iteration": 3.3499057292938232 + }, + { + "auxiliary_loss_clip": 0.01294636, + "auxiliary_loss_mlp": 0.01051104, + "balance_loss_clip": 1.06721878, + "balance_loss_mlp": 1.03806901, + "epoch": 0.057836830397402755, + "flos": 25009196745600.0, + "grad_norm": 2.3928223693688935, + "language_loss": 0.78511894, + "learning_rate": 3.991910637662408e-06, + "loss": 0.80857635, + "num_input_tokens_seen": 10413975, + "step": 481, + "time_per_iteration": 2.8415801525115967 + }, + { + "auxiliary_loss_clip": 0.0128008, + "auxiliary_loss_mlp": 0.01040518, + "balance_loss_clip": 1.07691503, + "balance_loss_mlp": 1.02925265, + "epoch": 0.057957073288041845, + "flos": 25593894334080.0, + "grad_norm": 1.853781086470067, + "language_loss": 0.80460429, + "learning_rate": 3.9918404960610355e-06, + "loss": 0.82781029, + "num_input_tokens_seen": 10433005, + "step": 482, + "time_per_iteration": 2.942938804626465 + }, + { + "auxiliary_loss_clip": 0.01295097, + "auxiliary_loss_mlp": 0.01049881, + "balance_loss_clip": 1.07610416, + "balance_loss_mlp": 1.03794217, + "epoch": 0.058077316178680935, + "flos": 20777411733120.0, + "grad_norm": 2.320408683691299, + "language_loss": 0.77270961, + "learning_rate": 3.991770052299043e-06, + "loss": 0.79615939, + "num_input_tokens_seen": 10451235, + "step": 483, + "time_per_iteration": 2.9692587852478027 + }, + { + "auxiliary_loss_clip": 0.01297839, + "auxiliary_loss_mlp": 0.01050586, + "balance_loss_clip": 1.07233834, + "balance_loss_mlp": 1.03896356, + "epoch": 0.058197559069320025, + "flos": 18916484941440.0, + "grad_norm": 2.6616943302176406, + "language_loss": 0.8759504, + "learning_rate": 3.991699306387118e-06, + "loss": 0.89943463, + "num_input_tokens_seen": 10469705, + "step": 484, + "time_per_iteration": 2.74149489402771 + }, + { + "auxiliary_loss_clip": 0.01289792, + "auxiliary_loss_mlp": 0.01057048, + "balance_loss_clip": 1.07584369, + "balance_loss_mlp": 1.04513335, + "epoch": 0.058317801959959115, + "flos": 24863327614080.0, + "grad_norm": 3.2267808197245262, + "language_loss": 0.78071254, + "learning_rate": 3.991628258335991e-06, + "loss": 0.80418098, + "num_input_tokens_seen": 10491910, + "step": 485, + "time_per_iteration": 3.7101409435272217 + }, + { + "auxiliary_loss_clip": 0.01292002, + "auxiliary_loss_mlp": 0.01040068, + "balance_loss_clip": 1.0692699, + "balance_loss_mlp": 1.02775943, + "epoch": 0.05843804485059821, + "flos": 23257977068160.0, + "grad_norm": 2.832700749115095, + "language_loss": 0.87682962, + "learning_rate": 3.991556908156442e-06, + "loss": 0.9001503, + "num_input_tokens_seen": 10508435, + "step": 486, + "time_per_iteration": 3.936037540435791 + }, + { + "auxiliary_loss_clip": 0.01298545, + "auxiliary_loss_mlp": 0.01053651, + "balance_loss_clip": 1.07188737, + "balance_loss_mlp": 1.04156303, + "epoch": 0.0585582877412373, + "flos": 23150532510720.0, + "grad_norm": 1.8810932802410325, + "language_loss": 0.87616682, + "learning_rate": 3.9914852558592914e-06, + "loss": 0.89968872, + "num_input_tokens_seen": 10529485, + "step": 487, + "time_per_iteration": 2.7993087768554688 + }, + { + "auxiliary_loss_clip": 0.01288546, + "auxiliary_loss_mlp": 0.01053984, + "balance_loss_clip": 1.07499027, + "balance_loss_mlp": 1.04130006, + "epoch": 0.05867853063187639, + "flos": 23506406507520.0, + "grad_norm": 3.2705974379994207, + "language_loss": 0.80246538, + "learning_rate": 3.991413301455413e-06, + "loss": 0.82589066, + "num_input_tokens_seen": 10545935, + "step": 488, + "time_per_iteration": 4.533856153488159 + }, + { + "auxiliary_loss_clip": 0.01281704, + "auxiliary_loss_mlp": 0.01054277, + "balance_loss_clip": 1.07132626, + "balance_loss_mlp": 1.04224277, + "epoch": 0.05879877352251548, + "flos": 29495803818240.0, + "grad_norm": 2.3927070199306235, + "language_loss": 0.7793231, + "learning_rate": 3.991341044955719e-06, + "loss": 0.80268294, + "num_input_tokens_seen": 10565690, + "step": 489, + "time_per_iteration": 2.83793306350708 + }, + { + "auxiliary_loss_clip": 0.01285996, + "auxiliary_loss_mlp": 0.01383039, + "balance_loss_clip": 1.07310081, + "balance_loss_mlp": 1.00037909, + "epoch": 0.05891901641315457, + "flos": 20157485880960.0, + "grad_norm": 2.2131682603610128, + "language_loss": 0.8155061, + "learning_rate": 3.991268486371172e-06, + "loss": 0.84219646, + "num_input_tokens_seen": 10584245, + "step": 490, + "time_per_iteration": 2.679490566253662 + }, + { + "auxiliary_loss_clip": 0.0129291, + "auxiliary_loss_mlp": 0.01044388, + "balance_loss_clip": 1.07005012, + "balance_loss_mlp": 1.03258061, + "epoch": 0.05903925930379366, + "flos": 24644200694400.0, + "grad_norm": 6.557698021257683, + "language_loss": 0.87761897, + "learning_rate": 3.991195625712779e-06, + "loss": 0.90099192, + "num_input_tokens_seen": 10601210, + "step": 491, + "time_per_iteration": 2.76204776763916 + }, + { + "auxiliary_loss_clip": 0.01281747, + "auxiliary_loss_mlp": 0.01042419, + "balance_loss_clip": 1.07844472, + "balance_loss_mlp": 1.03114784, + "epoch": 0.05915950219443276, + "flos": 21250391045760.0, + "grad_norm": 2.4716363878134224, + "language_loss": 0.81770486, + "learning_rate": 3.991122462991592e-06, + "loss": 0.84094656, + "num_input_tokens_seen": 10620730, + "step": 492, + "time_per_iteration": 2.704596519470215 + }, + { + "auxiliary_loss_clip": 0.01282263, + "auxiliary_loss_mlp": 0.01043218, + "balance_loss_clip": 1.07629693, + "balance_loss_mlp": 1.03076041, + "epoch": 0.05927974508507185, + "flos": 9902727319680.0, + "grad_norm": 3.686687170909808, + "language_loss": 0.8127079, + "learning_rate": 3.991048998218712e-06, + "loss": 0.83596265, + "num_input_tokens_seen": 10634035, + "step": 493, + "time_per_iteration": 2.59844970703125 + }, + { + "auxiliary_loss_clip": 0.01285172, + "auxiliary_loss_mlp": 0.01048505, + "balance_loss_clip": 1.07230353, + "balance_loss_mlp": 1.0365603, + "epoch": 0.05939998797571094, + "flos": 18259499232000.0, + "grad_norm": 2.482395333925963, + "language_loss": 0.76119006, + "learning_rate": 3.990975231405281e-06, + "loss": 0.78452682, + "num_input_tokens_seen": 10652485, + "step": 494, + "time_per_iteration": 2.654465436935425 + }, + { + "auxiliary_loss_clip": 0.01286765, + "auxiliary_loss_mlp": 0.01049919, + "balance_loss_clip": 1.07654929, + "balance_loss_mlp": 1.03815877, + "epoch": 0.05952023086635003, + "flos": 28256598558720.0, + "grad_norm": 1.838205808503201, + "language_loss": 0.78688526, + "learning_rate": 3.990901162562491e-06, + "loss": 0.81025219, + "num_input_tokens_seen": 10673175, + "step": 495, + "time_per_iteration": 2.7682135105133057 + }, + { + "auxiliary_loss_clip": 0.01297865, + "auxiliary_loss_mlp": 0.01382738, + "balance_loss_clip": 1.06527305, + "balance_loss_mlp": 1.00037313, + "epoch": 0.05964047375698912, + "flos": 14902498045440.0, + "grad_norm": 1.9864883110444103, + "language_loss": 0.90544295, + "learning_rate": 3.9908267917015765e-06, + "loss": 0.93224895, + "num_input_tokens_seen": 10691235, + "step": 496, + "time_per_iteration": 2.7021806240081787 + }, + { + "auxiliary_loss_clip": 0.01271503, + "auxiliary_loss_mlp": 0.01040164, + "balance_loss_clip": 1.0696311, + "balance_loss_mlp": 1.02808774, + "epoch": 0.059760716647628206, + "flos": 23185581206400.0, + "grad_norm": 2.2252128532413233, + "language_loss": 0.93012333, + "learning_rate": 3.990752118833821e-06, + "loss": 0.9532401, + "num_input_tokens_seen": 10708675, + "step": 497, + "time_per_iteration": 2.7194628715515137 + }, + { + "auxiliary_loss_clip": 0.01281524, + "auxiliary_loss_mlp": 0.01047955, + "balance_loss_clip": 1.07884264, + "balance_loss_mlp": 1.03599191, + "epoch": 0.0598809595382673, + "flos": 22746968231040.0, + "grad_norm": 1.7829258445594511, + "language_loss": 0.77992815, + "learning_rate": 3.990677143970553e-06, + "loss": 0.80322289, + "num_input_tokens_seen": 10729485, + "step": 498, + "time_per_iteration": 2.6941134929656982 + }, + { + "auxiliary_loss_clip": 0.01298601, + "auxiliary_loss_mlp": 0.0104877, + "balance_loss_clip": 1.07570529, + "balance_loss_mlp": 1.03641987, + "epoch": 0.06000120242890639, + "flos": 22127221946880.0, + "grad_norm": 6.168674719311637, + "language_loss": 0.81395477, + "learning_rate": 3.990601867123144e-06, + "loss": 0.83742845, + "num_input_tokens_seen": 10749210, + "step": 499, + "time_per_iteration": 2.8175835609436035 + }, + { + "auxiliary_loss_clip": 0.01299086, + "auxiliary_loss_mlp": 0.01046483, + "balance_loss_clip": 1.06626844, + "balance_loss_mlp": 1.03449106, + "epoch": 0.06012144531954548, + "flos": 19171773878400.0, + "grad_norm": 2.167634809381349, + "language_loss": 0.8475678, + "learning_rate": 3.990526288303014e-06, + "loss": 0.87102354, + "num_input_tokens_seen": 10768000, + "step": 500, + "time_per_iteration": 2.816568613052368 + }, + { + "auxiliary_loss_clip": 0.01290741, + "auxiliary_loss_mlp": 0.01382753, + "balance_loss_clip": 1.07178104, + "balance_loss_mlp": 1.00029469, + "epoch": 0.06024168821018457, + "flos": 22783345729920.0, + "grad_norm": 3.5145882269873456, + "language_loss": 0.90975344, + "learning_rate": 3.9904504075216295e-06, + "loss": 0.93648839, + "num_input_tokens_seen": 10788760, + "step": 501, + "time_per_iteration": 2.783174991607666 + }, + { + "auxiliary_loss_clip": 0.01298526, + "auxiliary_loss_mlp": 0.01048495, + "balance_loss_clip": 1.06727338, + "balance_loss_mlp": 1.03560865, + "epoch": 0.06036193110082366, + "flos": 18770687637120.0, + "grad_norm": 2.307617683304687, + "language_loss": 0.93877572, + "learning_rate": 3.990374224790501e-06, + "loss": 0.96224588, + "num_input_tokens_seen": 10806965, + "step": 502, + "time_per_iteration": 2.772639036178589 + }, + { + "auxiliary_loss_clip": 0.01295856, + "auxiliary_loss_mlp": 0.01046069, + "balance_loss_clip": 1.07434773, + "balance_loss_mlp": 1.03362906, + "epoch": 0.06048217399146275, + "flos": 17201570935680.0, + "grad_norm": 2.1999664867884774, + "language_loss": 0.70960361, + "learning_rate": 3.990297740121185e-06, + "loss": 0.73302281, + "num_input_tokens_seen": 10824900, + "step": 503, + "time_per_iteration": 2.7304792404174805 + }, + { + "auxiliary_loss_clip": 0.01285889, + "auxiliary_loss_mlp": 0.01382282, + "balance_loss_clip": 1.07463121, + "balance_loss_mlp": 1.00030053, + "epoch": 0.06060241688210185, + "flos": 24024131187840.0, + "grad_norm": 1.8734048573839204, + "language_loss": 0.7840811, + "learning_rate": 3.990220953525284e-06, + "loss": 0.81076276, + "num_input_tokens_seen": 10842010, + "step": 504, + "time_per_iteration": 2.7260677814483643 + }, + { + "auxiliary_loss_clip": 0.01273314, + "auxiliary_loss_mlp": 0.01051135, + "balance_loss_clip": 1.06545675, + "balance_loss_mlp": 1.03873134, + "epoch": 0.06072265977274094, + "flos": 14611190745600.0, + "grad_norm": 3.0670023115789595, + "language_loss": 0.7411021, + "learning_rate": 3.9901438650144465e-06, + "loss": 0.7643466, + "num_input_tokens_seen": 10858260, + "step": 505, + "time_per_iteration": 2.7004878520965576 + }, + { + "auxiliary_loss_clip": 0.01271996, + "auxiliary_loss_mlp": 0.01048691, + "balance_loss_clip": 1.07145143, + "balance_loss_mlp": 1.03616202, + "epoch": 0.06084290266338003, + "flos": 20558284813440.0, + "grad_norm": 3.991851191322872, + "language_loss": 0.9166317, + "learning_rate": 3.990066474600367e-06, + "loss": 0.93983853, + "num_input_tokens_seen": 10876230, + "step": 506, + "time_per_iteration": 2.707181930541992 + }, + { + "auxiliary_loss_clip": 0.01272231, + "auxiliary_loss_mlp": 0.01043974, + "balance_loss_clip": 1.06860685, + "balance_loss_mlp": 1.03258312, + "epoch": 0.06096314555401912, + "flos": 22309217182080.0, + "grad_norm": 2.0183781389534827, + "language_loss": 0.6766867, + "learning_rate": 3.989988782294786e-06, + "loss": 0.69984877, + "num_input_tokens_seen": 10896320, + "step": 507, + "time_per_iteration": 2.74017071723938 + }, + { + "auxiliary_loss_clip": 0.01275423, + "auxiliary_loss_mlp": 0.01048456, + "balance_loss_clip": 1.0630188, + "balance_loss_mlp": 1.0365293, + "epoch": 0.06108338844465821, + "flos": 19131374056320.0, + "grad_norm": 1.6917743062452455, + "language_loss": 0.94756615, + "learning_rate": 3.989910788109489e-06, + "loss": 0.97080505, + "num_input_tokens_seen": 10912970, + "step": 508, + "time_per_iteration": 2.9066200256347656 + }, + { + "auxiliary_loss_clip": 0.01294441, + "auxiliary_loss_mlp": 0.01053519, + "balance_loss_clip": 1.06852794, + "balance_loss_mlp": 1.04094839, + "epoch": 0.0612036313352973, + "flos": 33584018169600.0, + "grad_norm": 2.5599946070046316, + "language_loss": 0.74875319, + "learning_rate": 3.989832492056307e-06, + "loss": 0.77223277, + "num_input_tokens_seen": 10933995, + "step": 509, + "time_per_iteration": 2.8571178913116455 + }, + { + "auxiliary_loss_clip": 0.01284589, + "auxiliary_loss_mlp": 0.01044741, + "balance_loss_clip": 1.0745647, + "balance_loss_mlp": 1.03267682, + "epoch": 0.06132387422593639, + "flos": 27490552179840.0, + "grad_norm": 4.30976042829379, + "language_loss": 0.8084591, + "learning_rate": 3.989753894147119e-06, + "loss": 0.83175242, + "num_input_tokens_seen": 10954120, + "step": 510, + "time_per_iteration": 2.6979448795318604 + }, + { + "auxiliary_loss_clip": 0.01280737, + "auxiliary_loss_mlp": 0.0105176, + "balance_loss_clip": 1.07607889, + "balance_loss_mlp": 1.04024434, + "epoch": 0.061444117116575485, + "flos": 25885057979520.0, + "grad_norm": 1.795357519954519, + "language_loss": 0.79997289, + "learning_rate": 3.989674994393846e-06, + "loss": 0.82329786, + "num_input_tokens_seen": 10973595, + "step": 511, + "time_per_iteration": 3.6602108478546143 + }, + { + "auxiliary_loss_clip": 0.01282683, + "auxiliary_loss_mlp": 0.01048679, + "balance_loss_clip": 1.07832384, + "balance_loss_mlp": 1.03659678, + "epoch": 0.061564360007214575, + "flos": 28512031150080.0, + "grad_norm": 2.7845465051410105, + "language_loss": 0.94214225, + "learning_rate": 3.98959579280846e-06, + "loss": 0.96545577, + "num_input_tokens_seen": 10991995, + "step": 512, + "time_per_iteration": 2.706982135772705 + }, + { + "auxiliary_loss_clip": 0.01284922, + "auxiliary_loss_mlp": 0.01057315, + "balance_loss_clip": 1.06795394, + "balance_loss_mlp": 1.04414201, + "epoch": 0.061684602897853665, + "flos": 12094355652480.0, + "grad_norm": 2.1085887790692257, + "language_loss": 0.83152258, + "learning_rate": 3.989516289402973e-06, + "loss": 0.85494494, + "num_input_tokens_seen": 11007625, + "step": 513, + "time_per_iteration": 3.672677516937256 + }, + { + "auxiliary_loss_clip": 0.01277009, + "auxiliary_loss_mlp": 0.01048766, + "balance_loss_clip": 1.0563252, + "balance_loss_mlp": 1.03665984, + "epoch": 0.061804845788492754, + "flos": 19532639865600.0, + "grad_norm": 3.079419271898475, + "language_loss": 0.80496764, + "learning_rate": 3.989436484189447e-06, + "loss": 0.82822537, + "num_input_tokens_seen": 11025570, + "step": 514, + "time_per_iteration": 4.771639108657837 + }, + { + "auxiliary_loss_clip": 0.01284151, + "auxiliary_loss_mlp": 0.01047907, + "balance_loss_clip": 1.07113016, + "balance_loss_mlp": 1.03548551, + "epoch": 0.061925088679131844, + "flos": 15341111020800.0, + "grad_norm": 2.8412083094602383, + "language_loss": 0.80667865, + "learning_rate": 3.9893563771799885e-06, + "loss": 0.82999921, + "num_input_tokens_seen": 11042045, + "step": 515, + "time_per_iteration": 3.1121065616607666 + }, + { + "auxiliary_loss_clip": 0.01278298, + "auxiliary_loss_mlp": 0.01045473, + "balance_loss_clip": 1.07702422, + "balance_loss_mlp": 1.03305137, + "epoch": 0.062045331569770934, + "flos": 25919927107200.0, + "grad_norm": 2.3179580727567184, + "language_loss": 0.85948473, + "learning_rate": 3.989275968386749e-06, + "loss": 0.8827225, + "num_input_tokens_seen": 11059955, + "step": 516, + "time_per_iteration": 2.640467882156372 + }, + { + "auxiliary_loss_clip": 0.01278069, + "auxiliary_loss_mlp": 0.01042938, + "balance_loss_clip": 1.06606317, + "balance_loss_mlp": 1.03074861, + "epoch": 0.06216557446041003, + "flos": 28110621686400.0, + "grad_norm": 2.2892445824827248, + "language_loss": 0.76666236, + "learning_rate": 3.989195257821926e-06, + "loss": 0.78987241, + "num_input_tokens_seen": 11078440, + "step": 517, + "time_per_iteration": 2.808838129043579 + }, + { + "auxiliary_loss_clip": 0.01291015, + "auxiliary_loss_mlp": 0.01050493, + "balance_loss_clip": 1.07493174, + "balance_loss_mlp": 1.03818476, + "epoch": 0.06228581735104912, + "flos": 23478181395840.0, + "grad_norm": 2.049755553605639, + "language_loss": 0.84555942, + "learning_rate": 3.989114245497765e-06, + "loss": 0.86897445, + "num_input_tokens_seen": 11098240, + "step": 518, + "time_per_iteration": 2.723484516143799 + }, + { + "auxiliary_loss_clip": 0.01284602, + "auxiliary_loss_mlp": 0.01048991, + "balance_loss_clip": 1.0719794, + "balance_loss_mlp": 1.03755903, + "epoch": 0.06240606024168821, + "flos": 15195205975680.0, + "grad_norm": 2.178340079581876, + "language_loss": 0.94563937, + "learning_rate": 3.989032931426554e-06, + "loss": 0.96897531, + "num_input_tokens_seen": 11115395, + "step": 519, + "time_per_iteration": 2.702824115753174 + }, + { + "auxiliary_loss_clip": 0.0128731, + "auxiliary_loss_mlp": 0.01043531, + "balance_loss_clip": 1.07103169, + "balance_loss_mlp": 1.03205061, + "epoch": 0.06252630313232731, + "flos": 20631829910400.0, + "grad_norm": 2.1832210609898097, + "language_loss": 0.86758983, + "learning_rate": 3.9889513156206295e-06, + "loss": 0.89089823, + "num_input_tokens_seen": 11134835, + "step": 520, + "time_per_iteration": 2.7074596881866455 + }, + { + "auxiliary_loss_clip": 0.01296227, + "auxiliary_loss_mlp": 0.01045947, + "balance_loss_clip": 1.06884515, + "balance_loss_mlp": 1.03371, + "epoch": 0.06264654602296639, + "flos": 20778058177920.0, + "grad_norm": 2.569648066122966, + "language_loss": 0.73458427, + "learning_rate": 3.988869398092371e-06, + "loss": 0.75800598, + "num_input_tokens_seen": 11154745, + "step": 521, + "time_per_iteration": 2.8289942741394043 + }, + { + "auxiliary_loss_clip": 0.01291421, + "auxiliary_loss_mlp": 0.01045142, + "balance_loss_clip": 1.07170892, + "balance_loss_mlp": 1.03229761, + "epoch": 0.06276678891360549, + "flos": 29605798241280.0, + "grad_norm": 2.269432694239991, + "language_loss": 0.78898257, + "learning_rate": 3.988787178854206e-06, + "loss": 0.81234819, + "num_input_tokens_seen": 11174280, + "step": 522, + "time_per_iteration": 2.8013925552368164 + }, + { + "auxiliary_loss_clip": 0.01277966, + "auxiliary_loss_mlp": 0.01053239, + "balance_loss_clip": 1.07596397, + "balance_loss_mlp": 1.03917205, + "epoch": 0.06288703180424457, + "flos": 22126288193280.0, + "grad_norm": 2.8180318943764493, + "language_loss": 0.87548524, + "learning_rate": 3.988704657918608e-06, + "loss": 0.89879727, + "num_input_tokens_seen": 11193340, + "step": 523, + "time_per_iteration": 2.654170513153076 + }, + { + "auxiliary_loss_clip": 0.01287914, + "auxiliary_loss_mlp": 0.01041177, + "balance_loss_clip": 1.07798696, + "balance_loss_mlp": 1.02950597, + "epoch": 0.06300727469488367, + "flos": 14976689587200.0, + "grad_norm": 2.445801302515699, + "language_loss": 0.79649425, + "learning_rate": 3.988621835298094e-06, + "loss": 0.81978512, + "num_input_tokens_seen": 11210555, + "step": 524, + "time_per_iteration": 2.663442611694336 + }, + { + "auxiliary_loss_clip": 0.01277145, + "auxiliary_loss_mlp": 0.01046978, + "balance_loss_clip": 1.07853544, + "balance_loss_mlp": 1.03530777, + "epoch": 0.06312751758552275, + "flos": 24535391420160.0, + "grad_norm": 2.1587138447359697, + "language_loss": 0.91557652, + "learning_rate": 3.988538711005229e-06, + "loss": 0.93881774, + "num_input_tokens_seen": 11230010, + "step": 525, + "time_per_iteration": 2.689793109893799 + }, + { + "auxiliary_loss_clip": 0.01274103, + "auxiliary_loss_mlp": 0.01055728, + "balance_loss_clip": 1.06976271, + "balance_loss_mlp": 1.04389608, + "epoch": 0.06324776047616185, + "flos": 21507008785920.0, + "grad_norm": 2.4870989132557333, + "language_loss": 0.88449395, + "learning_rate": 3.988455285052622e-06, + "loss": 0.90779221, + "num_input_tokens_seen": 11246190, + "step": 526, + "time_per_iteration": 2.6288039684295654 + }, + { + "auxiliary_loss_clip": 0.01276303, + "auxiliary_loss_mlp": 0.01049691, + "balance_loss_clip": 1.07256532, + "balance_loss_mlp": 1.03808594, + "epoch": 0.06336800336680094, + "flos": 21688034353920.0, + "grad_norm": 2.926780930285324, + "language_loss": 0.83734834, + "learning_rate": 3.98837155745293e-06, + "loss": 0.86060834, + "num_input_tokens_seen": 11264230, + "step": 527, + "time_per_iteration": 2.5793583393096924 + }, + { + "auxiliary_loss_clip": 0.01285738, + "auxiliary_loss_mlp": 0.01044574, + "balance_loss_clip": 1.0778172, + "balance_loss_mlp": 1.03347564, + "epoch": 0.06348824625744003, + "flos": 19500895221120.0, + "grad_norm": 3.777597441634588, + "language_loss": 0.75811696, + "learning_rate": 3.988287528218854e-06, + "loss": 0.78142011, + "num_input_tokens_seen": 11283015, + "step": 528, + "time_per_iteration": 2.5873467922210693 + }, + { + "auxiliary_loss_clip": 0.01284484, + "auxiliary_loss_mlp": 0.01056861, + "balance_loss_clip": 1.07931948, + "balance_loss_mlp": 1.04579258, + "epoch": 0.06360848914807912, + "flos": 15481233976320.0, + "grad_norm": 1.9372033440387515, + "language_loss": 0.90347195, + "learning_rate": 3.98820319736314e-06, + "loss": 0.92688537, + "num_input_tokens_seen": 11299630, + "step": 529, + "time_per_iteration": 2.523776054382324 + }, + { + "auxiliary_loss_clip": 0.01298661, + "auxiliary_loss_mlp": 0.01047074, + "balance_loss_clip": 1.06929708, + "balance_loss_mlp": 1.03499234, + "epoch": 0.0637287320387182, + "flos": 20593369422720.0, + "grad_norm": 1.9516672009790363, + "language_loss": 0.85460091, + "learning_rate": 3.988118564898582e-06, + "loss": 0.87805837, + "num_input_tokens_seen": 11319170, + "step": 530, + "time_per_iteration": 2.7321274280548096 + }, + { + "auxiliary_loss_clip": 0.01289228, + "auxiliary_loss_mlp": 0.01383089, + "balance_loss_clip": 1.07187343, + "balance_loss_mlp": 1.00040269, + "epoch": 0.0638489749293573, + "flos": 17412222245760.0, + "grad_norm": 2.6567981685190944, + "language_loss": 0.89233536, + "learning_rate": 3.988033630838019e-06, + "loss": 0.9190585, + "num_input_tokens_seen": 11333210, + "step": 531, + "time_per_iteration": 2.688222885131836 + }, + { + "auxiliary_loss_clip": 0.01284938, + "auxiliary_loss_mlp": 0.0104689, + "balance_loss_clip": 1.07753277, + "balance_loss_mlp": 1.03536844, + "epoch": 0.0639692178199964, + "flos": 23807661874560.0, + "grad_norm": 1.818599553005039, + "language_loss": 0.88319594, + "learning_rate": 3.987948395194334e-06, + "loss": 0.90651423, + "num_input_tokens_seen": 11355590, + "step": 532, + "time_per_iteration": 2.73616886138916 + }, + { + "auxiliary_loss_clip": 0.01272133, + "auxiliary_loss_mlp": 0.01056228, + "balance_loss_clip": 1.07117701, + "balance_loss_mlp": 1.04478359, + "epoch": 0.06408946071063548, + "flos": 18477225521280.0, + "grad_norm": 2.5145922449800175, + "language_loss": 0.7655493, + "learning_rate": 3.987862857980458e-06, + "loss": 0.7888329, + "num_input_tokens_seen": 11371535, + "step": 533, + "time_per_iteration": 2.6529417037963867 + }, + { + "auxiliary_loss_clip": 0.01304433, + "auxiliary_loss_mlp": 0.01053183, + "balance_loss_clip": 1.07201767, + "balance_loss_mlp": 1.04079103, + "epoch": 0.06420970360127458, + "flos": 27162220936320.0, + "grad_norm": 2.3202250728116116, + "language_loss": 0.76831609, + "learning_rate": 3.987777019209368e-06, + "loss": 0.79189229, + "num_input_tokens_seen": 11392050, + "step": 534, + "time_per_iteration": 2.809584140777588 + }, + { + "auxiliary_loss_clip": 0.01278449, + "auxiliary_loss_mlp": 0.01048571, + "balance_loss_clip": 1.07878864, + "balance_loss_mlp": 1.0366621, + "epoch": 0.06432994649191366, + "flos": 23659673840640.0, + "grad_norm": 4.18178051680393, + "language_loss": 0.80998784, + "learning_rate": 3.987690878894084e-06, + "loss": 0.83325803, + "num_input_tokens_seen": 11411765, + "step": 535, + "time_per_iteration": 2.7172837257385254 + }, + { + "auxiliary_loss_clip": 0.01290765, + "auxiliary_loss_mlp": 0.01049039, + "balance_loss_clip": 1.0725832, + "balance_loss_mlp": 1.03679049, + "epoch": 0.06445018938255276, + "flos": 23403953940480.0, + "grad_norm": 2.2971400148663474, + "language_loss": 0.84997952, + "learning_rate": 3.987604437047673e-06, + "loss": 0.87337756, + "num_input_tokens_seen": 11431565, + "step": 536, + "time_per_iteration": 2.732740640640259 + }, + { + "auxiliary_loss_clip": 0.01284829, + "auxiliary_loss_mlp": 0.01040344, + "balance_loss_clip": 1.07587385, + "balance_loss_mlp": 1.02902555, + "epoch": 0.06457043227319184, + "flos": 19646692525440.0, + "grad_norm": 2.137490759348878, + "language_loss": 0.77690911, + "learning_rate": 3.987517693683251e-06, + "loss": 0.80016083, + "num_input_tokens_seen": 11450140, + "step": 537, + "time_per_iteration": 3.563037633895874 + }, + { + "auxiliary_loss_clip": 0.01284046, + "auxiliary_loss_mlp": 0.01050168, + "balance_loss_clip": 1.07469404, + "balance_loss_mlp": 1.03847969, + "epoch": 0.06469067516383094, + "flos": 16978744915200.0, + "grad_norm": 2.8440766892714455, + "language_loss": 0.9620207, + "learning_rate": 3.9874306488139745e-06, + "loss": 0.98536283, + "num_input_tokens_seen": 11465400, + "step": 538, + "time_per_iteration": 2.6534340381622314 + }, + { + "auxiliary_loss_clip": 0.01293503, + "auxiliary_loss_mlp": 0.0104209, + "balance_loss_clip": 1.07088542, + "balance_loss_mlp": 1.0308187, + "epoch": 0.06481091805447003, + "flos": 23296401642240.0, + "grad_norm": 2.0206278127683817, + "language_loss": 0.8795383, + "learning_rate": 3.987343302453049e-06, + "loss": 0.90289426, + "num_input_tokens_seen": 11486675, + "step": 539, + "time_per_iteration": 3.6805825233459473 + }, + { + "auxiliary_loss_clip": 0.01286355, + "auxiliary_loss_mlp": 0.01050424, + "balance_loss_clip": 1.07416534, + "balance_loss_mlp": 1.03825235, + "epoch": 0.06493116094510912, + "flos": 29172356824320.0, + "grad_norm": 1.7300217800579845, + "language_loss": 0.8256703, + "learning_rate": 3.987255654613724e-06, + "loss": 0.84903806, + "num_input_tokens_seen": 11510440, + "step": 540, + "time_per_iteration": 3.668525457382202 + }, + { + "auxiliary_loss_clip": 0.01293484, + "auxiliary_loss_mlp": 0.01051152, + "balance_loss_clip": 1.0700171, + "balance_loss_mlp": 1.03938627, + "epoch": 0.06505140383574821, + "flos": 19865065259520.0, + "grad_norm": 2.513175330480901, + "language_loss": 0.70321906, + "learning_rate": 3.987167705309296e-06, + "loss": 0.72666538, + "num_input_tokens_seen": 11529715, + "step": 541, + "time_per_iteration": 2.6802258491516113 + }, + { + "auxiliary_loss_clip": 0.01283702, + "auxiliary_loss_mlp": 0.01382256, + "balance_loss_clip": 1.07503271, + "balance_loss_mlp": 1.0003593, + "epoch": 0.0651716467263873, + "flos": 17924703540480.0, + "grad_norm": 2.1811852257114652, + "language_loss": 0.9543274, + "learning_rate": 3.987079454553108e-06, + "loss": 0.98098695, + "num_input_tokens_seen": 11547665, + "step": 542, + "time_per_iteration": 2.6474602222442627 + }, + { + "auxiliary_loss_clip": 0.01290409, + "auxiliary_loss_mlp": 0.01040577, + "balance_loss_clip": 1.07334399, + "balance_loss_mlp": 1.02928221, + "epoch": 0.0652918896170264, + "flos": 20842840356480.0, + "grad_norm": 1.8043276982793526, + "language_loss": 0.91206962, + "learning_rate": 3.986990902358546e-06, + "loss": 0.93537951, + "num_input_tokens_seen": 11564605, + "step": 543, + "time_per_iteration": 2.674638032913208 + }, + { + "auxiliary_loss_clip": 0.01282279, + "auxiliary_loss_mlp": 0.01053356, + "balance_loss_clip": 1.07332635, + "balance_loss_mlp": 1.04068398, + "epoch": 0.06541213250766549, + "flos": 21872507627520.0, + "grad_norm": 2.101832066202977, + "language_loss": 0.93098974, + "learning_rate": 3.986902048739045e-06, + "loss": 0.95434612, + "num_input_tokens_seen": 11584550, + "step": 544, + "time_per_iteration": 2.631817102432251 + }, + { + "auxiliary_loss_clip": 0.01291507, + "auxiliary_loss_mlp": 0.01044322, + "balance_loss_clip": 1.07329035, + "balance_loss_mlp": 1.03192401, + "epoch": 0.06553237539830457, + "flos": 23110743219840.0, + "grad_norm": 3.3045642232727195, + "language_loss": 0.79686511, + "learning_rate": 3.986812893708082e-06, + "loss": 0.82022333, + "num_input_tokens_seen": 11600740, + "step": 545, + "time_per_iteration": 2.6695072650909424 + }, + { + "auxiliary_loss_clip": 0.0129211, + "auxiliary_loss_mlp": 0.01047951, + "balance_loss_clip": 1.07013416, + "balance_loss_mlp": 1.03477836, + "epoch": 0.06565261828894367, + "flos": 17923769786880.0, + "grad_norm": 1.9940766109246306, + "language_loss": 0.81238735, + "learning_rate": 3.9867234372791826e-06, + "loss": 0.83578795, + "num_input_tokens_seen": 11618695, + "step": 546, + "time_per_iteration": 2.700145959854126 + }, + { + "auxiliary_loss_clip": 0.0128018, + "auxiliary_loss_mlp": 0.01040001, + "balance_loss_clip": 1.07439506, + "balance_loss_mlp": 1.02825868, + "epoch": 0.06577286117958275, + "flos": 22783058421120.0, + "grad_norm": 1.5996289706712536, + "language_loss": 0.87071115, + "learning_rate": 3.986633679465918e-06, + "loss": 0.89391297, + "num_input_tokens_seen": 11638850, + "step": 547, + "time_per_iteration": 2.7447190284729004 + }, + { + "auxiliary_loss_clip": 0.01304028, + "auxiliary_loss_mlp": 0.01050256, + "balance_loss_clip": 1.06962943, + "balance_loss_mlp": 1.03819251, + "epoch": 0.06589310407022185, + "flos": 23696194993920.0, + "grad_norm": 2.0731878933116086, + "language_loss": 0.80656731, + "learning_rate": 3.986543620281904e-06, + "loss": 0.83011013, + "num_input_tokens_seen": 11658500, + "step": 548, + "time_per_iteration": 3.1247732639312744 + }, + { + "auxiliary_loss_clip": 0.01269455, + "auxiliary_loss_mlp": 0.01049632, + "balance_loss_clip": 1.0707022, + "balance_loss_mlp": 1.03726387, + "epoch": 0.06601334696086093, + "flos": 26864772410880.0, + "grad_norm": 1.728783639469958, + "language_loss": 0.9131664, + "learning_rate": 3.986453259740802e-06, + "loss": 0.93635738, + "num_input_tokens_seen": 11676670, + "step": 549, + "time_per_iteration": 2.9940640926361084 + }, + { + "auxiliary_loss_clip": 0.01292396, + "auxiliary_loss_mlp": 0.01048784, + "balance_loss_clip": 1.07618952, + "balance_loss_mlp": 1.03736961, + "epoch": 0.06613358985150003, + "flos": 12567694101120.0, + "grad_norm": 3.511758724118501, + "language_loss": 0.79120791, + "learning_rate": 3.986362597856319e-06, + "loss": 0.81461978, + "num_input_tokens_seen": 11693170, + "step": 550, + "time_per_iteration": 2.6634278297424316 + }, + { + "auxiliary_loss_clip": 0.01277252, + "auxiliary_loss_mlp": 0.01383338, + "balance_loss_clip": 1.06782627, + "balance_loss_mlp": 1.00031793, + "epoch": 0.06625383274213913, + "flos": 18332505624960.0, + "grad_norm": 2.3291214444409376, + "language_loss": 0.81418848, + "learning_rate": 3.986271634642211e-06, + "loss": 0.84079432, + "num_input_tokens_seen": 11710150, + "step": 551, + "time_per_iteration": 2.7383804321289062 + }, + { + "auxiliary_loss_clip": 0.01272902, + "auxiliary_loss_mlp": 0.01038363, + "balance_loss_clip": 1.07669592, + "balance_loss_mlp": 1.02538729, + "epoch": 0.06637407563277821, + "flos": 15375585098880.0, + "grad_norm": 2.175281466528723, + "language_loss": 0.81557596, + "learning_rate": 3.986180370112274e-06, + "loss": 0.83868861, + "num_input_tokens_seen": 11726670, + "step": 552, + "time_per_iteration": 2.597367763519287 + }, + { + "auxiliary_loss_clip": 0.01283756, + "auxiliary_loss_mlp": 0.01382766, + "balance_loss_clip": 1.07684517, + "balance_loss_mlp": 1.00029302, + "epoch": 0.0664943185234173, + "flos": 24025244509440.0, + "grad_norm": 1.7249995253227388, + "language_loss": 0.74494249, + "learning_rate": 3.986088804280354e-06, + "loss": 0.77160776, + "num_input_tokens_seen": 11746400, + "step": 553, + "time_per_iteration": 2.7655389308929443 + }, + { + "auxiliary_loss_clip": 0.01282921, + "auxiliary_loss_mlp": 0.01043672, + "balance_loss_clip": 1.06902957, + "balance_loss_mlp": 1.03170323, + "epoch": 0.06661456141405639, + "flos": 20957503547520.0, + "grad_norm": 2.638844898852347, + "language_loss": 0.94379735, + "learning_rate": 3.985996937160342e-06, + "loss": 0.96706331, + "num_input_tokens_seen": 11765590, + "step": 554, + "time_per_iteration": 2.754848003387451 + }, + { + "auxiliary_loss_clip": 0.01280055, + "auxiliary_loss_mlp": 0.01035347, + "balance_loss_clip": 1.07698393, + "balance_loss_mlp": 1.02340245, + "epoch": 0.06673480430469549, + "flos": 52223953322880.0, + "grad_norm": 2.119834846392197, + "language_loss": 0.68720347, + "learning_rate": 3.985904768766173e-06, + "loss": 0.71035743, + "num_input_tokens_seen": 11788365, + "step": 555, + "time_per_iteration": 3.000251293182373 + }, + { + "auxiliary_loss_clip": 0.01297254, + "auxiliary_loss_mlp": 0.01042095, + "balance_loss_clip": 1.07122457, + "balance_loss_mlp": 1.02929163, + "epoch": 0.06685504719533458, + "flos": 16217079995520.0, + "grad_norm": 3.053867075743763, + "language_loss": 0.75920463, + "learning_rate": 3.98581229911183e-06, + "loss": 0.78259814, + "num_input_tokens_seen": 11807285, + "step": 556, + "time_per_iteration": 2.7281718254089355 + }, + { + "auxiliary_loss_clip": 0.01280128, + "auxiliary_loss_mlp": 0.01053417, + "balance_loss_clip": 1.07257915, + "balance_loss_mlp": 1.04103768, + "epoch": 0.06697529008597367, + "flos": 22491535639680.0, + "grad_norm": 2.362392400735536, + "language_loss": 0.92127812, + "learning_rate": 3.985719528211341e-06, + "loss": 0.94461358, + "num_input_tokens_seen": 11826655, + "step": 557, + "time_per_iteration": 2.6738171577453613 + }, + { + "auxiliary_loss_clip": 0.01189373, + "auxiliary_loss_mlp": 0.01012142, + "balance_loss_clip": 1.04520226, + "balance_loss_mlp": 1.00825548, + "epoch": 0.06709553297661276, + "flos": 62688216936960.0, + "grad_norm": 0.8468176659684239, + "language_loss": 0.62997329, + "learning_rate": 3.985626456078777e-06, + "loss": 0.65198839, + "num_input_tokens_seen": 11891310, + "step": 558, + "time_per_iteration": 3.381007432937622 + }, + { + "auxiliary_loss_clip": 0.01296209, + "auxiliary_loss_mlp": 0.01049885, + "balance_loss_clip": 1.07173467, + "balance_loss_mlp": 1.03822041, + "epoch": 0.06721577586725185, + "flos": 11216590997760.0, + "grad_norm": 2.75720275365305, + "language_loss": 0.86134112, + "learning_rate": 3.985533082728259e-06, + "loss": 0.8848021, + "num_input_tokens_seen": 11906965, + "step": 559, + "time_per_iteration": 2.7352752685546875 + }, + { + "auxiliary_loss_clip": 0.01269105, + "auxiliary_loss_mlp": 0.01039056, + "balance_loss_clip": 1.07198668, + "balance_loss_mlp": 1.02675319, + "epoch": 0.06733601875789094, + "flos": 25922189664000.0, + "grad_norm": 2.5050042108486807, + "language_loss": 0.74692702, + "learning_rate": 3.985439408173951e-06, + "loss": 0.77000868, + "num_input_tokens_seen": 11927190, + "step": 560, + "time_per_iteration": 2.704852342605591 + }, + { + "auxiliary_loss_clip": 0.01279844, + "auxiliary_loss_mlp": 0.01049884, + "balance_loss_clip": 1.0798316, + "balance_loss_mlp": 1.03807628, + "epoch": 0.06745626164853002, + "flos": 20813645577600.0, + "grad_norm": 3.828010274272392, + "language_loss": 0.70519847, + "learning_rate": 3.9853454324300634e-06, + "loss": 0.72849578, + "num_input_tokens_seen": 11946400, + "step": 561, + "time_per_iteration": 2.6167585849761963 + }, + { + "auxiliary_loss_clip": 0.01312917, + "auxiliary_loss_mlp": 0.01047969, + "balance_loss_clip": 1.0640502, + "balance_loss_mlp": 1.03581572, + "epoch": 0.06757650453916912, + "flos": 19829262378240.0, + "grad_norm": 2.137915083408154, + "language_loss": 0.78026378, + "learning_rate": 3.985251155510852e-06, + "loss": 0.80387259, + "num_input_tokens_seen": 11965430, + "step": 562, + "time_per_iteration": 3.031951427459717 + }, + { + "auxiliary_loss_clip": 0.01302922, + "auxiliary_loss_mlp": 0.01038452, + "balance_loss_clip": 1.06980062, + "balance_loss_mlp": 1.026227, + "epoch": 0.06769674742980822, + "flos": 25739224761600.0, + "grad_norm": 1.821127770672979, + "language_loss": 0.80337441, + "learning_rate": 3.98515657743062e-06, + "loss": 0.82678819, + "num_input_tokens_seen": 11984895, + "step": 563, + "time_per_iteration": 3.9482812881469727 + }, + { + "auxiliary_loss_clip": 0.01287434, + "auxiliary_loss_mlp": 0.01049091, + "balance_loss_clip": 1.071805, + "balance_loss_mlp": 1.03692603, + "epoch": 0.0678169903204473, + "flos": 13074788355840.0, + "grad_norm": 1.856383188897722, + "language_loss": 0.77853191, + "learning_rate": 3.985061698203711e-06, + "loss": 0.80189717, + "num_input_tokens_seen": 12002010, + "step": 564, + "time_per_iteration": 2.7179415225982666 + }, + { + "auxiliary_loss_clip": 0.01167513, + "auxiliary_loss_mlp": 0.01009886, + "balance_loss_clip": 1.04666972, + "balance_loss_mlp": 1.00521338, + "epoch": 0.0679372332110864, + "flos": 70865830788480.0, + "grad_norm": 0.8860481837258503, + "language_loss": 0.6380055, + "learning_rate": 3.984966517844523e-06, + "loss": 0.65977955, + "num_input_tokens_seen": 12057255, + "step": 565, + "time_per_iteration": 4.2034666538238525 + }, + { + "auxiliary_loss_clip": 0.01275101, + "auxiliary_loss_mlp": 0.01041976, + "balance_loss_clip": 1.07587528, + "balance_loss_mlp": 1.02951217, + "epoch": 0.06805747610172548, + "flos": 28256418990720.0, + "grad_norm": 2.8490035911287612, + "language_loss": 0.80353922, + "learning_rate": 3.984871036367492e-06, + "loss": 0.82670999, + "num_input_tokens_seen": 12077280, + "step": 566, + "time_per_iteration": 3.633204221725464 + }, + { + "auxiliary_loss_clip": 0.01280117, + "auxiliary_loss_mlp": 0.01382633, + "balance_loss_clip": 1.07502675, + "balance_loss_mlp": 1.00034797, + "epoch": 0.06817771899236458, + "flos": 20120533764480.0, + "grad_norm": 2.1119084546111866, + "language_loss": 0.83228111, + "learning_rate": 3.984775253787102e-06, + "loss": 0.85890865, + "num_input_tokens_seen": 12095570, + "step": 567, + "time_per_iteration": 2.6764540672302246 + }, + { + "auxiliary_loss_clip": 0.0128365, + "auxiliary_loss_mlp": 0.0104308, + "balance_loss_clip": 1.07306373, + "balance_loss_mlp": 1.03128982, + "epoch": 0.06829796188300366, + "flos": 17930629284480.0, + "grad_norm": 3.143604878973736, + "language_loss": 0.8805455, + "learning_rate": 3.984679170117885e-06, + "loss": 0.90381283, + "num_input_tokens_seen": 12111775, + "step": 568, + "time_per_iteration": 2.691500663757324 + }, + { + "auxiliary_loss_clip": 0.01274398, + "auxiliary_loss_mlp": 0.01046393, + "balance_loss_clip": 1.07157946, + "balance_loss_mlp": 1.03472829, + "epoch": 0.06841820477364276, + "flos": 14501627285760.0, + "grad_norm": 2.485745949865142, + "language_loss": 0.7845788, + "learning_rate": 3.984582785374415e-06, + "loss": 0.8077867, + "num_input_tokens_seen": 12129215, + "step": 569, + "time_per_iteration": 2.64306640625 + }, + { + "auxiliary_loss_clip": 0.01286701, + "auxiliary_loss_mlp": 0.01382373, + "balance_loss_clip": 1.07326126, + "balance_loss_mlp": 1.00023401, + "epoch": 0.06853844766428185, + "flos": 21938474954880.0, + "grad_norm": 2.0365458725572783, + "language_loss": 0.80619144, + "learning_rate": 3.9844860995713155e-06, + "loss": 0.83288223, + "num_input_tokens_seen": 12148755, + "step": 570, + "time_per_iteration": 2.7964928150177 + }, + { + "auxiliary_loss_clip": 0.01282536, + "auxiliary_loss_mlp": 0.01040966, + "balance_loss_clip": 1.07951021, + "balance_loss_mlp": 1.02932549, + "epoch": 0.06865869055492094, + "flos": 16800628348800.0, + "grad_norm": 2.73138357094909, + "language_loss": 0.82671231, + "learning_rate": 3.9843891127232524e-06, + "loss": 0.84994733, + "num_input_tokens_seen": 12166290, + "step": 571, + "time_per_iteration": 2.710481882095337 + }, + { + "auxiliary_loss_clip": 0.01289893, + "auxiliary_loss_mlp": 0.01040086, + "balance_loss_clip": 1.06526756, + "balance_loss_mlp": 1.02852297, + "epoch": 0.06877893344556003, + "flos": 19937281553280.0, + "grad_norm": 2.5566616933876167, + "language_loss": 0.66875398, + "learning_rate": 3.984291824844938e-06, + "loss": 0.69205374, + "num_input_tokens_seen": 12181385, + "step": 572, + "time_per_iteration": 2.756087064743042 + }, + { + "auxiliary_loss_clip": 0.01270421, + "auxiliary_loss_mlp": 0.01038602, + "balance_loss_clip": 1.07426536, + "balance_loss_mlp": 1.02616239, + "epoch": 0.06889917633619912, + "flos": 23039388852480.0, + "grad_norm": 2.522951282072005, + "language_loss": 0.85111398, + "learning_rate": 3.984194235951132e-06, + "loss": 0.87420422, + "num_input_tokens_seen": 12197530, + "step": 573, + "time_per_iteration": 2.6300299167633057 + }, + { + "auxiliary_loss_clip": 0.01275579, + "auxiliary_loss_mlp": 0.01046814, + "balance_loss_clip": 1.07878733, + "balance_loss_mlp": 1.03467846, + "epoch": 0.06901941922683821, + "flos": 20960556203520.0, + "grad_norm": 2.315127896593087, + "language_loss": 0.84738898, + "learning_rate": 3.9840963460566375e-06, + "loss": 0.87061286, + "num_input_tokens_seen": 12216310, + "step": 574, + "time_per_iteration": 2.6369147300720215 + }, + { + "auxiliary_loss_clip": 0.01287859, + "auxiliary_loss_mlp": 0.01041678, + "balance_loss_clip": 1.06147969, + "balance_loss_mlp": 1.0302099, + "epoch": 0.06913966211747731, + "flos": 24821850384000.0, + "grad_norm": 1.6569431197939037, + "language_loss": 0.89386332, + "learning_rate": 3.983998155176305e-06, + "loss": 0.91715872, + "num_input_tokens_seen": 12236670, + "step": 575, + "time_per_iteration": 2.8763928413391113 + }, + { + "auxiliary_loss_clip": 0.01156706, + "auxiliary_loss_mlp": 0.01003699, + "balance_loss_clip": 1.03893447, + "balance_loss_mlp": 0.99959832, + "epoch": 0.06925990500811639, + "flos": 58367446957440.0, + "grad_norm": 0.8169886826509524, + "language_loss": 0.56994188, + "learning_rate": 3.9838996633250305e-06, + "loss": 0.59154594, + "num_input_tokens_seen": 12297185, + "step": 576, + "time_per_iteration": 3.1662392616271973 + }, + { + "auxiliary_loss_clip": 0.01278557, + "auxiliary_loss_mlp": 0.01056763, + "balance_loss_clip": 1.07149863, + "balance_loss_mlp": 1.04478288, + "epoch": 0.06938014789875549, + "flos": 12749940731520.0, + "grad_norm": 2.5602510473688196, + "language_loss": 0.88220942, + "learning_rate": 3.983800870517753e-06, + "loss": 0.90556258, + "num_input_tokens_seen": 12313975, + "step": 577, + "time_per_iteration": 2.7065024375915527 + }, + { + "auxiliary_loss_clip": 0.01278924, + "auxiliary_loss_mlp": 0.01040713, + "balance_loss_clip": 1.07966828, + "balance_loss_mlp": 1.02941191, + "epoch": 0.06950039078939457, + "flos": 22820226019200.0, + "grad_norm": 6.8502993313011125, + "language_loss": 0.78069526, + "learning_rate": 3.983701776769463e-06, + "loss": 0.8038916, + "num_input_tokens_seen": 12331385, + "step": 578, + "time_per_iteration": 2.668621063232422 + }, + { + "auxiliary_loss_clip": 0.01263154, + "auxiliary_loss_mlp": 0.01043721, + "balance_loss_clip": 1.07102251, + "balance_loss_mlp": 1.03244948, + "epoch": 0.06962063368003367, + "flos": 21941348042880.0, + "grad_norm": 2.226407994192909, + "language_loss": 0.85883927, + "learning_rate": 3.9836023820951885e-06, + "loss": 0.88190806, + "num_input_tokens_seen": 12350600, + "step": 579, + "time_per_iteration": 2.7145354747772217 + }, + { + "auxiliary_loss_clip": 0.01285362, + "auxiliary_loss_mlp": 0.01046164, + "balance_loss_clip": 1.06704652, + "balance_loss_mlp": 1.03511333, + "epoch": 0.06974087657067275, + "flos": 20706021452160.0, + "grad_norm": 2.284338404423597, + "language_loss": 0.68467414, + "learning_rate": 3.983502686510011e-06, + "loss": 0.70798945, + "num_input_tokens_seen": 12371430, + "step": 580, + "time_per_iteration": 2.776210069656372 + }, + { + "auxiliary_loss_clip": 0.01278533, + "auxiliary_loss_mlp": 0.01382356, + "balance_loss_clip": 1.07288718, + "balance_loss_mlp": 1.00011253, + "epoch": 0.06986111946131185, + "flos": 22638230784000.0, + "grad_norm": 2.0271946452599967, + "language_loss": 0.73479283, + "learning_rate": 3.9834026900290525e-06, + "loss": 0.76140165, + "num_input_tokens_seen": 12390825, + "step": 581, + "time_per_iteration": 2.752289056777954 + }, + { + "auxiliary_loss_clip": 0.01272359, + "auxiliary_loss_mlp": 0.01040916, + "balance_loss_clip": 1.07663274, + "balance_loss_mlp": 1.0288285, + "epoch": 0.06998136235195095, + "flos": 26943453152640.0, + "grad_norm": 2.3694038350542574, + "language_loss": 1.00024498, + "learning_rate": 3.983302392667482e-06, + "loss": 1.02337766, + "num_input_tokens_seen": 12411670, + "step": 582, + "time_per_iteration": 2.678539276123047 + }, + { + "auxiliary_loss_clip": 0.01280627, + "auxiliary_loss_mlp": 0.01041833, + "balance_loss_clip": 1.07795107, + "balance_loss_mlp": 1.03104472, + "epoch": 0.07010160524259003, + "flos": 22492505306880.0, + "grad_norm": 1.8063031060097559, + "language_loss": 0.93485796, + "learning_rate": 3.983201794440517e-06, + "loss": 0.95808256, + "num_input_tokens_seen": 12431245, + "step": 583, + "time_per_iteration": 2.682666540145874 + }, + { + "auxiliary_loss_clip": 0.0127135, + "auxiliary_loss_mlp": 0.01048408, + "balance_loss_clip": 1.06971133, + "balance_loss_mlp": 1.03629029, + "epoch": 0.07022184813322913, + "flos": 18332541538560.0, + "grad_norm": 1.8916423775025737, + "language_loss": 0.67567295, + "learning_rate": 3.9831008953634165e-06, + "loss": 0.69887054, + "num_input_tokens_seen": 12450535, + "step": 584, + "time_per_iteration": 2.6807334423065186 + }, + { + "auxiliary_loss_clip": 0.01278454, + "auxiliary_loss_mlp": 0.01049445, + "balance_loss_clip": 1.06296563, + "balance_loss_mlp": 1.03760767, + "epoch": 0.07034209102386821, + "flos": 24675550289280.0, + "grad_norm": 2.0694502988109096, + "language_loss": 0.81150341, + "learning_rate": 3.9829996954514864e-06, + "loss": 0.83478242, + "num_input_tokens_seen": 12469675, + "step": 585, + "time_per_iteration": 2.7705581188201904 + }, + { + "auxiliary_loss_clip": 0.01266001, + "auxiliary_loss_mlp": 0.0104288, + "balance_loss_clip": 1.07191563, + "balance_loss_mlp": 1.03174639, + "epoch": 0.0704623339145073, + "flos": 25995878415360.0, + "grad_norm": 1.912596429167658, + "language_loss": 0.84406388, + "learning_rate": 3.982898194720079e-06, + "loss": 0.86715269, + "num_input_tokens_seen": 12490405, + "step": 586, + "time_per_iteration": 2.7415971755981445 + }, + { + "auxiliary_loss_clip": 0.01279604, + "auxiliary_loss_mlp": 0.01382583, + "balance_loss_clip": 1.07383084, + "balance_loss_mlp": 1.00011599, + "epoch": 0.0705825768051464, + "flos": 25338318088320.0, + "grad_norm": 3.3571860142235344, + "language_loss": 0.8259896, + "learning_rate": 3.982796393184592e-06, + "loss": 0.85261142, + "num_input_tokens_seen": 12509485, + "step": 587, + "time_per_iteration": 2.727501153945923 + }, + { + "auxiliary_loss_clip": 0.0116038, + "auxiliary_loss_mlp": 0.01004915, + "balance_loss_clip": 1.03467822, + "balance_loss_mlp": 1.00121999, + "epoch": 0.07070281969578548, + "flos": 66047552507520.0, + "grad_norm": 0.7903037229355261, + "language_loss": 0.62692475, + "learning_rate": 3.98269429086047e-06, + "loss": 0.64857775, + "num_input_tokens_seen": 12567325, + "step": 588, + "time_per_iteration": 3.1211044788360596 + }, + { + "auxiliary_loss_clip": 0.01270347, + "auxiliary_loss_mlp": 0.01049333, + "balance_loss_clip": 1.07199609, + "balance_loss_mlp": 1.03666127, + "epoch": 0.07082306258642458, + "flos": 23653568528640.0, + "grad_norm": 10.804979468353348, + "language_loss": 0.8654092, + "learning_rate": 3.982591887763199e-06, + "loss": 0.88860595, + "num_input_tokens_seen": 12584785, + "step": 589, + "time_per_iteration": 3.6056275367736816 + }, + { + "auxiliary_loss_clip": 0.01263302, + "auxiliary_loss_mlp": 0.01041056, + "balance_loss_clip": 1.05809045, + "balance_loss_mlp": 1.02964771, + "epoch": 0.07094330547706366, + "flos": 13880049408000.0, + "grad_norm": 2.3366268724435213, + "language_loss": 0.81934375, + "learning_rate": 3.982489183908316e-06, + "loss": 0.84238732, + "num_input_tokens_seen": 12601205, + "step": 590, + "time_per_iteration": 2.6892590522766113 + }, + { + "auxiliary_loss_clip": 0.01271938, + "auxiliary_loss_mlp": 0.01046057, + "balance_loss_clip": 1.05695105, + "balance_loss_mlp": 1.03532195, + "epoch": 0.07106354836770276, + "flos": 24645098534400.0, + "grad_norm": 2.057740357381453, + "language_loss": 0.84830159, + "learning_rate": 3.982386179311399e-06, + "loss": 0.87148154, + "num_input_tokens_seen": 12621725, + "step": 591, + "time_per_iteration": 3.720888614654541 + }, + { + "auxiliary_loss_clip": 0.01282399, + "auxiliary_loss_mlp": 0.01058524, + "balance_loss_clip": 1.076056, + "balance_loss_mlp": 1.04643071, + "epoch": 0.07118379125834184, + "flos": 16217223649920.0, + "grad_norm": 2.4607232044651983, + "language_loss": 0.87563789, + "learning_rate": 3.982282873988075e-06, + "loss": 0.89904714, + "num_input_tokens_seen": 12639600, + "step": 592, + "time_per_iteration": 3.6971325874328613 + }, + { + "auxiliary_loss_clip": 0.01287321, + "auxiliary_loss_mlp": 0.01047856, + "balance_loss_clip": 1.07515442, + "balance_loss_mlp": 1.03631687, + "epoch": 0.07130403414898094, + "flos": 19719986227200.0, + "grad_norm": 1.673807303459961, + "language_loss": 0.86970299, + "learning_rate": 3.982179267954016e-06, + "loss": 0.89305472, + "num_input_tokens_seen": 12660030, + "step": 593, + "time_per_iteration": 2.7033023834228516 + }, + { + "auxiliary_loss_clip": 0.01270753, + "auxiliary_loss_mlp": 0.0105079, + "balance_loss_clip": 1.07483339, + "balance_loss_mlp": 1.0390842, + "epoch": 0.07142427703962004, + "flos": 21871933009920.0, + "grad_norm": 3.093490660143408, + "language_loss": 0.95680046, + "learning_rate": 3.982075361224937e-06, + "loss": 0.98001587, + "num_input_tokens_seen": 12678395, + "step": 594, + "time_per_iteration": 2.5950889587402344 + }, + { + "auxiliary_loss_clip": 0.01275767, + "auxiliary_loss_mlp": 0.01382179, + "balance_loss_clip": 1.07647729, + "balance_loss_mlp": 1.00009465, + "epoch": 0.07154451993025912, + "flos": 18296595002880.0, + "grad_norm": 5.965271780035816, + "language_loss": 0.88235927, + "learning_rate": 3.981971153816602e-06, + "loss": 0.90893877, + "num_input_tokens_seen": 12696000, + "step": 595, + "time_per_iteration": 2.5865914821624756 + }, + { + "auxiliary_loss_clip": 0.01270942, + "auxiliary_loss_mlp": 0.01044269, + "balance_loss_clip": 1.08018875, + "balance_loss_mlp": 1.03318822, + "epoch": 0.07166476282089822, + "flos": 22160690444160.0, + "grad_norm": 1.7259993051016322, + "language_loss": 0.96357381, + "learning_rate": 3.981866645744819e-06, + "loss": 0.98672593, + "num_input_tokens_seen": 12716715, + "step": 596, + "time_per_iteration": 2.6409826278686523 + }, + { + "auxiliary_loss_clip": 0.01273578, + "auxiliary_loss_mlp": 0.01381994, + "balance_loss_clip": 1.07731462, + "balance_loss_mlp": 1.0000242, + "epoch": 0.0717850057115373, + "flos": 14136343925760.0, + "grad_norm": 3.2738853736723654, + "language_loss": 0.8138938, + "learning_rate": 3.9817618370254416e-06, + "loss": 0.84044945, + "num_input_tokens_seen": 12733370, + "step": 597, + "time_per_iteration": 2.5588605403900146 + }, + { + "auxiliary_loss_clip": 0.01268994, + "auxiliary_loss_mlp": 0.0104216, + "balance_loss_clip": 1.07534385, + "balance_loss_mlp": 1.02994728, + "epoch": 0.0719052486021764, + "flos": 30917794412160.0, + "grad_norm": 5.105680174536041, + "language_loss": 0.87480617, + "learning_rate": 3.9816567276743684e-06, + "loss": 0.89791769, + "num_input_tokens_seen": 12753235, + "step": 598, + "time_per_iteration": 2.663069009780884 + }, + { + "auxiliary_loss_clip": 0.0127697, + "auxiliary_loss_mlp": 0.0104597, + "balance_loss_clip": 1.06966436, + "balance_loss_mlp": 1.03406692, + "epoch": 0.0720254914928155, + "flos": 21287019939840.0, + "grad_norm": 2.6976270468364802, + "language_loss": 0.77400863, + "learning_rate": 3.9815513177075466e-06, + "loss": 0.79723799, + "num_input_tokens_seen": 12772020, + "step": 599, + "time_per_iteration": 2.708580255508423 + }, + { + "auxiliary_loss_clip": 0.01267052, + "auxiliary_loss_mlp": 0.01040292, + "balance_loss_clip": 1.07376361, + "balance_loss_mlp": 1.02864563, + "epoch": 0.07214573438345458, + "flos": 27819170732160.0, + "grad_norm": 1.5453128590492546, + "language_loss": 0.70238805, + "learning_rate": 3.9814456071409646e-06, + "loss": 0.72546148, + "num_input_tokens_seen": 12792555, + "step": 600, + "time_per_iteration": 2.682587146759033 + }, + { + "auxiliary_loss_clip": 0.0130094, + "auxiliary_loss_mlp": 0.01039342, + "balance_loss_clip": 1.06661141, + "balance_loss_mlp": 1.02714646, + "epoch": 0.07226597727409367, + "flos": 25483576688640.0, + "grad_norm": 2.967351886414557, + "language_loss": 0.85180938, + "learning_rate": 3.981339595990659e-06, + "loss": 0.87521219, + "num_input_tokens_seen": 12811085, + "step": 601, + "time_per_iteration": 2.7379956245422363 + }, + { + "auxiliary_loss_clip": 0.01275395, + "auxiliary_loss_mlp": 0.01045499, + "balance_loss_clip": 1.07437038, + "balance_loss_mlp": 1.03369761, + "epoch": 0.07238622016473276, + "flos": 23513840622720.0, + "grad_norm": 1.9657797045235919, + "language_loss": 0.81329596, + "learning_rate": 3.981233284272713e-06, + "loss": 0.83650488, + "num_input_tokens_seen": 12830830, + "step": 602, + "time_per_iteration": 2.6474058628082275 + }, + { + "auxiliary_loss_clip": 0.01287815, + "auxiliary_loss_mlp": 0.01041066, + "balance_loss_clip": 1.06890404, + "balance_loss_mlp": 1.02931798, + "epoch": 0.07250646305537185, + "flos": 25453519983360.0, + "grad_norm": 1.5610707280014529, + "language_loss": 0.89995825, + "learning_rate": 3.981126672003253e-06, + "loss": 0.92324704, + "num_input_tokens_seen": 12853505, + "step": 603, + "time_per_iteration": 2.8373517990112305 + }, + { + "auxiliary_loss_clip": 0.0128692, + "auxiliary_loss_mlp": 0.01043073, + "balance_loss_clip": 1.0685761, + "balance_loss_mlp": 1.03225434, + "epoch": 0.07262670594601094, + "flos": 27155038216320.0, + "grad_norm": 2.9259070722183766, + "language_loss": 0.78223801, + "learning_rate": 3.981019759198451e-06, + "loss": 0.80553794, + "num_input_tokens_seen": 12872455, + "step": 604, + "time_per_iteration": 2.691317081451416 + }, + { + "auxiliary_loss_clip": 0.0128794, + "auxiliary_loss_mlp": 0.01038514, + "balance_loss_clip": 1.07011724, + "balance_loss_mlp": 1.02769589, + "epoch": 0.07274694883665003, + "flos": 26651607148800.0, + "grad_norm": 2.2541077635767115, + "language_loss": 0.84452462, + "learning_rate": 3.980912545874528e-06, + "loss": 0.86778915, + "num_input_tokens_seen": 12892620, + "step": 605, + "time_per_iteration": 2.740063190460205 + }, + { + "auxiliary_loss_clip": 0.01265824, + "auxiliary_loss_mlp": 0.01382561, + "balance_loss_clip": 1.06994247, + "balance_loss_mlp": 1.00002789, + "epoch": 0.07286719172728913, + "flos": 29862344154240.0, + "grad_norm": 2.2570199409527536, + "language_loss": 0.85767806, + "learning_rate": 3.980805032047746e-06, + "loss": 0.88416195, + "num_input_tokens_seen": 12914090, + "step": 606, + "time_per_iteration": 2.6993408203125 + }, + { + "auxiliary_loss_clip": 0.01279435, + "auxiliary_loss_mlp": 0.01041496, + "balance_loss_clip": 1.07005203, + "balance_loss_mlp": 1.02906871, + "epoch": 0.07298743461792821, + "flos": 17382057799680.0, + "grad_norm": 2.0041933913702934, + "language_loss": 0.80938429, + "learning_rate": 3.980697217734415e-06, + "loss": 0.83259368, + "num_input_tokens_seen": 12931830, + "step": 607, + "time_per_iteration": 2.648196220397949 + }, + { + "auxiliary_loss_clip": 0.01300299, + "auxiliary_loss_mlp": 0.01382348, + "balance_loss_clip": 1.06516576, + "balance_loss_mlp": 0.99999142, + "epoch": 0.07310767750856731, + "flos": 19498201701120.0, + "grad_norm": 2.0299942279842815, + "language_loss": 0.91442031, + "learning_rate": 3.980589102950891e-06, + "loss": 0.94124681, + "num_input_tokens_seen": 12949995, + "step": 608, + "time_per_iteration": 2.7212164402008057 + }, + { + "auxiliary_loss_clip": 0.01278535, + "auxiliary_loss_mlp": 0.01047515, + "balance_loss_clip": 1.07165647, + "balance_loss_mlp": 1.03533208, + "epoch": 0.07322792039920639, + "flos": 29168693637120.0, + "grad_norm": 2.60076873813455, + "language_loss": 0.75500625, + "learning_rate": 3.9804806877135755e-06, + "loss": 0.77826679, + "num_input_tokens_seen": 12968040, + "step": 609, + "time_per_iteration": 2.706712245941162 + }, + { + "auxiliary_loss_clip": 0.01273608, + "auxiliary_loss_mlp": 0.01382478, + "balance_loss_clip": 1.06884277, + "balance_loss_mlp": 1.00000513, + "epoch": 0.07334816328984549, + "flos": 23477822259840.0, + "grad_norm": 2.0829938557512517, + "language_loss": 0.86132914, + "learning_rate": 3.980371972038915e-06, + "loss": 0.88788998, + "num_input_tokens_seen": 12988530, + "step": 610, + "time_per_iteration": 2.6564104557037354 + }, + { + "auxiliary_loss_clip": 0.01275676, + "auxiliary_loss_mlp": 0.01049502, + "balance_loss_clip": 1.08034408, + "balance_loss_mlp": 1.03753328, + "epoch": 0.07346840618048459, + "flos": 22962467877120.0, + "grad_norm": 1.7782361571803607, + "language_loss": 0.8430863, + "learning_rate": 3.980262955943399e-06, + "loss": 0.86633807, + "num_input_tokens_seen": 13008195, + "step": 611, + "time_per_iteration": 2.6503419876098633 + }, + { + "auxiliary_loss_clip": 0.01277428, + "auxiliary_loss_mlp": 0.01049522, + "balance_loss_clip": 1.07244515, + "balance_loss_mlp": 1.0380187, + "epoch": 0.07358864907112367, + "flos": 17673903803520.0, + "grad_norm": 2.4794362962543333, + "language_loss": 0.86553729, + "learning_rate": 3.980153639443569e-06, + "loss": 0.88880682, + "num_input_tokens_seen": 13024180, + "step": 612, + "time_per_iteration": 2.6598944664001465 + }, + { + "auxiliary_loss_clip": 0.01289807, + "auxiliary_loss_mlp": 0.0104593, + "balance_loss_clip": 1.07127225, + "balance_loss_mlp": 1.03344882, + "epoch": 0.07370889196176277, + "flos": 24097029840000.0, + "grad_norm": 1.880119338697321, + "language_loss": 0.79659724, + "learning_rate": 3.980044022556005e-06, + "loss": 0.81995463, + "num_input_tokens_seen": 13043865, + "step": 613, + "time_per_iteration": 2.713930130004883 + }, + { + "auxiliary_loss_clip": 0.01274142, + "auxiliary_loss_mlp": 0.01052464, + "balance_loss_clip": 1.07368517, + "balance_loss_mlp": 1.03988767, + "epoch": 0.07382913485240185, + "flos": 25885919905920.0, + "grad_norm": 2.1103932452111107, + "language_loss": 0.72666627, + "learning_rate": 3.9799341052973375e-06, + "loss": 0.74993235, + "num_input_tokens_seen": 13063700, + "step": 614, + "time_per_iteration": 2.7005929946899414 + }, + { + "auxiliary_loss_clip": 0.01278798, + "auxiliary_loss_mlp": 0.01050505, + "balance_loss_clip": 1.07191253, + "balance_loss_mlp": 1.03770757, + "epoch": 0.07394937774304094, + "flos": 16873850223360.0, + "grad_norm": 4.275070566655856, + "language_loss": 0.75500548, + "learning_rate": 3.979823887684241e-06, + "loss": 0.7782985, + "num_input_tokens_seen": 13082640, + "step": 615, + "time_per_iteration": 3.5518898963928223 + }, + { + "auxiliary_loss_clip": 0.01266968, + "auxiliary_loss_mlp": 0.01051246, + "balance_loss_clip": 1.07479787, + "balance_loss_mlp": 1.03961742, + "epoch": 0.07406962063368003, + "flos": 20703471586560.0, + "grad_norm": 2.300898962563342, + "language_loss": 0.84657091, + "learning_rate": 3.979713369733434e-06, + "loss": 0.86975306, + "num_input_tokens_seen": 13100505, + "step": 616, + "time_per_iteration": 2.6211323738098145 + }, + { + "auxiliary_loss_clip": 0.01267422, + "auxiliary_loss_mlp": 0.01046221, + "balance_loss_clip": 1.0716176, + "balance_loss_mlp": 1.03394842, + "epoch": 0.07418986352431912, + "flos": 21430985650560.0, + "grad_norm": 2.0111863777773897, + "language_loss": 0.85105956, + "learning_rate": 3.979602551461683e-06, + "loss": 0.87419599, + "num_input_tokens_seen": 13121285, + "step": 617, + "time_per_iteration": 3.613299608230591 + }, + { + "auxiliary_loss_clip": 0.01282502, + "auxiliary_loss_mlp": 0.01059579, + "balance_loss_clip": 1.07457995, + "balance_loss_mlp": 1.04750323, + "epoch": 0.07431010641495822, + "flos": 12021133777920.0, + "grad_norm": 2.3269794895763374, + "language_loss": 0.91705513, + "learning_rate": 3.979491432885799e-06, + "loss": 0.94047594, + "num_input_tokens_seen": 13137550, + "step": 618, + "time_per_iteration": 4.488139867782593 + }, + { + "auxiliary_loss_clip": 0.01269072, + "auxiliary_loss_mlp": 0.01381441, + "balance_loss_clip": 1.06668162, + "balance_loss_mlp": 0.99998909, + "epoch": 0.0744303493055973, + "flos": 20957575374720.0, + "grad_norm": 2.1873255749683773, + "language_loss": 0.83218956, + "learning_rate": 3.97938001402264e-06, + "loss": 0.85869467, + "num_input_tokens_seen": 13156675, + "step": 619, + "time_per_iteration": 2.7869553565979004 + }, + { + "auxiliary_loss_clip": 0.01283552, + "auxiliary_loss_mlp": 0.01044938, + "balance_loss_clip": 1.06845975, + "balance_loss_mlp": 1.03318381, + "epoch": 0.0745505921962364, + "flos": 16253134272000.0, + "grad_norm": 3.147421101092579, + "language_loss": 0.7996875, + "learning_rate": 3.979268294889105e-06, + "loss": 0.82297242, + "num_input_tokens_seen": 13172225, + "step": 620, + "time_per_iteration": 2.778351306915283 + }, + { + "auxiliary_loss_clip": 0.01266606, + "auxiliary_loss_mlp": 0.01052465, + "balance_loss_clip": 1.07457149, + "balance_loss_mlp": 1.04108, + "epoch": 0.07467083508687548, + "flos": 50944635550080.0, + "grad_norm": 1.8681368195729895, + "language_loss": 0.73798275, + "learning_rate": 3.979156275502143e-06, + "loss": 0.76117349, + "num_input_tokens_seen": 13195885, + "step": 621, + "time_per_iteration": 2.8700218200683594 + }, + { + "auxiliary_loss_clip": 0.01290306, + "auxiliary_loss_mlp": 0.01043038, + "balance_loss_clip": 1.07112992, + "balance_loss_mlp": 1.03041339, + "epoch": 0.07479107797751458, + "flos": 17529686697600.0, + "grad_norm": 2.6781637149212636, + "language_loss": 0.91576326, + "learning_rate": 3.979043955878749e-06, + "loss": 0.93909675, + "num_input_tokens_seen": 13213730, + "step": 622, + "time_per_iteration": 2.7461936473846436 + }, + { + "auxiliary_loss_clip": 0.01278323, + "auxiliary_loss_mlp": 0.01049446, + "balance_loss_clip": 1.07192016, + "balance_loss_mlp": 1.03816855, + "epoch": 0.07491132086815366, + "flos": 23473943591040.0, + "grad_norm": 1.9577984257549879, + "language_loss": 0.83325255, + "learning_rate": 3.978931336035959e-06, + "loss": 0.85653025, + "num_input_tokens_seen": 13232540, + "step": 623, + "time_per_iteration": 2.6859312057495117 + }, + { + "auxiliary_loss_clip": 0.01274522, + "auxiliary_loss_mlp": 0.01046967, + "balance_loss_clip": 1.0748347, + "balance_loss_mlp": 1.03543925, + "epoch": 0.07503156375879276, + "flos": 20157557708160.0, + "grad_norm": 2.34779797885042, + "language_loss": 0.82529628, + "learning_rate": 3.9788184159908595e-06, + "loss": 0.84851116, + "num_input_tokens_seen": 13249670, + "step": 624, + "time_per_iteration": 2.6525650024414062 + }, + { + "auxiliary_loss_clip": 0.01268055, + "auxiliary_loss_mlp": 0.01055148, + "balance_loss_clip": 1.06845176, + "balance_loss_mlp": 1.04284608, + "epoch": 0.07515180664943186, + "flos": 15115519653120.0, + "grad_norm": 4.629777212615311, + "language_loss": 0.8283481, + "learning_rate": 3.97870519576058e-06, + "loss": 0.85158014, + "num_input_tokens_seen": 13266095, + "step": 625, + "time_per_iteration": 2.7016615867614746 + }, + { + "auxiliary_loss_clip": 0.0128231, + "auxiliary_loss_mlp": 0.01382, + "balance_loss_clip": 1.06666183, + "balance_loss_mlp": 0.99999136, + "epoch": 0.07527204954007094, + "flos": 21287702298240.0, + "grad_norm": 2.171257297017563, + "language_loss": 0.80973685, + "learning_rate": 3.978591675362295e-06, + "loss": 0.83638, + "num_input_tokens_seen": 13284810, + "step": 626, + "time_per_iteration": 2.7677149772644043 + }, + { + "auxiliary_loss_clip": 0.012959, + "auxiliary_loss_mlp": 0.01046694, + "balance_loss_clip": 1.07049775, + "balance_loss_mlp": 1.03426075, + "epoch": 0.07539229243071004, + "flos": 21324187537920.0, + "grad_norm": 2.00645189390948, + "language_loss": 0.87645102, + "learning_rate": 3.978477854813226e-06, + "loss": 0.89987695, + "num_input_tokens_seen": 13304150, + "step": 627, + "time_per_iteration": 2.7685158252716064 + }, + { + "auxiliary_loss_clip": 0.01273072, + "auxiliary_loss_mlp": 0.01039006, + "balance_loss_clip": 1.0713824, + "balance_loss_mlp": 1.02713895, + "epoch": 0.07551253532134912, + "flos": 13042540920960.0, + "grad_norm": 2.0268160760006615, + "language_loss": 0.82313019, + "learning_rate": 3.97836373413064e-06, + "loss": 0.84625101, + "num_input_tokens_seen": 13322205, + "step": 628, + "time_per_iteration": 2.6491289138793945 + }, + { + "auxiliary_loss_clip": 0.01265411, + "auxiliary_loss_mlp": 0.0105276, + "balance_loss_clip": 1.07360554, + "balance_loss_mlp": 1.04084492, + "epoch": 0.07563277821198822, + "flos": 19208761908480.0, + "grad_norm": 2.0833428709949846, + "language_loss": 0.74543494, + "learning_rate": 3.978249313331848e-06, + "loss": 0.76861668, + "num_input_tokens_seen": 13340435, + "step": 629, + "time_per_iteration": 2.6300694942474365 + }, + { + "auxiliary_loss_clip": 0.01276716, + "auxiliary_loss_mlp": 0.01381978, + "balance_loss_clip": 1.07009518, + "balance_loss_mlp": 1.00000608, + "epoch": 0.07575302110262731, + "flos": 19537200892800.0, + "grad_norm": 3.6223437574492605, + "language_loss": 0.62019598, + "learning_rate": 3.978134592434208e-06, + "loss": 0.64678288, + "num_input_tokens_seen": 13358185, + "step": 630, + "time_per_iteration": 2.6927921772003174 + }, + { + "auxiliary_loss_clip": 0.01183037, + "auxiliary_loss_mlp": 0.01007425, + "balance_loss_clip": 1.03521466, + "balance_loss_mlp": 1.00413501, + "epoch": 0.0758732639932664, + "flos": 67961808017280.0, + "grad_norm": 1.0463284214030657, + "language_loss": 0.59374082, + "learning_rate": 3.978019571455123e-06, + "loss": 0.61564547, + "num_input_tokens_seen": 13410130, + "step": 631, + "time_per_iteration": 3.3279755115509033 + }, + { + "auxiliary_loss_clip": 0.01264689, + "auxiliary_loss_mlp": 0.01051377, + "balance_loss_clip": 1.07537389, + "balance_loss_mlp": 1.0396291, + "epoch": 0.07599350688390549, + "flos": 18989204025600.0, + "grad_norm": 2.319048199786777, + "language_loss": 0.83866858, + "learning_rate": 3.977904250412042e-06, + "loss": 0.86182922, + "num_input_tokens_seen": 13429085, + "step": 632, + "time_per_iteration": 2.6206843852996826 + }, + { + "auxiliary_loss_clip": 0.01285866, + "auxiliary_loss_mlp": 0.01050453, + "balance_loss_clip": 1.07229745, + "balance_loss_mlp": 1.0393312, + "epoch": 0.07611374977454458, + "flos": 21069006341760.0, + "grad_norm": 2.6395584823958154, + "language_loss": 0.85683239, + "learning_rate": 3.97778862932246e-06, + "loss": 0.88019562, + "num_input_tokens_seen": 13446250, + "step": 633, + "time_per_iteration": 2.7309200763702393 + }, + { + "auxiliary_loss_clip": 0.01288266, + "auxiliary_loss_mlp": 0.01054928, + "balance_loss_clip": 1.05151057, + "balance_loss_mlp": 1.04273844, + "epoch": 0.07623399266518367, + "flos": 18514536773760.0, + "grad_norm": 2.2970202240249864, + "language_loss": 0.93950868, + "learning_rate": 3.9776727082039144e-06, + "loss": 0.96294057, + "num_input_tokens_seen": 13463220, + "step": 634, + "time_per_iteration": 2.863375663757324 + }, + { + "auxiliary_loss_clip": 0.01148352, + "auxiliary_loss_mlp": 0.01003264, + "balance_loss_clip": 1.03583395, + "balance_loss_mlp": 1.00016499, + "epoch": 0.07635423555582276, + "flos": 44663036077440.0, + "grad_norm": 0.8170032272805334, + "language_loss": 0.55512238, + "learning_rate": 3.977556487073991e-06, + "loss": 0.57663858, + "num_input_tokens_seen": 13517775, + "step": 635, + "time_per_iteration": 3.1394424438476562 + }, + { + "auxiliary_loss_clip": 0.01266706, + "auxiliary_loss_mlp": 0.01041132, + "balance_loss_clip": 1.06460941, + "balance_loss_mlp": 1.03015304, + "epoch": 0.07647447844646185, + "flos": 21761148487680.0, + "grad_norm": 1.8254915439747654, + "language_loss": 0.81762564, + "learning_rate": 3.97743996595032e-06, + "loss": 0.84070402, + "num_input_tokens_seen": 13537815, + "step": 636, + "time_per_iteration": 2.7771925926208496 + }, + { + "auxiliary_loss_clip": 0.01267012, + "auxiliary_loss_mlp": 0.01041705, + "balance_loss_clip": 1.07454252, + "balance_loss_mlp": 1.0296824, + "epoch": 0.07659472133710095, + "flos": 23806799948160.0, + "grad_norm": 1.7013640301094972, + "language_loss": 0.81952727, + "learning_rate": 3.9773231448505804e-06, + "loss": 0.84261447, + "num_input_tokens_seen": 13559605, + "step": 637, + "time_per_iteration": 2.7193336486816406 + }, + { + "auxiliary_loss_clip": 0.01278681, + "auxiliary_loss_mlp": 0.01382072, + "balance_loss_clip": 1.07002366, + "balance_loss_mlp": 0.9999187, + "epoch": 0.07671496422774003, + "flos": 21469984842240.0, + "grad_norm": 2.2313165380196427, + "language_loss": 0.78130591, + "learning_rate": 3.977206023792491e-06, + "loss": 0.80791342, + "num_input_tokens_seen": 13579495, + "step": 638, + "time_per_iteration": 2.737410545349121 + }, + { + "auxiliary_loss_clip": 0.01270641, + "auxiliary_loss_mlp": 0.01040444, + "balance_loss_clip": 1.07369876, + "balance_loss_mlp": 1.02818978, + "epoch": 0.07683520711837913, + "flos": 16980971558400.0, + "grad_norm": 2.712164004895122, + "language_loss": 0.8098892, + "learning_rate": 3.97708860279382e-06, + "loss": 0.833, + "num_input_tokens_seen": 13597605, + "step": 639, + "time_per_iteration": 2.677635431289673 + }, + { + "auxiliary_loss_clip": 0.01279022, + "auxiliary_loss_mlp": 0.01053451, + "balance_loss_clip": 1.06602859, + "balance_loss_mlp": 1.04259706, + "epoch": 0.07695545000901821, + "flos": 23476744851840.0, + "grad_norm": 1.6451125290958468, + "language_loss": 0.77943277, + "learning_rate": 3.97697088187238e-06, + "loss": 0.8027575, + "num_input_tokens_seen": 13618120, + "step": 640, + "time_per_iteration": 2.811051368713379 + }, + { + "auxiliary_loss_clip": 0.01282006, + "auxiliary_loss_mlp": 0.01045491, + "balance_loss_clip": 1.07548833, + "balance_loss_mlp": 1.03399277, + "epoch": 0.07707569289965731, + "flos": 17634258167040.0, + "grad_norm": 1.9934725674370934, + "language_loss": 0.92257375, + "learning_rate": 3.976852861046029e-06, + "loss": 0.9458487, + "num_input_tokens_seen": 13634735, + "step": 641, + "time_per_iteration": 3.530000686645508 + }, + { + "auxiliary_loss_clip": 0.01287211, + "auxiliary_loss_mlp": 0.01045143, + "balance_loss_clip": 1.06509924, + "balance_loss_mlp": 1.03384233, + "epoch": 0.0771959357902964, + "flos": 25775674087680.0, + "grad_norm": 2.1870807733769118, + "language_loss": 0.80221486, + "learning_rate": 3.97673454033267e-06, + "loss": 0.8255384, + "num_input_tokens_seen": 13656835, + "step": 642, + "time_per_iteration": 2.815446138381958 + }, + { + "auxiliary_loss_clip": 0.01278306, + "auxiliary_loss_mlp": 0.01042356, + "balance_loss_clip": 1.06860733, + "balance_loss_mlp": 1.03070307, + "epoch": 0.07731617868093549, + "flos": 19828651847040.0, + "grad_norm": 2.2533229921143785, + "language_loss": 0.82777572, + "learning_rate": 3.976615919750254e-06, + "loss": 0.85098231, + "num_input_tokens_seen": 13674535, + "step": 643, + "time_per_iteration": 3.6461849212646484 + }, + { + "auxiliary_loss_clip": 0.0127213, + "auxiliary_loss_mlp": 0.01045987, + "balance_loss_clip": 1.07484841, + "balance_loss_mlp": 1.03419769, + "epoch": 0.07743642157157458, + "flos": 21324654414720.0, + "grad_norm": 2.432736584095714, + "language_loss": 0.86783326, + "learning_rate": 3.976496999316775e-06, + "loss": 0.89101446, + "num_input_tokens_seen": 13693290, + "step": 644, + "time_per_iteration": 3.5903499126434326 + }, + { + "auxiliary_loss_clip": 0.01279408, + "auxiliary_loss_mlp": 0.01047263, + "balance_loss_clip": 1.07562637, + "balance_loss_mlp": 1.03554463, + "epoch": 0.07755666446221367, + "flos": 19969133938560.0, + "grad_norm": 2.017678059918407, + "language_loss": 0.83901, + "learning_rate": 3.976377779050271e-06, + "loss": 0.86227673, + "num_input_tokens_seen": 13711420, + "step": 645, + "time_per_iteration": 2.8036060333251953 + }, + { + "auxiliary_loss_clip": 0.01256624, + "auxiliary_loss_mlp": 0.01046752, + "balance_loss_clip": 1.0690794, + "balance_loss_mlp": 1.0344733, + "epoch": 0.07767690735285276, + "flos": 23623224514560.0, + "grad_norm": 1.9411006070605967, + "language_loss": 0.8399424, + "learning_rate": 3.976258258968831e-06, + "loss": 0.86297619, + "num_input_tokens_seen": 13729965, + "step": 646, + "time_per_iteration": 2.6935973167419434 + }, + { + "auxiliary_loss_clip": 0.01291462, + "auxiliary_loss_mlp": 0.01047635, + "balance_loss_clip": 1.07307231, + "balance_loss_mlp": 1.03627968, + "epoch": 0.07779715024349185, + "flos": 22236246702720.0, + "grad_norm": 2.2733981815759834, + "language_loss": 0.74227118, + "learning_rate": 3.976138439090583e-06, + "loss": 0.76566213, + "num_input_tokens_seen": 13748045, + "step": 647, + "time_per_iteration": 2.8211722373962402 + }, + { + "auxiliary_loss_clip": 0.01289487, + "auxiliary_loss_mlp": 0.01053385, + "balance_loss_clip": 1.07105482, + "balance_loss_mlp": 1.04073668, + "epoch": 0.07791739313413094, + "flos": 20955097336320.0, + "grad_norm": 2.1879800437727654, + "language_loss": 0.85526872, + "learning_rate": 3.976018319433706e-06, + "loss": 0.8786974, + "num_input_tokens_seen": 13765590, + "step": 648, + "time_per_iteration": 2.7673299312591553 + }, + { + "auxiliary_loss_clip": 0.01271813, + "auxiliary_loss_mlp": 0.01041714, + "balance_loss_clip": 1.07357657, + "balance_loss_mlp": 1.02920866, + "epoch": 0.07803763602477004, + "flos": 19312327797120.0, + "grad_norm": 2.299757732309599, + "language_loss": 0.911874, + "learning_rate": 3.9758979000164205e-06, + "loss": 0.93500924, + "num_input_tokens_seen": 13782410, + "step": 649, + "time_per_iteration": 2.689335584640503 + }, + { + "auxiliary_loss_clip": 0.01290909, + "auxiliary_loss_mlp": 0.01045792, + "balance_loss_clip": 1.06940448, + "balance_loss_mlp": 1.03332865, + "epoch": 0.07815787891540912, + "flos": 22710806213760.0, + "grad_norm": 1.8680825494177487, + "language_loss": 0.7193768, + "learning_rate": 3.975777180856995e-06, + "loss": 0.74274379, + "num_input_tokens_seen": 13801530, + "step": 650, + "time_per_iteration": 2.7815372943878174 + }, + { + "auxiliary_loss_clip": 0.01266355, + "auxiliary_loss_mlp": 0.0105395, + "balance_loss_clip": 1.07347322, + "balance_loss_mlp": 1.04153383, + "epoch": 0.07827812180604822, + "flos": 22711129436160.0, + "grad_norm": 2.3329491073809048, + "language_loss": 0.86227417, + "learning_rate": 3.975656161973742e-06, + "loss": 0.88547719, + "num_input_tokens_seen": 13820615, + "step": 651, + "time_per_iteration": 2.637251615524292 + }, + { + "auxiliary_loss_clip": 0.01265931, + "auxiliary_loss_mlp": 0.01042947, + "balance_loss_clip": 1.07350385, + "balance_loss_mlp": 1.03076386, + "epoch": 0.0783983646966873, + "flos": 21725597001600.0, + "grad_norm": 2.5811453731283955, + "language_loss": 0.88666183, + "learning_rate": 3.9755348433850194e-06, + "loss": 0.90975058, + "num_input_tokens_seen": 13835955, + "step": 652, + "time_per_iteration": 2.631772518157959 + }, + { + "auxiliary_loss_clip": 0.01158037, + "auxiliary_loss_mlp": 0.01008521, + "balance_loss_clip": 1.02816427, + "balance_loss_mlp": 1.0049212, + "epoch": 0.0785186075873264, + "flos": 60640877537280.0, + "grad_norm": 0.9604387172138625, + "language_loss": 0.63643724, + "learning_rate": 3.975413225109232e-06, + "loss": 0.65810281, + "num_input_tokens_seen": 13896505, + "step": 653, + "time_per_iteration": 3.2646756172180176 + }, + { + "auxiliary_loss_clip": 0.01270876, + "auxiliary_loss_mlp": 0.01035617, + "balance_loss_clip": 1.07190323, + "balance_loss_mlp": 1.02406549, + "epoch": 0.0786388504779655, + "flos": 23877902920320.0, + "grad_norm": 3.7653889906365063, + "language_loss": 0.93605506, + "learning_rate": 3.975291307164829e-06, + "loss": 0.95911998, + "num_input_tokens_seen": 13915150, + "step": 654, + "time_per_iteration": 2.66363787651062 + }, + { + "auxiliary_loss_clip": 0.01273109, + "auxiliary_loss_mlp": 0.01045468, + "balance_loss_clip": 1.06445539, + "balance_loss_mlp": 1.03392208, + "epoch": 0.07875909336860458, + "flos": 15158684822400.0, + "grad_norm": 2.1297444149743474, + "language_loss": 0.85111696, + "learning_rate": 3.975169089570306e-06, + "loss": 0.87430269, + "num_input_tokens_seen": 13933525, + "step": 655, + "time_per_iteration": 2.6819920539855957 + }, + { + "auxiliary_loss_clip": 0.01253194, + "auxiliary_loss_mlp": 0.01043563, + "balance_loss_clip": 1.07012296, + "balance_loss_mlp": 1.03129661, + "epoch": 0.07887933625924368, + "flos": 22236857233920.0, + "grad_norm": 2.2147049865396005, + "language_loss": 0.91460311, + "learning_rate": 3.975046572344202e-06, + "loss": 0.93757069, + "num_input_tokens_seen": 13949985, + "step": 656, + "time_per_iteration": 2.661581039428711 + }, + { + "auxiliary_loss_clip": 0.0128555, + "auxiliary_loss_mlp": 0.01044757, + "balance_loss_clip": 1.06672549, + "balance_loss_mlp": 1.03358078, + "epoch": 0.07899957914988276, + "flos": 20777734955520.0, + "grad_norm": 2.018991328543865, + "language_loss": 0.71231556, + "learning_rate": 3.974923755505103e-06, + "loss": 0.73561859, + "num_input_tokens_seen": 13969215, + "step": 657, + "time_per_iteration": 2.779662847518921 + }, + { + "auxiliary_loss_clip": 0.01270161, + "auxiliary_loss_mlp": 0.01045997, + "balance_loss_clip": 1.06613445, + "balance_loss_mlp": 1.03213835, + "epoch": 0.07911982204052186, + "flos": 23003047267200.0, + "grad_norm": 1.8819855388823419, + "language_loss": 0.91107357, + "learning_rate": 3.974800639071641e-06, + "loss": 0.93423516, + "num_input_tokens_seen": 13989935, + "step": 658, + "time_per_iteration": 2.7969212532043457 + }, + { + "auxiliary_loss_clip": 0.01284295, + "auxiliary_loss_mlp": 0.01381938, + "balance_loss_clip": 1.05967641, + "balance_loss_mlp": 0.99985135, + "epoch": 0.07924006493116094, + "flos": 23111389664640.0, + "grad_norm": 2.994675038938138, + "language_loss": 1.00666046, + "learning_rate": 3.974677223062492e-06, + "loss": 1.03332281, + "num_input_tokens_seen": 14007150, + "step": 659, + "time_per_iteration": 2.75994873046875 + }, + { + "auxiliary_loss_clip": 0.01277835, + "auxiliary_loss_mlp": 0.01041325, + "balance_loss_clip": 1.07194507, + "balance_loss_mlp": 1.02935028, + "epoch": 0.07936030782180004, + "flos": 16472153450880.0, + "grad_norm": 2.065045295342362, + "language_loss": 0.74309659, + "learning_rate": 3.974553507496378e-06, + "loss": 0.76628822, + "num_input_tokens_seen": 14025725, + "step": 660, + "time_per_iteration": 2.620663642883301 + }, + { + "auxiliary_loss_clip": 0.01260388, + "auxiliary_loss_mlp": 0.01051596, + "balance_loss_clip": 1.06771171, + "balance_loss_mlp": 1.03905559, + "epoch": 0.07948055071243913, + "flos": 23733290764800.0, + "grad_norm": 2.5958493442348476, + "language_loss": 0.8901819, + "learning_rate": 3.974429492392068e-06, + "loss": 0.91330177, + "num_input_tokens_seen": 14045750, + "step": 661, + "time_per_iteration": 2.749390125274658 + }, + { + "auxiliary_loss_clip": 0.01265956, + "auxiliary_loss_mlp": 0.01382147, + "balance_loss_clip": 1.07490361, + "balance_loss_mlp": 0.99987948, + "epoch": 0.07960079360307822, + "flos": 19573326996480.0, + "grad_norm": 2.0097799502299045, + "language_loss": 0.91047508, + "learning_rate": 3.974305177768373e-06, + "loss": 0.93695611, + "num_input_tokens_seen": 14063960, + "step": 662, + "time_per_iteration": 2.6421048641204834 + }, + { + "auxiliary_loss_clip": 0.0127623, + "auxiliary_loss_mlp": 0.01048572, + "balance_loss_clip": 1.06785274, + "balance_loss_mlp": 1.03646612, + "epoch": 0.07972103649371731, + "flos": 23513409659520.0, + "grad_norm": 2.8874137818224557, + "language_loss": 0.8623302, + "learning_rate": 3.974180563644152e-06, + "loss": 0.88557827, + "num_input_tokens_seen": 14082525, + "step": 663, + "time_per_iteration": 2.780571222305298 + }, + { + "auxiliary_loss_clip": 0.01276743, + "auxiliary_loss_mlp": 0.01043132, + "balance_loss_clip": 1.06886458, + "balance_loss_mlp": 1.03152728, + "epoch": 0.0798412793843564, + "flos": 16726867770240.0, + "grad_norm": 2.534315687114384, + "language_loss": 0.89350283, + "learning_rate": 3.97405565003831e-06, + "loss": 0.91670156, + "num_input_tokens_seen": 14098610, + "step": 664, + "time_per_iteration": 2.6793735027313232 + }, + { + "auxiliary_loss_clip": 0.01290074, + "auxiliary_loss_mlp": 0.0104029, + "balance_loss_clip": 1.06987262, + "balance_loss_mlp": 1.02860808, + "epoch": 0.07996152227499549, + "flos": 18223337214720.0, + "grad_norm": 2.1813084617528498, + "language_loss": 0.7841922, + "learning_rate": 3.973930436969794e-06, + "loss": 0.80749583, + "num_input_tokens_seen": 14117065, + "step": 665, + "time_per_iteration": 2.729094982147217 + }, + { + "auxiliary_loss_clip": 0.01263148, + "auxiliary_loss_mlp": 0.01058104, + "balance_loss_clip": 1.06827784, + "balance_loss_mlp": 1.04591429, + "epoch": 0.08008176516563459, + "flos": 20594877793920.0, + "grad_norm": 5.333544511746557, + "language_loss": 0.86135423, + "learning_rate": 3.973804924457602e-06, + "loss": 0.88456678, + "num_input_tokens_seen": 14135145, + "step": 666, + "time_per_iteration": 2.6880290508270264 + }, + { + "auxiliary_loss_clip": 0.01264549, + "auxiliary_loss_mlp": 0.01045574, + "balance_loss_clip": 1.0667733, + "balance_loss_mlp": 1.03328955, + "epoch": 0.08020200805627367, + "flos": 31834306863360.0, + "grad_norm": 1.7721630744348738, + "language_loss": 0.85886168, + "learning_rate": 3.973679112520771e-06, + "loss": 0.8819629, + "num_input_tokens_seen": 14156860, + "step": 667, + "time_per_iteration": 3.7228925228118896 + }, + { + "auxiliary_loss_clip": 0.01274985, + "auxiliary_loss_mlp": 0.01050784, + "balance_loss_clip": 1.06528687, + "balance_loss_mlp": 1.03782547, + "epoch": 0.08032225094691277, + "flos": 17783503176960.0, + "grad_norm": 2.4961518663584057, + "language_loss": 0.98992795, + "learning_rate": 3.973553001178389e-06, + "loss": 1.01318562, + "num_input_tokens_seen": 14174365, + "step": 668, + "time_per_iteration": 2.718242645263672 + }, + { + "auxiliary_loss_clip": 0.01287359, + "auxiliary_loss_mlp": 0.01042623, + "balance_loss_clip": 1.07049286, + "balance_loss_mlp": 1.03035641, + "epoch": 0.08044249383755185, + "flos": 24061693835520.0, + "grad_norm": 2.183354301932254, + "language_loss": 0.75671875, + "learning_rate": 3.973426590449585e-06, + "loss": 0.78001857, + "num_input_tokens_seen": 14192320, + "step": 669, + "time_per_iteration": 4.58692479133606 + }, + { + "auxiliary_loss_clip": 0.01293635, + "auxiliary_loss_mlp": 0.01051692, + "balance_loss_clip": 1.06718755, + "balance_loss_mlp": 1.03925252, + "epoch": 0.08056273672819095, + "flos": 18223624523520.0, + "grad_norm": 2.0506883975995787, + "language_loss": 0.75581634, + "learning_rate": 3.9732998803535364e-06, + "loss": 0.7792697, + "num_input_tokens_seen": 14210380, + "step": 670, + "time_per_iteration": 3.6097447872161865 + }, + { + "auxiliary_loss_clip": 0.01263528, + "auxiliary_loss_mlp": 0.01058315, + "balance_loss_clip": 1.07299757, + "balance_loss_mlp": 1.04637623, + "epoch": 0.08068297961883003, + "flos": 19676856971520.0, + "grad_norm": 2.8692269456807717, + "language_loss": 0.85102069, + "learning_rate": 3.973172870909465e-06, + "loss": 0.87423915, + "num_input_tokens_seen": 14225145, + "step": 671, + "time_per_iteration": 2.639697551727295 + }, + { + "auxiliary_loss_clip": 0.01285448, + "auxiliary_loss_mlp": 0.01035833, + "balance_loss_clip": 1.07061338, + "balance_loss_mlp": 1.02359617, + "epoch": 0.08080322250946913, + "flos": 23148736830720.0, + "grad_norm": 2.780753336789474, + "language_loss": 0.80751562, + "learning_rate": 3.973045562136638e-06, + "loss": 0.83072841, + "num_input_tokens_seen": 14241960, + "step": 672, + "time_per_iteration": 2.7323427200317383 + }, + { + "auxiliary_loss_clip": 0.01273348, + "auxiliary_loss_mlp": 0.01047265, + "balance_loss_clip": 1.07246602, + "balance_loss_mlp": 1.03593469, + "epoch": 0.08092346540010822, + "flos": 21763626526080.0, + "grad_norm": 2.078910010720214, + "language_loss": 0.9146874, + "learning_rate": 3.972917954054368e-06, + "loss": 0.93789351, + "num_input_tokens_seen": 14260515, + "step": 673, + "time_per_iteration": 2.700065851211548 + }, + { + "auxiliary_loss_clip": 0.01272922, + "auxiliary_loss_mlp": 0.01046013, + "balance_loss_clip": 1.07067919, + "balance_loss_mlp": 1.03307879, + "epoch": 0.08104370829074731, + "flos": 21032485188480.0, + "grad_norm": 2.230827526231449, + "language_loss": 0.82277095, + "learning_rate": 3.972790046682013e-06, + "loss": 0.84596032, + "num_input_tokens_seen": 14279190, + "step": 674, + "time_per_iteration": 2.706210136413574 + }, + { + "auxiliary_loss_clip": 0.01284733, + "auxiliary_loss_mlp": 0.01047524, + "balance_loss_clip": 1.06494367, + "balance_loss_mlp": 1.03603244, + "epoch": 0.0811639511813864, + "flos": 20083186598400.0, + "grad_norm": 1.8609015890595453, + "language_loss": 0.7907095, + "learning_rate": 3.972661840038977e-06, + "loss": 0.81403202, + "num_input_tokens_seen": 14299480, + "step": 675, + "time_per_iteration": 2.798875570297241 + }, + { + "auxiliary_loss_clip": 0.01269389, + "auxiliary_loss_mlp": 0.01050048, + "balance_loss_clip": 1.072047, + "balance_loss_mlp": 1.03849113, + "epoch": 0.08128419407202549, + "flos": 16836718538880.0, + "grad_norm": 2.3760694349585103, + "language_loss": 0.83353877, + "learning_rate": 3.972533334144707e-06, + "loss": 0.8567332, + "num_input_tokens_seen": 14316405, + "step": 676, + "time_per_iteration": 2.674659252166748 + }, + { + "auxiliary_loss_clip": 0.01274306, + "auxiliary_loss_mlp": 0.01046825, + "balance_loss_clip": 1.07278705, + "balance_loss_mlp": 1.03436184, + "epoch": 0.08140443696266458, + "flos": 23769273214080.0, + "grad_norm": 2.360297430850892, + "language_loss": 0.78461313, + "learning_rate": 3.972404529018699e-06, + "loss": 0.80782437, + "num_input_tokens_seen": 14336265, + "step": 677, + "time_per_iteration": 2.668602228164673 + }, + { + "auxiliary_loss_clip": 0.01265833, + "auxiliary_loss_mlp": 0.01035871, + "balance_loss_clip": 1.06339884, + "balance_loss_mlp": 1.02414072, + "epoch": 0.08152467985330367, + "flos": 24390132819840.0, + "grad_norm": 1.8958690655334336, + "language_loss": 0.85543221, + "learning_rate": 3.972275424680493e-06, + "loss": 0.8784492, + "num_input_tokens_seen": 14356375, + "step": 678, + "time_per_iteration": 2.7974932193756104 + }, + { + "auxiliary_loss_clip": 0.01262771, + "auxiliary_loss_mlp": 0.01056162, + "balance_loss_clip": 1.0737977, + "balance_loss_mlp": 1.04419923, + "epoch": 0.08164492274394276, + "flos": 19317750750720.0, + "grad_norm": 2.169911853006817, + "language_loss": 0.91842365, + "learning_rate": 3.972146021149673e-06, + "loss": 0.9416129, + "num_input_tokens_seen": 14374650, + "step": 679, + "time_per_iteration": 2.7444217205047607 + }, + { + "auxiliary_loss_clip": 0.01286193, + "auxiliary_loss_mlp": 0.01037814, + "balance_loss_clip": 1.06927752, + "balance_loss_mlp": 1.02768707, + "epoch": 0.08176516563458186, + "flos": 14830461319680.0, + "grad_norm": 2.469536542318767, + "language_loss": 0.7881611, + "learning_rate": 3.972016318445868e-06, + "loss": 0.81140113, + "num_input_tokens_seen": 14392650, + "step": 680, + "time_per_iteration": 2.737028121948242 + }, + { + "auxiliary_loss_clip": 0.01271031, + "auxiliary_loss_mlp": 0.01038694, + "balance_loss_clip": 1.07152784, + "balance_loss_mlp": 1.02598023, + "epoch": 0.08188540852522094, + "flos": 22602320161920.0, + "grad_norm": 2.0621542650945246, + "language_loss": 0.92255431, + "learning_rate": 3.971886316588757e-06, + "loss": 0.94565159, + "num_input_tokens_seen": 14413155, + "step": 681, + "time_per_iteration": 2.694951295852661 + }, + { + "auxiliary_loss_clip": 0.01269038, + "auxiliary_loss_mlp": 0.01045186, + "balance_loss_clip": 1.06590116, + "balance_loss_mlp": 1.03265095, + "epoch": 0.08200565141586004, + "flos": 19463727623040.0, + "grad_norm": 2.642470248418341, + "language_loss": 0.73183358, + "learning_rate": 3.9717560155980595e-06, + "loss": 0.7549758, + "num_input_tokens_seen": 14428805, + "step": 682, + "time_per_iteration": 2.744990110397339 + }, + { + "auxiliary_loss_clip": 0.01271738, + "auxiliary_loss_mlp": 0.01045025, + "balance_loss_clip": 1.07228065, + "balance_loss_mlp": 1.03380704, + "epoch": 0.08212589430649912, + "flos": 20594662312320.0, + "grad_norm": 1.9812601279277637, + "language_loss": 0.91904688, + "learning_rate": 3.971625415493542e-06, + "loss": 0.94221449, + "num_input_tokens_seen": 14447125, + "step": 683, + "time_per_iteration": 2.657259941101074 + }, + { + "auxiliary_loss_clip": 0.01277102, + "auxiliary_loss_mlp": 0.01036726, + "balance_loss_clip": 1.06496882, + "balance_loss_mlp": 1.02475762, + "epoch": 0.08224613719713822, + "flos": 25953611086080.0, + "grad_norm": 1.9445553102139537, + "language_loss": 0.87356293, + "learning_rate": 3.971494516295017e-06, + "loss": 0.89670122, + "num_input_tokens_seen": 14466575, + "step": 684, + "time_per_iteration": 2.8839123249053955 + }, + { + "auxiliary_loss_clip": 0.01280023, + "auxiliary_loss_mlp": 0.01036248, + "balance_loss_clip": 1.06309414, + "balance_loss_mlp": 1.02558517, + "epoch": 0.08236638008777732, + "flos": 23768734510080.0, + "grad_norm": 2.39856261405829, + "language_loss": 0.85277569, + "learning_rate": 3.971363318022341e-06, + "loss": 0.87593842, + "num_input_tokens_seen": 14487915, + "step": 685, + "time_per_iteration": 2.8047127723693848 + }, + { + "auxiliary_loss_clip": 0.01270633, + "auxiliary_loss_mlp": 0.01039235, + "balance_loss_clip": 1.06387568, + "balance_loss_mlp": 1.02688539, + "epoch": 0.0824866229784164, + "flos": 38799144887040.0, + "grad_norm": 1.9739780085266254, + "language_loss": 0.68377233, + "learning_rate": 3.971231820695417e-06, + "loss": 0.70687103, + "num_input_tokens_seen": 14511530, + "step": 686, + "time_per_iteration": 2.8298768997192383 + }, + { + "auxiliary_loss_clip": 0.01280317, + "auxiliary_loss_mlp": 0.01042568, + "balance_loss_clip": 1.0680095, + "balance_loss_mlp": 1.03095675, + "epoch": 0.0826068658690555, + "flos": 23107762391040.0, + "grad_norm": 2.0968193990652524, + "language_loss": 0.81368232, + "learning_rate": 3.971100024334193e-06, + "loss": 0.8369112, + "num_input_tokens_seen": 14529050, + "step": 687, + "time_per_iteration": 2.7144596576690674 + }, + { + "auxiliary_loss_clip": 0.01260042, + "auxiliary_loss_mlp": 0.01042009, + "balance_loss_clip": 1.06111908, + "balance_loss_mlp": 1.03139889, + "epoch": 0.08272710875969458, + "flos": 21136374299520.0, + "grad_norm": 2.2379233923932, + "language_loss": 0.86703622, + "learning_rate": 3.970967928958663e-06, + "loss": 0.89005673, + "num_input_tokens_seen": 14546165, + "step": 688, + "time_per_iteration": 2.691943407058716 + }, + { + "auxiliary_loss_clip": 0.01296415, + "auxiliary_loss_mlp": 0.01051916, + "balance_loss_clip": 1.06598544, + "balance_loss_mlp": 1.04092479, + "epoch": 0.08284735165033368, + "flos": 19063000517760.0, + "grad_norm": 1.7401850000713621, + "language_loss": 0.83520114, + "learning_rate": 3.970835534588865e-06, + "loss": 0.85868448, + "num_input_tokens_seen": 14563660, + "step": 689, + "time_per_iteration": 2.7841594219207764 + }, + { + "auxiliary_loss_clip": 0.01279151, + "auxiliary_loss_mlp": 0.01049734, + "balance_loss_clip": 1.07242656, + "balance_loss_mlp": 1.03888619, + "epoch": 0.08296759454097276, + "flos": 16727442387840.0, + "grad_norm": 1.8997775743882546, + "language_loss": 0.85461092, + "learning_rate": 3.970702841244883e-06, + "loss": 0.87789977, + "num_input_tokens_seen": 14581980, + "step": 690, + "time_per_iteration": 2.6888315677642822 + }, + { + "auxiliary_loss_clip": 0.01275557, + "auxiliary_loss_mlp": 0.0104476, + "balance_loss_clip": 1.07432795, + "balance_loss_mlp": 1.03365016, + "epoch": 0.08308783743161186, + "flos": 18004928567040.0, + "grad_norm": 2.620103198856793, + "language_loss": 0.82357746, + "learning_rate": 3.970569848946847e-06, + "loss": 0.84678066, + "num_input_tokens_seen": 14601795, + "step": 691, + "time_per_iteration": 2.8064794540405273 + }, + { + "auxiliary_loss_clip": 0.01250605, + "auxiliary_loss_mlp": 0.01037514, + "balance_loss_clip": 1.06920397, + "balance_loss_mlp": 1.02617681, + "epoch": 0.08320808032225095, + "flos": 15079788599040.0, + "grad_norm": 2.3699317210206936, + "language_loss": 0.83176291, + "learning_rate": 3.970436557714932e-06, + "loss": 0.85464418, + "num_input_tokens_seen": 14618315, + "step": 692, + "time_per_iteration": 3.6236202716827393 + }, + { + "auxiliary_loss_clip": 0.0126942, + "auxiliary_loss_mlp": 0.01043378, + "balance_loss_clip": 1.06712139, + "balance_loss_mlp": 1.03164816, + "epoch": 0.08332832321289003, + "flos": 22383085501440.0, + "grad_norm": 2.1165391970134317, + "language_loss": 0.86578035, + "learning_rate": 3.970302967569358e-06, + "loss": 0.88890833, + "num_input_tokens_seen": 14636905, + "step": 693, + "time_per_iteration": 2.7853269577026367 + }, + { + "auxiliary_loss_clip": 0.01266178, + "auxiliary_loss_mlp": 0.01043368, + "balance_loss_clip": 1.07146716, + "balance_loss_mlp": 1.03066623, + "epoch": 0.08344856610352913, + "flos": 24717386655360.0, + "grad_norm": 2.230519119300184, + "language_loss": 0.68572843, + "learning_rate": 3.9701690785303896e-06, + "loss": 0.70882392, + "num_input_tokens_seen": 14656100, + "step": 694, + "time_per_iteration": 2.769578218460083 + }, + { + "auxiliary_loss_clip": 0.0126966, + "auxiliary_loss_mlp": 0.01047659, + "balance_loss_clip": 1.07102346, + "balance_loss_mlp": 1.03628027, + "epoch": 0.08356880899416821, + "flos": 25370206387200.0, + "grad_norm": 2.104985192932198, + "language_loss": 0.88428319, + "learning_rate": 3.970034890618339e-06, + "loss": 0.90745634, + "num_input_tokens_seen": 14675790, + "step": 695, + "time_per_iteration": 3.652106761932373 + }, + { + "auxiliary_loss_clip": 0.012504, + "auxiliary_loss_mlp": 0.01037098, + "balance_loss_clip": 1.06791508, + "balance_loss_mlp": 1.02613664, + "epoch": 0.08368905188480731, + "flos": 24353072962560.0, + "grad_norm": 2.3914441425167676, + "language_loss": 0.87903595, + "learning_rate": 3.969900403853562e-06, + "loss": 0.9019109, + "num_input_tokens_seen": 14694830, + "step": 696, + "time_per_iteration": 3.5851995944976807 + }, + { + "auxiliary_loss_clip": 0.01264308, + "auxiliary_loss_mlp": 0.01046609, + "balance_loss_clip": 1.07523632, + "balance_loss_mlp": 1.0349679, + "epoch": 0.08380929477544641, + "flos": 18037319656320.0, + "grad_norm": 2.2157301660324786, + "language_loss": 0.78024232, + "learning_rate": 3.96976561825646e-06, + "loss": 0.80335152, + "num_input_tokens_seen": 14711920, + "step": 697, + "time_per_iteration": 2.656012773513794 + }, + { + "auxiliary_loss_clip": 0.01295493, + "auxiliary_loss_mlp": 0.01037109, + "balance_loss_clip": 1.0646975, + "balance_loss_mlp": 1.02633274, + "epoch": 0.08392953766608549, + "flos": 26286287875200.0, + "grad_norm": 2.1292528875468246, + "language_loss": 0.86866933, + "learning_rate": 3.969630533847479e-06, + "loss": 0.89199531, + "num_input_tokens_seen": 14730880, + "step": 698, + "time_per_iteration": 2.7805798053741455 + }, + { + "auxiliary_loss_clip": 0.01264033, + "auxiliary_loss_mlp": 0.01040255, + "balance_loss_clip": 1.06711411, + "balance_loss_mlp": 1.02986002, + "epoch": 0.08404978055672459, + "flos": 22492146170880.0, + "grad_norm": 2.0520992642807028, + "language_loss": 0.84349108, + "learning_rate": 3.969495150647113e-06, + "loss": 0.86653399, + "num_input_tokens_seen": 14749050, + "step": 699, + "time_per_iteration": 2.7080554962158203 + }, + { + "auxiliary_loss_clip": 0.01276893, + "auxiliary_loss_mlp": 0.01039196, + "balance_loss_clip": 1.06814146, + "balance_loss_mlp": 1.02754354, + "epoch": 0.08417002344736367, + "flos": 24826878288000.0, + "grad_norm": 2.1344532850607294, + "language_loss": 0.76652801, + "learning_rate": 3.969359468675899e-06, + "loss": 0.78968894, + "num_input_tokens_seen": 14769180, + "step": 700, + "time_per_iteration": 2.745610237121582 + }, + { + "auxiliary_loss_clip": 0.01264552, + "auxiliary_loss_mlp": 0.01034555, + "balance_loss_clip": 1.06965137, + "balance_loss_mlp": 1.0232774, + "epoch": 0.08429026633800277, + "flos": 16945922862720.0, + "grad_norm": 1.9943356315074372, + "language_loss": 0.89326978, + "learning_rate": 3.969223487954418e-06, + "loss": 0.9162609, + "num_input_tokens_seen": 14786640, + "step": 701, + "time_per_iteration": 2.716325283050537 + }, + { + "auxiliary_loss_clip": 0.01283488, + "auxiliary_loss_mlp": 0.01035411, + "balance_loss_clip": 1.06539035, + "balance_loss_mlp": 1.02437234, + "epoch": 0.08441050922864185, + "flos": 23841920471040.0, + "grad_norm": 2.255801562126828, + "language_loss": 0.82876706, + "learning_rate": 3.969087208503301e-06, + "loss": 0.85195601, + "num_input_tokens_seen": 14806720, + "step": 702, + "time_per_iteration": 2.7695693969726562 + }, + { + "auxiliary_loss_clip": 0.01261013, + "auxiliary_loss_mlp": 0.01035042, + "balance_loss_clip": 1.06422901, + "balance_loss_mlp": 1.02398551, + "epoch": 0.08453075211928095, + "flos": 25520205582720.0, + "grad_norm": 2.410499823760644, + "language_loss": 0.84657896, + "learning_rate": 3.968950630343219e-06, + "loss": 0.8695395, + "num_input_tokens_seen": 14823705, + "step": 703, + "time_per_iteration": 2.8227579593658447 + }, + { + "auxiliary_loss_clip": 0.0126724, + "auxiliary_loss_mlp": 0.0103914, + "balance_loss_clip": 1.0640986, + "balance_loss_mlp": 1.02805924, + "epoch": 0.08465099500992004, + "flos": 19532496211200.0, + "grad_norm": 2.6336303231949576, + "language_loss": 0.93545187, + "learning_rate": 3.968813753494892e-06, + "loss": 0.95851564, + "num_input_tokens_seen": 14841865, + "step": 704, + "time_per_iteration": 2.714332342147827 + }, + { + "auxiliary_loss_clip": 0.01267977, + "auxiliary_loss_mlp": 0.0138185, + "balance_loss_clip": 1.06259143, + "balance_loss_mlp": 0.99986565, + "epoch": 0.08477123790055913, + "flos": 29351299403520.0, + "grad_norm": 2.603910812474726, + "language_loss": 0.75467634, + "learning_rate": 3.968676577979084e-06, + "loss": 0.78117454, + "num_input_tokens_seen": 14861415, + "step": 705, + "time_per_iteration": 2.8089559078216553 + }, + { + "auxiliary_loss_clip": 0.01286335, + "auxiliary_loss_mlp": 0.01041374, + "balance_loss_clip": 1.06318104, + "balance_loss_mlp": 1.03019834, + "epoch": 0.08489148079119822, + "flos": 18624495283200.0, + "grad_norm": 3.118956852289312, + "language_loss": 0.78372312, + "learning_rate": 3.968539103816605e-06, + "loss": 0.80700022, + "num_input_tokens_seen": 14879215, + "step": 706, + "time_per_iteration": 2.802781105041504 + }, + { + "auxiliary_loss_clip": 0.01269206, + "auxiliary_loss_mlp": 0.01381929, + "balance_loss_clip": 1.06715941, + "balance_loss_mlp": 0.99989891, + "epoch": 0.0850117236818373, + "flos": 23471393725440.0, + "grad_norm": 1.7793106288858949, + "language_loss": 0.89347541, + "learning_rate": 3.9684013310283085e-06, + "loss": 0.91998684, + "num_input_tokens_seen": 14897900, + "step": 707, + "time_per_iteration": 2.798614501953125 + }, + { + "auxiliary_loss_clip": 0.01276837, + "auxiliary_loss_mlp": 0.01038097, + "balance_loss_clip": 1.07244873, + "balance_loss_mlp": 1.027619, + "epoch": 0.0851319665724764, + "flos": 40625058896640.0, + "grad_norm": 1.7997804492592993, + "language_loss": 0.64214265, + "learning_rate": 3.9682632596350956e-06, + "loss": 0.66529197, + "num_input_tokens_seen": 14919065, + "step": 708, + "time_per_iteration": 2.879823923110962 + }, + { + "auxiliary_loss_clip": 0.01267812, + "auxiliary_loss_mlp": 0.01045444, + "balance_loss_clip": 1.07387543, + "balance_loss_mlp": 1.03343964, + "epoch": 0.0852522094631155, + "flos": 15879554870400.0, + "grad_norm": 2.6962920620675983, + "language_loss": 0.78477991, + "learning_rate": 3.968124889657911e-06, + "loss": 0.80791247, + "num_input_tokens_seen": 14934165, + "step": 709, + "time_per_iteration": 2.6690831184387207 + }, + { + "auxiliary_loss_clip": 0.01284026, + "auxiliary_loss_mlp": 0.0103581, + "balance_loss_clip": 1.06205833, + "balance_loss_mlp": 1.02451491, + "epoch": 0.08537245235375458, + "flos": 14567091822720.0, + "grad_norm": 2.3374455939859056, + "language_loss": 0.90644598, + "learning_rate": 3.967986221117746e-06, + "loss": 0.92964435, + "num_input_tokens_seen": 14950105, + "step": 710, + "time_per_iteration": 2.827065944671631 + }, + { + "auxiliary_loss_clip": 0.01309395, + "auxiliary_loss_mlp": 0.01043156, + "balance_loss_clip": 1.06063461, + "balance_loss_mlp": 1.03186643, + "epoch": 0.08549269524439368, + "flos": 26468929555200.0, + "grad_norm": 2.265096794473243, + "language_loss": 0.86422026, + "learning_rate": 3.967847254035635e-06, + "loss": 0.88774574, + "num_input_tokens_seen": 14969490, + "step": 711, + "time_per_iteration": 2.9359776973724365 + }, + { + "auxiliary_loss_clip": 0.01285727, + "auxiliary_loss_mlp": 0.01043122, + "balance_loss_clip": 1.06773126, + "balance_loss_mlp": 1.03145695, + "epoch": 0.08561293813503276, + "flos": 13590214565760.0, + "grad_norm": 2.4390496329747013, + "language_loss": 0.8643899, + "learning_rate": 3.967707988432661e-06, + "loss": 0.88767838, + "num_input_tokens_seen": 14987195, + "step": 712, + "time_per_iteration": 3.068124532699585 + }, + { + "auxiliary_loss_clip": 0.01256238, + "auxiliary_loss_mlp": 0.01037822, + "balance_loss_clip": 1.06875944, + "balance_loss_mlp": 1.02674186, + "epoch": 0.08573318102567186, + "flos": 26943524979840.0, + "grad_norm": 2.220193677911842, + "language_loss": 0.88068485, + "learning_rate": 3.967568424329949e-06, + "loss": 0.90362549, + "num_input_tokens_seen": 15007620, + "step": 713, + "time_per_iteration": 2.7133076190948486 + }, + { + "auxiliary_loss_clip": 0.01175933, + "auxiliary_loss_mlp": 0.01010499, + "balance_loss_clip": 1.0356791, + "balance_loss_mlp": 1.00642228, + "epoch": 0.08585342391631094, + "flos": 67302739319040.0, + "grad_norm": 0.8160431369450866, + "language_loss": 0.55500722, + "learning_rate": 3.967428561748671e-06, + "loss": 0.57687151, + "num_input_tokens_seen": 15075590, + "step": 714, + "time_per_iteration": 3.615741729736328 + }, + { + "auxiliary_loss_clip": 0.01275775, + "auxiliary_loss_mlp": 0.0104677, + "balance_loss_clip": 1.0591799, + "balance_loss_mlp": 1.03475332, + "epoch": 0.08597366680695004, + "flos": 22456594684800.0, + "grad_norm": 1.9352488376087817, + "language_loss": 0.87541884, + "learning_rate": 3.967288400710045e-06, + "loss": 0.89864433, + "num_input_tokens_seen": 15095055, + "step": 715, + "time_per_iteration": 2.796123504638672 + }, + { + "auxiliary_loss_clip": 0.01279189, + "auxiliary_loss_mlp": 0.01038685, + "balance_loss_clip": 1.06950736, + "balance_loss_mlp": 1.02678823, + "epoch": 0.08609390969758914, + "flos": 23550505430400.0, + "grad_norm": 1.9047100419202667, + "language_loss": 0.8878988, + "learning_rate": 3.9671479412353335e-06, + "loss": 0.9110775, + "num_input_tokens_seen": 15113520, + "step": 716, + "time_per_iteration": 2.677969217300415 + }, + { + "auxiliary_loss_clip": 0.01261907, + "auxiliary_loss_mlp": 0.01040286, + "balance_loss_clip": 1.06932282, + "balance_loss_mlp": 1.02908635, + "epoch": 0.08621415258822822, + "flos": 25885848078720.0, + "grad_norm": 7.135059210190599, + "language_loss": 0.74174553, + "learning_rate": 3.967007183345843e-06, + "loss": 0.76476747, + "num_input_tokens_seen": 15133375, + "step": 717, + "time_per_iteration": 2.7075417041778564 + }, + { + "auxiliary_loss_clip": 0.01260199, + "auxiliary_loss_mlp": 0.0105257, + "balance_loss_clip": 1.06957638, + "balance_loss_mlp": 1.04117393, + "epoch": 0.08633439547886732, + "flos": 13589568120960.0, + "grad_norm": 2.191623165788272, + "language_loss": 0.89553452, + "learning_rate": 3.966866127062927e-06, + "loss": 0.91866219, + "num_input_tokens_seen": 15150500, + "step": 718, + "time_per_iteration": 3.632659673690796 + }, + { + "auxiliary_loss_clip": 0.01160235, + "auxiliary_loss_mlp": 0.01008381, + "balance_loss_clip": 1.0369339, + "balance_loss_mlp": 1.00456619, + "epoch": 0.0864546383695064, + "flos": 57767342434560.0, + "grad_norm": 0.8951114350203411, + "language_loss": 0.62677312, + "learning_rate": 3.966724772407982e-06, + "loss": 0.64845926, + "num_input_tokens_seen": 15208015, + "step": 719, + "time_per_iteration": 3.075124979019165 + }, + { + "auxiliary_loss_clip": 0.01269684, + "auxiliary_loss_mlp": 0.01052852, + "balance_loss_clip": 1.06510758, + "balance_loss_mlp": 1.04190814, + "epoch": 0.0865748812601455, + "flos": 20046952753920.0, + "grad_norm": 2.281491993459996, + "language_loss": 0.88805699, + "learning_rate": 3.966583119402454e-06, + "loss": 0.91128236, + "num_input_tokens_seen": 15224780, + "step": 720, + "time_per_iteration": 2.77402663230896 + }, + { + "auxiliary_loss_clip": 0.01265099, + "auxiliary_loss_mlp": 0.01381455, + "balance_loss_clip": 1.07068264, + "balance_loss_mlp": 1.00004792, + "epoch": 0.08669512415078459, + "flos": 35262446935680.0, + "grad_norm": 1.724716739510157, + "language_loss": 0.82196236, + "learning_rate": 3.9664411680678305e-06, + "loss": 0.84842789, + "num_input_tokens_seen": 15246535, + "step": 721, + "time_per_iteration": 4.683412313461304 + }, + { + "auxiliary_loss_clip": 0.0117656, + "auxiliary_loss_mlp": 0.01004909, + "balance_loss_clip": 1.03128195, + "balance_loss_mlp": 1.00114155, + "epoch": 0.08681536704142367, + "flos": 65654870048640.0, + "grad_norm": 0.8537643108306767, + "language_loss": 0.61459577, + "learning_rate": 3.966298918425644e-06, + "loss": 0.63641047, + "num_input_tokens_seen": 15304025, + "step": 722, + "time_per_iteration": 4.101659059524536 + }, + { + "auxiliary_loss_clip": 0.01266641, + "auxiliary_loss_mlp": 0.010456, + "balance_loss_clip": 1.06878698, + "balance_loss_mlp": 1.03466272, + "epoch": 0.08693560993206277, + "flos": 34529940881280.0, + "grad_norm": 1.9410216839189087, + "language_loss": 0.82605433, + "learning_rate": 3.966156370497476e-06, + "loss": 0.84917682, + "num_input_tokens_seen": 15327635, + "step": 723, + "time_per_iteration": 2.8239455223083496 + }, + { + "auxiliary_loss_clip": 0.01270815, + "auxiliary_loss_mlp": 0.01043087, + "balance_loss_clip": 1.07031155, + "balance_loss_mlp": 1.03162503, + "epoch": 0.08705585282270185, + "flos": 23149419189120.0, + "grad_norm": 1.855781499620921, + "language_loss": 0.88574094, + "learning_rate": 3.96601352430495e-06, + "loss": 0.90888, + "num_input_tokens_seen": 15347405, + "step": 724, + "time_per_iteration": 2.762645721435547 + }, + { + "auxiliary_loss_clip": 0.01274944, + "auxiliary_loss_mlp": 0.01039295, + "balance_loss_clip": 1.07208908, + "balance_loss_mlp": 1.02844703, + "epoch": 0.08717609571334095, + "flos": 29497599498240.0, + "grad_norm": 1.5461595969414001, + "language_loss": 0.83094269, + "learning_rate": 3.965870379869735e-06, + "loss": 0.85408509, + "num_input_tokens_seen": 15369450, + "step": 725, + "time_per_iteration": 2.782634973526001 + }, + { + "auxiliary_loss_clip": 0.01262282, + "auxiliary_loss_mlp": 0.01039575, + "balance_loss_clip": 1.06638992, + "balance_loss_mlp": 1.02934074, + "epoch": 0.08729633860398003, + "flos": 20667489137280.0, + "grad_norm": 2.576366736117941, + "language_loss": 0.86846536, + "learning_rate": 3.965726937213547e-06, + "loss": 0.8914839, + "num_input_tokens_seen": 15388085, + "step": 726, + "time_per_iteration": 2.728752851486206 + }, + { + "auxiliary_loss_clip": 0.01257106, + "auxiliary_loss_mlp": 0.01044637, + "balance_loss_clip": 1.06538367, + "balance_loss_mlp": 1.03268039, + "epoch": 0.08741658149461913, + "flos": 18369493655040.0, + "grad_norm": 1.9303211847952977, + "language_loss": 0.807338, + "learning_rate": 3.965583196358144e-06, + "loss": 0.83035541, + "num_input_tokens_seen": 15407120, + "step": 727, + "time_per_iteration": 2.6966090202331543 + }, + { + "auxiliary_loss_clip": 0.01259636, + "auxiliary_loss_mlp": 0.01038236, + "balance_loss_clip": 1.07153034, + "balance_loss_mlp": 1.02616024, + "epoch": 0.08753682438525823, + "flos": 18729677283840.0, + "grad_norm": 2.226438900539452, + "language_loss": 0.74534041, + "learning_rate": 3.965439157325335e-06, + "loss": 0.76831913, + "num_input_tokens_seen": 15424485, + "step": 728, + "time_per_iteration": 2.654573917388916 + }, + { + "auxiliary_loss_clip": 0.01265559, + "auxiliary_loss_mlp": 0.01038831, + "balance_loss_clip": 1.06431293, + "balance_loss_mlp": 1.02736866, + "epoch": 0.08765706727589731, + "flos": 27776113303680.0, + "grad_norm": 1.8917402841273738, + "language_loss": 0.76051378, + "learning_rate": 3.965294820136968e-06, + "loss": 0.78355765, + "num_input_tokens_seen": 15446285, + "step": 729, + "time_per_iteration": 2.8330564498901367 + }, + { + "auxiliary_loss_clip": 0.0127169, + "auxiliary_loss_mlp": 0.01037704, + "balance_loss_clip": 1.06967735, + "balance_loss_mlp": 1.02611709, + "epoch": 0.08777731016653641, + "flos": 24389127239040.0, + "grad_norm": 1.9752903540895295, + "language_loss": 0.86960196, + "learning_rate": 3.965150184814938e-06, + "loss": 0.8926959, + "num_input_tokens_seen": 15465770, + "step": 730, + "time_per_iteration": 2.7928316593170166 + }, + { + "auxiliary_loss_clip": 0.01256628, + "auxiliary_loss_mlp": 0.01052302, + "balance_loss_clip": 1.06676316, + "balance_loss_mlp": 1.04067278, + "epoch": 0.08789755305717549, + "flos": 21981855605760.0, + "grad_norm": 2.417456889968886, + "language_loss": 0.76688504, + "learning_rate": 3.965005251381189e-06, + "loss": 0.78997433, + "num_input_tokens_seen": 15483705, + "step": 731, + "time_per_iteration": 2.7541511058807373 + }, + { + "auxiliary_loss_clip": 0.01160658, + "auxiliary_loss_mlp": 0.01006778, + "balance_loss_clip": 1.03358281, + "balance_loss_mlp": 1.00293922, + "epoch": 0.08801779594781459, + "flos": 58360120583040.0, + "grad_norm": 1.5241364354924107, + "language_loss": 0.64597857, + "learning_rate": 3.964860019857705e-06, + "loss": 0.66765285, + "num_input_tokens_seen": 15548620, + "step": 732, + "time_per_iteration": 3.403531789779663 + }, + { + "auxiliary_loss_clip": 0.01256092, + "auxiliary_loss_mlp": 0.01042744, + "balance_loss_clip": 1.07219005, + "balance_loss_mlp": 1.03215814, + "epoch": 0.08813803883845367, + "flos": 23294785530240.0, + "grad_norm": 1.801586961814516, + "language_loss": 0.84180391, + "learning_rate": 3.964714490266518e-06, + "loss": 0.86479223, + "num_input_tokens_seen": 15569265, + "step": 733, + "time_per_iteration": 2.770189046859741 + }, + { + "auxiliary_loss_clip": 0.01155875, + "auxiliary_loss_mlp": 0.01004218, + "balance_loss_clip": 1.03253937, + "balance_loss_mlp": 1.00002158, + "epoch": 0.08825828172909277, + "flos": 63424924882560.0, + "grad_norm": 0.9120504094050682, + "language_loss": 0.64619482, + "learning_rate": 3.964568662629706e-06, + "loss": 0.66779578, + "num_input_tokens_seen": 15630570, + "step": 734, + "time_per_iteration": 3.1728675365448 + }, + { + "auxiliary_loss_clip": 0.01261309, + "auxiliary_loss_mlp": 0.01038263, + "balance_loss_clip": 1.06713879, + "balance_loss_mlp": 1.02711141, + "epoch": 0.08837852461973186, + "flos": 26720986268160.0, + "grad_norm": 2.0484066748223437, + "language_loss": 0.84283525, + "learning_rate": 3.9644225369693895e-06, + "loss": 0.86583096, + "num_input_tokens_seen": 15650870, + "step": 735, + "time_per_iteration": 2.84027099609375 + }, + { + "auxiliary_loss_clip": 0.01259853, + "auxiliary_loss_mlp": 0.01048614, + "balance_loss_clip": 1.07485306, + "balance_loss_mlp": 1.03765893, + "epoch": 0.08849876751037095, + "flos": 27265427688960.0, + "grad_norm": 2.0901339421264833, + "language_loss": 0.86616933, + "learning_rate": 3.964276113307735e-06, + "loss": 0.88925397, + "num_input_tokens_seen": 15670835, + "step": 736, + "time_per_iteration": 2.7695462703704834 + }, + { + "auxiliary_loss_clip": 0.01271733, + "auxiliary_loss_mlp": 0.01043015, + "balance_loss_clip": 1.06575489, + "balance_loss_mlp": 1.03104639, + "epoch": 0.08861901040101004, + "flos": 19828759587840.0, + "grad_norm": 1.922566675184825, + "language_loss": 0.80565059, + "learning_rate": 3.9641293916669574e-06, + "loss": 0.82879806, + "num_input_tokens_seen": 15689795, + "step": 737, + "time_per_iteration": 2.76375675201416 + }, + { + "auxiliary_loss_clip": 0.01272197, + "auxiliary_loss_mlp": 0.01045883, + "balance_loss_clip": 1.06370783, + "balance_loss_mlp": 1.03453374, + "epoch": 0.08873925329164913, + "flos": 23658704173440.0, + "grad_norm": 1.8542865099505592, + "language_loss": 0.82539856, + "learning_rate": 3.9639823720693115e-06, + "loss": 0.84857935, + "num_input_tokens_seen": 15711650, + "step": 738, + "time_per_iteration": 2.9284534454345703 + }, + { + "auxiliary_loss_clip": 0.01167386, + "auxiliary_loss_mlp": 0.01009965, + "balance_loss_clip": 1.03081989, + "balance_loss_mlp": 1.00686514, + "epoch": 0.08885949618228822, + "flos": 71831541893760.0, + "grad_norm": 0.842912975043884, + "language_loss": 0.6002537, + "learning_rate": 3.963835054537102e-06, + "loss": 0.62202722, + "num_input_tokens_seen": 15780615, + "step": 739, + "time_per_iteration": 3.352262496948242 + }, + { + "auxiliary_loss_clip": 0.01263545, + "auxiliary_loss_mlp": 0.01047477, + "balance_loss_clip": 1.06313384, + "balance_loss_mlp": 1.03562796, + "epoch": 0.08897973907292732, + "flos": 22346169298560.0, + "grad_norm": 2.0591751559225258, + "language_loss": 0.60467505, + "learning_rate": 3.963687439092676e-06, + "loss": 0.62778527, + "num_input_tokens_seen": 15801300, + "step": 740, + "time_per_iteration": 2.805924415588379 + }, + { + "auxiliary_loss_clip": 0.0126325, + "auxiliary_loss_mlp": 0.01050098, + "balance_loss_clip": 1.07086754, + "balance_loss_mlp": 1.03874326, + "epoch": 0.0890999819635664, + "flos": 21251827589760.0, + "grad_norm": 1.9823052666184389, + "language_loss": 0.80295432, + "learning_rate": 3.963539525758427e-06, + "loss": 0.82608777, + "num_input_tokens_seen": 15820860, + "step": 741, + "time_per_iteration": 2.720853328704834 + }, + { + "auxiliary_loss_clip": 0.01274448, + "auxiliary_loss_mlp": 0.01053147, + "balance_loss_clip": 1.06880116, + "balance_loss_mlp": 1.04262114, + "epoch": 0.0892202248542055, + "flos": 25370888745600.0, + "grad_norm": 1.9912118721698813, + "language_loss": 0.67479885, + "learning_rate": 3.9633913145567925e-06, + "loss": 0.69807482, + "num_input_tokens_seen": 15841350, + "step": 742, + "time_per_iteration": 2.813215494155884 + }, + { + "auxiliary_loss_clip": 0.01274477, + "auxiliary_loss_mlp": 0.01035286, + "balance_loss_clip": 1.07055652, + "balance_loss_mlp": 1.02428865, + "epoch": 0.08934046774484458, + "flos": 24457895827200.0, + "grad_norm": 1.9015985291418658, + "language_loss": 0.81498134, + "learning_rate": 3.9632428055102575e-06, + "loss": 0.83807898, + "num_input_tokens_seen": 15861360, + "step": 743, + "time_per_iteration": 2.7939062118530273 + }, + { + "auxiliary_loss_clip": 0.0126768, + "auxiliary_loss_mlp": 0.01050094, + "balance_loss_clip": 1.07215095, + "balance_loss_mlp": 1.03800654, + "epoch": 0.08946071063548368, + "flos": 35772773414400.0, + "grad_norm": 2.011036687257273, + "language_loss": 0.67218512, + "learning_rate": 3.9630939986413495e-06, + "loss": 0.69536293, + "num_input_tokens_seen": 15883160, + "step": 744, + "time_per_iteration": 3.908590316772461 + }, + { + "auxiliary_loss_clip": 0.0126282, + "auxiliary_loss_mlp": 0.01041536, + "balance_loss_clip": 1.06437588, + "balance_loss_mlp": 1.03010952, + "epoch": 0.08958095352612276, + "flos": 14356584167040.0, + "grad_norm": 1.6305874160816438, + "language_loss": 0.78175014, + "learning_rate": 3.962944893972643e-06, + "loss": 0.80479372, + "num_input_tokens_seen": 15901610, + "step": 745, + "time_per_iteration": 2.8494229316711426 + }, + { + "auxiliary_loss_clip": 0.0126788, + "auxiliary_loss_mlp": 0.01042172, + "balance_loss_clip": 1.06862497, + "balance_loss_mlp": 1.03061485, + "epoch": 0.08970119641676186, + "flos": 17853277345920.0, + "grad_norm": 2.6924519194913397, + "language_loss": 0.90840077, + "learning_rate": 3.962795491526756e-06, + "loss": 0.93150127, + "num_input_tokens_seen": 15918770, + "step": 746, + "time_per_iteration": 2.7203738689422607 + }, + { + "auxiliary_loss_clip": 0.01260557, + "auxiliary_loss_mlp": 0.01041912, + "balance_loss_clip": 1.07280886, + "balance_loss_mlp": 1.03132594, + "epoch": 0.08982143930740095, + "flos": 20811670329600.0, + "grad_norm": 3.1165436725186706, + "language_loss": 0.89367074, + "learning_rate": 3.962645791326354e-06, + "loss": 0.91669542, + "num_input_tokens_seen": 15938025, + "step": 747, + "time_per_iteration": 5.576411485671997 + }, + { + "auxiliary_loss_clip": 0.01259216, + "auxiliary_loss_mlp": 0.01056857, + "balance_loss_clip": 1.07098293, + "balance_loss_mlp": 1.04594922, + "epoch": 0.08994168219804004, + "flos": 24097712198400.0, + "grad_norm": 1.9894411289364142, + "language_loss": 0.83311146, + "learning_rate": 3.962495793394146e-06, + "loss": 0.85627222, + "num_input_tokens_seen": 15957215, + "step": 748, + "time_per_iteration": 2.7613413333892822 + }, + { + "auxiliary_loss_clip": 0.01139934, + "auxiliary_loss_mlp": 0.01024042, + "balance_loss_clip": 1.02968693, + "balance_loss_mlp": 1.02022743, + "epoch": 0.09006192508867913, + "flos": 57188893812480.0, + "grad_norm": 0.7452622635389927, + "language_loss": 0.61207968, + "learning_rate": 3.9623454977528864e-06, + "loss": 0.63371944, + "num_input_tokens_seen": 16015870, + "step": 749, + "time_per_iteration": 3.0577685832977295 + }, + { + "auxiliary_loss_clip": 0.01278989, + "auxiliary_loss_mlp": 0.01044275, + "balance_loss_clip": 1.06569684, + "balance_loss_mlp": 1.03288436, + "epoch": 0.09018216797931822, + "flos": 20487505063680.0, + "grad_norm": 1.7896664719526496, + "language_loss": 0.84943932, + "learning_rate": 3.962194904425375e-06, + "loss": 0.87267196, + "num_input_tokens_seen": 16036500, + "step": 750, + "time_per_iteration": 2.7837793827056885 + }, + { + "auxiliary_loss_clip": 0.01259831, + "auxiliary_loss_mlp": 0.01048463, + "balance_loss_clip": 1.07120657, + "balance_loss_mlp": 1.03663111, + "epoch": 0.09030241086995731, + "flos": 22638123043200.0, + "grad_norm": 2.179217080758297, + "language_loss": 0.6825732, + "learning_rate": 3.9620440134344566e-06, + "loss": 0.70565611, + "num_input_tokens_seen": 16054655, + "step": 751, + "time_per_iteration": 2.6919398307800293 + }, + { + "auxiliary_loss_clip": 0.01278865, + "auxiliary_loss_mlp": 0.01051869, + "balance_loss_clip": 1.06829596, + "balance_loss_mlp": 1.04016256, + "epoch": 0.09042265376059641, + "flos": 21871502046720.0, + "grad_norm": 2.204011992623906, + "language_loss": 0.82538617, + "learning_rate": 3.9618928248030215e-06, + "loss": 0.84869355, + "num_input_tokens_seen": 16074165, + "step": 752, + "time_per_iteration": 2.7260115146636963 + }, + { + "auxiliary_loss_clip": 0.01262362, + "auxiliary_loss_mlp": 0.01039282, + "balance_loss_clip": 1.07173622, + "balance_loss_mlp": 1.02815986, + "epoch": 0.0905428966512355, + "flos": 24316192673280.0, + "grad_norm": 2.287916311557811, + "language_loss": 0.82852101, + "learning_rate": 3.961741338554005e-06, + "loss": 0.85153747, + "num_input_tokens_seen": 16092505, + "step": 753, + "time_per_iteration": 2.700364589691162 + }, + { + "auxiliary_loss_clip": 0.01281735, + "auxiliary_loss_mlp": 0.01044215, + "balance_loss_clip": 1.07253575, + "balance_loss_mlp": 1.03302729, + "epoch": 0.09066313954187459, + "flos": 35845061535360.0, + "grad_norm": 1.9947425515088129, + "language_loss": 0.75978816, + "learning_rate": 3.9615895547103865e-06, + "loss": 0.78304762, + "num_input_tokens_seen": 16116150, + "step": 754, + "time_per_iteration": 2.8557252883911133 + }, + { + "auxiliary_loss_clip": 0.01270613, + "auxiliary_loss_mlp": 0.01046361, + "balance_loss_clip": 1.06897593, + "balance_loss_mlp": 1.03530407, + "epoch": 0.09078338243251367, + "flos": 29168729550720.0, + "grad_norm": 2.093857033897267, + "language_loss": 0.77832305, + "learning_rate": 3.961437473295193e-06, + "loss": 0.80149281, + "num_input_tokens_seen": 16136295, + "step": 755, + "time_per_iteration": 2.837494373321533 + }, + { + "auxiliary_loss_clip": 0.01264895, + "auxiliary_loss_mlp": 0.0104567, + "balance_loss_clip": 1.05630243, + "balance_loss_mlp": 1.03528059, + "epoch": 0.09090362532315277, + "flos": 21907699977600.0, + "grad_norm": 2.4748658736237332, + "language_loss": 0.72293317, + "learning_rate": 3.961285094331495e-06, + "loss": 0.74603891, + "num_input_tokens_seen": 16154210, + "step": 756, + "time_per_iteration": 2.837938070297241 + }, + { + "auxiliary_loss_clip": 0.01254674, + "auxiliary_loss_mlp": 0.01048295, + "balance_loss_clip": 1.07282388, + "balance_loss_mlp": 1.03785825, + "epoch": 0.09102386821379185, + "flos": 27344503480320.0, + "grad_norm": 1.8735069018192851, + "language_loss": 0.85765231, + "learning_rate": 3.961132417842406e-06, + "loss": 0.88068199, + "num_input_tokens_seen": 16173995, + "step": 757, + "time_per_iteration": 2.6934545040130615 + }, + { + "auxiliary_loss_clip": 0.0124883, + "auxiliary_loss_mlp": 0.01046177, + "balance_loss_clip": 1.06665182, + "balance_loss_mlp": 1.03536499, + "epoch": 0.09114411110443095, + "flos": 20813501923200.0, + "grad_norm": 2.9294079850195693, + "language_loss": 0.75575346, + "learning_rate": 3.960979443851089e-06, + "loss": 0.77870357, + "num_input_tokens_seen": 16191020, + "step": 758, + "time_per_iteration": 2.9018630981445312 + }, + { + "auxiliary_loss_clip": 0.01265764, + "auxiliary_loss_mlp": 0.01039698, + "balance_loss_clip": 1.0670712, + "balance_loss_mlp": 1.02773535, + "epoch": 0.09126435399507005, + "flos": 26145949438080.0, + "grad_norm": 1.7822647524343878, + "language_loss": 0.78881729, + "learning_rate": 3.96082617238075e-06, + "loss": 0.81187189, + "num_input_tokens_seen": 16213645, + "step": 759, + "time_per_iteration": 2.7252869606018066 + }, + { + "auxiliary_loss_clip": 0.01266811, + "auxiliary_loss_mlp": 0.01039112, + "balance_loss_clip": 1.06626225, + "balance_loss_mlp": 1.02860343, + "epoch": 0.09138459688570913, + "flos": 24388911757440.0, + "grad_norm": 2.4931174541983623, + "language_loss": 0.79778177, + "learning_rate": 3.960672603454639e-06, + "loss": 0.82084095, + "num_input_tokens_seen": 16233625, + "step": 760, + "time_per_iteration": 2.7707905769348145 + }, + { + "auxiliary_loss_clip": 0.01253591, + "auxiliary_loss_mlp": 0.01041398, + "balance_loss_clip": 1.06932616, + "balance_loss_mlp": 1.02982271, + "epoch": 0.09150483977634823, + "flos": 21032664756480.0, + "grad_norm": 5.754478248775613, + "language_loss": 0.76830858, + "learning_rate": 3.960518737096054e-06, + "loss": 0.79125845, + "num_input_tokens_seen": 16253255, + "step": 761, + "time_per_iteration": 2.6809611320495605 + }, + { + "auxiliary_loss_clip": 0.01266514, + "auxiliary_loss_mlp": 0.01035109, + "balance_loss_clip": 1.07183671, + "balance_loss_mlp": 1.024225, + "epoch": 0.09162508266698731, + "flos": 22856998567680.0, + "grad_norm": 4.80272123237637, + "language_loss": 0.7299155, + "learning_rate": 3.960364573328334e-06, + "loss": 0.75293171, + "num_input_tokens_seen": 16272580, + "step": 762, + "time_per_iteration": 2.68796968460083 + }, + { + "auxiliary_loss_clip": 0.01280115, + "auxiliary_loss_mlp": 0.01041161, + "balance_loss_clip": 1.06723416, + "balance_loss_mlp": 1.02983069, + "epoch": 0.0917453255576264, + "flos": 21724411852800.0, + "grad_norm": 2.2270249362789025, + "language_loss": 0.88972986, + "learning_rate": 3.9602101121748675e-06, + "loss": 0.91294265, + "num_input_tokens_seen": 16293075, + "step": 763, + "time_per_iteration": 2.7321557998657227 + }, + { + "auxiliary_loss_clip": 0.01270911, + "auxiliary_loss_mlp": 0.01034866, + "balance_loss_clip": 1.07013893, + "balance_loss_mlp": 1.02425051, + "epoch": 0.0918655684482655, + "flos": 14609215497600.0, + "grad_norm": 2.4710393350844493, + "language_loss": 0.72761905, + "learning_rate": 3.960055353659085e-06, + "loss": 0.75067687, + "num_input_tokens_seen": 16310185, + "step": 764, + "time_per_iteration": 2.6455953121185303 + }, + { + "auxiliary_loss_clip": 0.01283822, + "auxiliary_loss_mlp": 0.01047353, + "balance_loss_clip": 1.06924522, + "balance_loss_mlp": 1.0369575, + "epoch": 0.09198581133890459, + "flos": 23435016226560.0, + "grad_norm": 1.9104621607954244, + "language_loss": 0.83568609, + "learning_rate": 3.959900297804465e-06, + "loss": 0.85899782, + "num_input_tokens_seen": 16330355, + "step": 765, + "time_per_iteration": 2.7818002700805664 + }, + { + "auxiliary_loss_clip": 0.01255562, + "auxiliary_loss_mlp": 0.01038201, + "balance_loss_clip": 1.06458402, + "balance_loss_mlp": 1.02825236, + "epoch": 0.09210605422954368, + "flos": 16795887753600.0, + "grad_norm": 1.8942769042644925, + "language_loss": 0.77359319, + "learning_rate": 3.9597449446345276e-06, + "loss": 0.79653078, + "num_input_tokens_seen": 16347600, + "step": 766, + "time_per_iteration": 2.678084373474121 + }, + { + "auxiliary_loss_clip": 0.01250734, + "auxiliary_loss_mlp": 0.01042888, + "balance_loss_clip": 1.06198418, + "balance_loss_mlp": 1.0322423, + "epoch": 0.09222629712018277, + "flos": 22674249146880.0, + "grad_norm": 3.152982070663724, + "language_loss": 0.83635515, + "learning_rate": 3.95958929417284e-06, + "loss": 0.85929132, + "num_input_tokens_seen": 16365755, + "step": 767, + "time_per_iteration": 2.7077183723449707 + }, + { + "auxiliary_loss_clip": 0.0114824, + "auxiliary_loss_mlp": 0.01005431, + "balance_loss_clip": 1.02682376, + "balance_loss_mlp": 1.00202203, + "epoch": 0.09234654001082186, + "flos": 69976756327680.0, + "grad_norm": 0.7300373791954388, + "language_loss": 0.58772397, + "learning_rate": 3.9594333464430145e-06, + "loss": 0.60926068, + "num_input_tokens_seen": 16435245, + "step": 768, + "time_per_iteration": 3.326063394546509 + }, + { + "auxiliary_loss_clip": 0.01283691, + "auxiliary_loss_mlp": 0.01053049, + "balance_loss_clip": 1.05544293, + "balance_loss_mlp": 1.04311299, + "epoch": 0.09246678290146094, + "flos": 20011437181440.0, + "grad_norm": 1.9510166729323604, + "language_loss": 0.87805688, + "learning_rate": 3.959277101468709e-06, + "loss": 0.90142435, + "num_input_tokens_seen": 16454795, + "step": 769, + "time_per_iteration": 2.8686656951904297 + }, + { + "auxiliary_loss_clip": 0.01252318, + "auxiliary_loss_mlp": 0.01041997, + "balance_loss_clip": 1.06592536, + "balance_loss_mlp": 1.03088045, + "epoch": 0.09258702579210004, + "flos": 17747448900480.0, + "grad_norm": 2.5534863088158595, + "language_loss": 0.78486443, + "learning_rate": 3.959120559273624e-06, + "loss": 0.80780756, + "num_input_tokens_seen": 16472580, + "step": 770, + "time_per_iteration": 3.813966751098633 + }, + { + "auxiliary_loss_clip": 0.01253488, + "auxiliary_loss_mlp": 0.01040915, + "balance_loss_clip": 1.06959248, + "balance_loss_mlp": 1.03015637, + "epoch": 0.09270726868273914, + "flos": 20886544229760.0, + "grad_norm": 1.7724361047558501, + "language_loss": 0.83632404, + "learning_rate": 3.958963719881509e-06, + "loss": 0.85926807, + "num_input_tokens_seen": 16490670, + "step": 771, + "time_per_iteration": 2.7063748836517334 + }, + { + "auxiliary_loss_clip": 0.01263082, + "auxiliary_loss_mlp": 0.01042018, + "balance_loss_clip": 1.07280409, + "balance_loss_mlp": 1.03176033, + "epoch": 0.09282751157337822, + "flos": 17015697031680.0, + "grad_norm": 2.8212718130139214, + "language_loss": 0.93606913, + "learning_rate": 3.958806583316154e-06, + "loss": 0.95912015, + "num_input_tokens_seen": 16508640, + "step": 772, + "time_per_iteration": 3.5889346599578857 + }, + { + "auxiliary_loss_clip": 0.01261135, + "auxiliary_loss_mlp": 0.01039568, + "balance_loss_clip": 1.07584846, + "balance_loss_mlp": 1.02904165, + "epoch": 0.09294775446401732, + "flos": 32523647748480.0, + "grad_norm": 2.1016901064211395, + "language_loss": 0.78806889, + "learning_rate": 3.9586491496013985e-06, + "loss": 0.81107593, + "num_input_tokens_seen": 16531035, + "step": 773, + "time_per_iteration": 3.6480188369750977 + }, + { + "auxiliary_loss_clip": 0.01267639, + "auxiliary_loss_mlp": 0.01044624, + "balance_loss_clip": 1.07346332, + "balance_loss_mlp": 1.03404438, + "epoch": 0.0930679973546564, + "flos": 18259750627200.0, + "grad_norm": 3.2660310914056474, + "language_loss": 0.83281255, + "learning_rate": 3.958491418761124e-06, + "loss": 0.8559351, + "num_input_tokens_seen": 16548605, + "step": 774, + "time_per_iteration": 3.5431034564971924 + }, + { + "auxiliary_loss_clip": 0.01267173, + "auxiliary_loss_mlp": 0.01038775, + "balance_loss_clip": 1.06451726, + "balance_loss_mlp": 1.02798676, + "epoch": 0.0931882402452955, + "flos": 21099745405440.0, + "grad_norm": 2.1971544682600808, + "language_loss": 0.72998339, + "learning_rate": 3.958333390819258e-06, + "loss": 0.75304294, + "num_input_tokens_seen": 16565535, + "step": 775, + "time_per_iteration": 2.723348379135132 + }, + { + "auxiliary_loss_clip": 0.01257913, + "auxiliary_loss_mlp": 0.01039473, + "balance_loss_clip": 1.07508349, + "balance_loss_mlp": 1.02906024, + "epoch": 0.0933084831359346, + "flos": 24207275658240.0, + "grad_norm": 2.3516586107329345, + "language_loss": 0.80048758, + "learning_rate": 3.9581750657997754e-06, + "loss": 0.82346141, + "num_input_tokens_seen": 16584900, + "step": 776, + "time_per_iteration": 2.6633756160736084 + }, + { + "auxiliary_loss_clip": 0.01266569, + "auxiliary_loss_mlp": 0.01038033, + "balance_loss_clip": 1.0661025, + "balance_loss_mlp": 1.02804947, + "epoch": 0.09342872602657368, + "flos": 25480272637440.0, + "grad_norm": 1.7806333860526908, + "language_loss": 0.89555895, + "learning_rate": 3.95801644372669e-06, + "loss": 0.91860497, + "num_input_tokens_seen": 16604805, + "step": 777, + "time_per_iteration": 2.7262537479400635 + }, + { + "auxiliary_loss_clip": 0.01275745, + "auxiliary_loss_mlp": 0.01041309, + "balance_loss_clip": 1.06723213, + "balance_loss_mlp": 1.03079486, + "epoch": 0.09354896891721277, + "flos": 23149060053120.0, + "grad_norm": 2.32871421193343, + "language_loss": 0.84488702, + "learning_rate": 3.957857524624068e-06, + "loss": 0.86805749, + "num_input_tokens_seen": 16623685, + "step": 778, + "time_per_iteration": 2.7338666915893555 + }, + { + "auxiliary_loss_clip": 0.01270938, + "auxiliary_loss_mlp": 0.01036298, + "balance_loss_clip": 1.06773186, + "balance_loss_mlp": 1.02573657, + "epoch": 0.09366921180785186, + "flos": 24279563779200.0, + "grad_norm": 1.6146294054535553, + "language_loss": 0.89612746, + "learning_rate": 3.957698308516016e-06, + "loss": 0.91919982, + "num_input_tokens_seen": 16644985, + "step": 779, + "time_per_iteration": 2.7112298011779785 + }, + { + "auxiliary_loss_clip": 0.01265374, + "auxiliary_loss_mlp": 0.01381418, + "balance_loss_clip": 1.07661748, + "balance_loss_mlp": 0.99998569, + "epoch": 0.09378945469849095, + "flos": 18730036419840.0, + "grad_norm": 2.5455834550432774, + "language_loss": 0.82241309, + "learning_rate": 3.957538795426688e-06, + "loss": 0.84888101, + "num_input_tokens_seen": 16662410, + "step": 780, + "time_per_iteration": 2.7465457916259766 + }, + { + "auxiliary_loss_clip": 0.01270306, + "auxiliary_loss_mlp": 0.01034212, + "balance_loss_clip": 1.06852007, + "balance_loss_mlp": 1.02385247, + "epoch": 0.09390969758913004, + "flos": 23218834222080.0, + "grad_norm": 2.5104112645236256, + "language_loss": 0.77104819, + "learning_rate": 3.9573789853802804e-06, + "loss": 0.79409337, + "num_input_tokens_seen": 16680885, + "step": 781, + "time_per_iteration": 2.7014825344085693 + }, + { + "auxiliary_loss_clip": 0.01276197, + "auxiliary_loss_mlp": 0.0138123, + "balance_loss_clip": 1.07405221, + "balance_loss_mlp": 0.99992472, + "epoch": 0.09402994047976913, + "flos": 19646728439040.0, + "grad_norm": 3.4028495446811653, + "language_loss": 0.73983634, + "learning_rate": 3.957218878401037e-06, + "loss": 0.76641059, + "num_input_tokens_seen": 16699375, + "step": 782, + "time_per_iteration": 2.717731237411499 + }, + { + "auxiliary_loss_clip": 0.01262506, + "auxiliary_loss_mlp": 0.01043887, + "balance_loss_clip": 1.07716775, + "balance_loss_mlp": 1.03298569, + "epoch": 0.09415018337040823, + "flos": 29420463041280.0, + "grad_norm": 2.1349004725449254, + "language_loss": 0.89365971, + "learning_rate": 3.957058474513246e-06, + "loss": 0.91672367, + "num_input_tokens_seen": 16719230, + "step": 783, + "time_per_iteration": 2.6692938804626465 + }, + { + "auxiliary_loss_clip": 0.01265248, + "auxiliary_loss_mlp": 0.0104435, + "balance_loss_clip": 1.07534909, + "balance_loss_mlp": 1.0337944, + "epoch": 0.09427042626104731, + "flos": 24572092141440.0, + "grad_norm": 1.730344394715699, + "language_loss": 0.7841872, + "learning_rate": 3.956897773741241e-06, + "loss": 0.80728316, + "num_input_tokens_seen": 16738220, + "step": 784, + "time_per_iteration": 2.7035555839538574 + }, + { + "auxiliary_loss_clip": 0.0125549, + "auxiliary_loss_mlp": 0.01038916, + "balance_loss_clip": 1.06576812, + "balance_loss_mlp": 1.02811575, + "epoch": 0.09439066915168641, + "flos": 26359581576960.0, + "grad_norm": 1.7939136404546892, + "language_loss": 0.71738434, + "learning_rate": 3.956736776109398e-06, + "loss": 0.74032843, + "num_input_tokens_seen": 16759395, + "step": 785, + "time_per_iteration": 2.7430055141448975 + }, + { + "auxiliary_loss_clip": 0.01253218, + "auxiliary_loss_mlp": 0.01381638, + "balance_loss_clip": 1.06964946, + "balance_loss_mlp": 0.9999789, + "epoch": 0.09451091204232549, + "flos": 19427278296960.0, + "grad_norm": 2.3083564139319797, + "language_loss": 0.83760035, + "learning_rate": 3.956575481642143e-06, + "loss": 0.863949, + "num_input_tokens_seen": 16778285, + "step": 786, + "time_per_iteration": 2.6888253688812256 + }, + { + "auxiliary_loss_clip": 0.01278073, + "auxiliary_loss_mlp": 0.01039414, + "balance_loss_clip": 1.06076503, + "balance_loss_mlp": 1.02777863, + "epoch": 0.09463115493296459, + "flos": 25368051571200.0, + "grad_norm": 2.357753606674932, + "language_loss": 0.75043637, + "learning_rate": 3.956413890363943e-06, + "loss": 0.77361119, + "num_input_tokens_seen": 16795265, + "step": 787, + "time_per_iteration": 2.757305860519409 + }, + { + "auxiliary_loss_clip": 0.01263622, + "auxiliary_loss_mlp": 0.01041955, + "balance_loss_clip": 1.07346487, + "balance_loss_mlp": 1.03116691, + "epoch": 0.09475139782360369, + "flos": 10123254869760.0, + "grad_norm": 1.9380533613520943, + "language_loss": 0.81779003, + "learning_rate": 3.956252002299312e-06, + "loss": 0.84084582, + "num_input_tokens_seen": 16811165, + "step": 788, + "time_per_iteration": 2.770026445388794 + }, + { + "auxiliary_loss_clip": 0.01256621, + "auxiliary_loss_mlp": 0.01033792, + "balance_loss_clip": 1.07343292, + "balance_loss_mlp": 1.02274168, + "epoch": 0.09487164071424277, + "flos": 17231088936960.0, + "grad_norm": 2.185269063691456, + "language_loss": 0.90956634, + "learning_rate": 3.956089817472807e-06, + "loss": 0.9324705, + "num_input_tokens_seen": 16828470, + "step": 789, + "time_per_iteration": 2.65024995803833 + }, + { + "auxiliary_loss_clip": 0.01268251, + "auxiliary_loss_mlp": 0.01050836, + "balance_loss_clip": 1.06974638, + "balance_loss_mlp": 1.04061413, + "epoch": 0.09499188360488187, + "flos": 30849564528000.0, + "grad_norm": 3.210964836664397, + "language_loss": 0.85723239, + "learning_rate": 3.955927335909032e-06, + "loss": 0.88042331, + "num_input_tokens_seen": 16851680, + "step": 790, + "time_per_iteration": 2.804936408996582 + }, + { + "auxiliary_loss_clip": 0.01283542, + "auxiliary_loss_mlp": 0.01045333, + "balance_loss_clip": 1.06831908, + "balance_loss_mlp": 1.03494978, + "epoch": 0.09511212649552095, + "flos": 29351694453120.0, + "grad_norm": 2.61491219719387, + "language_loss": 0.76353335, + "learning_rate": 3.955764557632634e-06, + "loss": 0.78682208, + "num_input_tokens_seen": 16871490, + "step": 791, + "time_per_iteration": 2.842067241668701 + }, + { + "auxiliary_loss_clip": 0.01260937, + "auxiliary_loss_mlp": 0.01041306, + "balance_loss_clip": 1.06525385, + "balance_loss_mlp": 1.03116691, + "epoch": 0.09523236938616005, + "flos": 10378687461120.0, + "grad_norm": 2.1241428108031464, + "language_loss": 0.94689429, + "learning_rate": 3.955601482668309e-06, + "loss": 0.9699167, + "num_input_tokens_seen": 16889350, + "step": 792, + "time_per_iteration": 2.6997151374816895 + }, + { + "auxiliary_loss_clip": 0.01275008, + "auxiliary_loss_mlp": 0.01046278, + "balance_loss_clip": 1.05723047, + "balance_loss_mlp": 1.03513801, + "epoch": 0.09535261227679913, + "flos": 19061815368960.0, + "grad_norm": 1.7583655161731613, + "language_loss": 0.88450521, + "learning_rate": 3.955438111040794e-06, + "loss": 0.90771806, + "num_input_tokens_seen": 16907625, + "step": 793, + "time_per_iteration": 2.728954792022705 + }, + { + "auxiliary_loss_clip": 0.01279634, + "auxiliary_loss_mlp": 0.01032438, + "balance_loss_clip": 1.06375718, + "balance_loss_mlp": 1.0216732, + "epoch": 0.09547285516743823, + "flos": 20922993555840.0, + "grad_norm": 2.5642383974443623, + "language_loss": 0.80543554, + "learning_rate": 3.955274442774873e-06, + "loss": 0.8285563, + "num_input_tokens_seen": 16926205, + "step": 794, + "time_per_iteration": 2.7385149002075195 + }, + { + "auxiliary_loss_clip": 0.01258034, + "auxiliary_loss_mlp": 0.01034395, + "balance_loss_clip": 1.06654847, + "balance_loss_mlp": 1.02335048, + "epoch": 0.09559309805807732, + "flos": 30154405639680.0, + "grad_norm": 2.1929202072539127, + "language_loss": 0.70586193, + "learning_rate": 3.9551104778953725e-06, + "loss": 0.72878617, + "num_input_tokens_seen": 16946500, + "step": 795, + "time_per_iteration": 2.7559826374053955 + }, + { + "auxiliary_loss_clip": 0.01278674, + "auxiliary_loss_mlp": 0.01035131, + "balance_loss_clip": 1.06593871, + "balance_loss_mlp": 1.02482486, + "epoch": 0.0957133409487164, + "flos": 21066743784960.0, + "grad_norm": 1.7770198486217244, + "language_loss": 0.85128534, + "learning_rate": 3.954946216427167e-06, + "loss": 0.87442338, + "num_input_tokens_seen": 16966960, + "step": 796, + "time_per_iteration": 2.721555709838867 + }, + { + "auxiliary_loss_clip": 0.01163566, + "auxiliary_loss_mlp": 0.01004121, + "balance_loss_clip": 1.02578259, + "balance_loss_mlp": 1.00129592, + "epoch": 0.0958335838393555, + "flos": 71297979315840.0, + "grad_norm": 0.8732482582769587, + "language_loss": 0.6155678, + "learning_rate": 3.954781658395176e-06, + "loss": 0.6372447, + "num_input_tokens_seen": 17023215, + "step": 797, + "time_per_iteration": 4.079889535903931 + }, + { + "auxiliary_loss_clip": 0.01276108, + "auxiliary_loss_mlp": 0.01048837, + "balance_loss_clip": 1.06729627, + "balance_loss_mlp": 1.03754771, + "epoch": 0.09595382672999458, + "flos": 21872974504320.0, + "grad_norm": 2.8973921432746628, + "language_loss": 0.92156482, + "learning_rate": 3.95461680382436e-06, + "loss": 0.94481426, + "num_input_tokens_seen": 17042140, + "step": 798, + "time_per_iteration": 3.5963377952575684 + }, + { + "auxiliary_loss_clip": 0.01266205, + "auxiliary_loss_mlp": 0.01040869, + "balance_loss_clip": 1.07486987, + "balance_loss_mlp": 1.02892435, + "epoch": 0.09607406962063368, + "flos": 18695562341760.0, + "grad_norm": 3.1968095640109064, + "language_loss": 0.85892618, + "learning_rate": 3.9544516527397295e-06, + "loss": 0.88199687, + "num_input_tokens_seen": 17058490, + "step": 799, + "time_per_iteration": 4.515854358673096 + }, + { + "auxiliary_loss_clip": 0.01251434, + "auxiliary_loss_mlp": 0.01042992, + "balance_loss_clip": 1.06586313, + "balance_loss_mlp": 1.03191149, + "epoch": 0.09619431251127276, + "flos": 22568456615040.0, + "grad_norm": 2.829198144703049, + "language_loss": 0.8037715, + "learning_rate": 3.954286205166338e-06, + "loss": 0.82671571, + "num_input_tokens_seen": 17079655, + "step": 800, + "time_per_iteration": 2.7049880027770996 + }, + { + "auxiliary_loss_clip": 0.01268563, + "auxiliary_loss_mlp": 0.01043868, + "balance_loss_clip": 1.07613218, + "balance_loss_mlp": 1.03275716, + "epoch": 0.09631455540191186, + "flos": 14246230608000.0, + "grad_norm": 2.5689744080616994, + "language_loss": 0.83861047, + "learning_rate": 3.954120461129282e-06, + "loss": 0.86173481, + "num_input_tokens_seen": 17097065, + "step": 801, + "time_per_iteration": 2.6491847038269043 + }, + { + "auxiliary_loss_clip": 0.01261474, + "auxiliary_loss_mlp": 0.01041973, + "balance_loss_clip": 1.07828903, + "balance_loss_mlp": 1.03147697, + "epoch": 0.09643479829255096, + "flos": 20740387789440.0, + "grad_norm": 2.1288479024187357, + "language_loss": 0.83700323, + "learning_rate": 3.953954420653706e-06, + "loss": 0.86003768, + "num_input_tokens_seen": 17114090, + "step": 802, + "time_per_iteration": 2.6190898418426514 + }, + { + "auxiliary_loss_clip": 0.01267768, + "auxiliary_loss_mlp": 0.01045517, + "balance_loss_clip": 1.0749346, + "balance_loss_mlp": 1.03448415, + "epoch": 0.09655504118319004, + "flos": 24420476833920.0, + "grad_norm": 1.9221167582716103, + "language_loss": 0.88304329, + "learning_rate": 3.953788083764798e-06, + "loss": 0.90617615, + "num_input_tokens_seen": 17133325, + "step": 803, + "time_per_iteration": 2.7077274322509766 + }, + { + "auxiliary_loss_clip": 0.0128615, + "auxiliary_loss_mlp": 0.01034973, + "balance_loss_clip": 1.06713772, + "balance_loss_mlp": 1.02454185, + "epoch": 0.09667528407382914, + "flos": 18441961344000.0, + "grad_norm": 1.8282159013775727, + "language_loss": 0.92110342, + "learning_rate": 3.953621450487792e-06, + "loss": 0.94431466, + "num_input_tokens_seen": 17151945, + "step": 804, + "time_per_iteration": 2.6619558334350586 + }, + { + "auxiliary_loss_clip": 0.01135352, + "auxiliary_loss_mlp": 0.010014, + "balance_loss_clip": 1.02874923, + "balance_loss_mlp": 0.99839556, + "epoch": 0.09679552696446822, + "flos": 70816455544320.0, + "grad_norm": 0.8751751168082863, + "language_loss": 0.61215997, + "learning_rate": 3.953454520847964e-06, + "loss": 0.63352752, + "num_input_tokens_seen": 17216790, + "step": 805, + "time_per_iteration": 3.3609158992767334 + }, + { + "auxiliary_loss_clip": 0.01262373, + "auxiliary_loss_mlp": 0.01046954, + "balance_loss_clip": 1.06641662, + "balance_loss_mlp": 1.03386497, + "epoch": 0.09691576985510732, + "flos": 21945514020480.0, + "grad_norm": 2.1510389327530404, + "language_loss": 0.73803437, + "learning_rate": 3.9532872948706395e-06, + "loss": 0.76112765, + "num_input_tokens_seen": 17236285, + "step": 806, + "time_per_iteration": 2.7237436771392822 + }, + { + "auxiliary_loss_clip": 0.01271882, + "auxiliary_loss_mlp": 0.01041129, + "balance_loss_clip": 1.06857944, + "balance_loss_mlp": 1.02982783, + "epoch": 0.09703601274574641, + "flos": 17965211103360.0, + "grad_norm": 6.957718637586831, + "language_loss": 0.8282088, + "learning_rate": 3.9531197725811845e-06, + "loss": 0.85133886, + "num_input_tokens_seen": 17251670, + "step": 807, + "time_per_iteration": 2.6541943550109863 + }, + { + "auxiliary_loss_clip": 0.01256613, + "auxiliary_loss_mlp": 0.01042989, + "balance_loss_clip": 1.07651031, + "balance_loss_mlp": 1.03270769, + "epoch": 0.0971562556363855, + "flos": 22162162901760.0, + "grad_norm": 2.8295576407147336, + "language_loss": 0.87737793, + "learning_rate": 3.952951954005013e-06, + "loss": 0.90037394, + "num_input_tokens_seen": 17271355, + "step": 808, + "time_per_iteration": 2.6537680625915527 + }, + { + "auxiliary_loss_clip": 0.01265838, + "auxiliary_loss_mlp": 0.01048768, + "balance_loss_clip": 1.06437588, + "balance_loss_mlp": 1.03797317, + "epoch": 0.0972764985270246, + "flos": 25848716394240.0, + "grad_norm": 1.686042288218959, + "language_loss": 0.84891129, + "learning_rate": 3.952783839167584e-06, + "loss": 0.87205738, + "num_input_tokens_seen": 17291400, + "step": 809, + "time_per_iteration": 2.746051549911499 + }, + { + "auxiliary_loss_clip": 0.01260836, + "auxiliary_loss_mlp": 0.01043475, + "balance_loss_clip": 1.07037258, + "balance_loss_mlp": 1.03241849, + "epoch": 0.09739674141766368, + "flos": 20339373375360.0, + "grad_norm": 4.429252567250951, + "language_loss": 0.74280572, + "learning_rate": 3.952615428094398e-06, + "loss": 0.76584882, + "num_input_tokens_seen": 17310920, + "step": 810, + "time_per_iteration": 2.715937614440918 + }, + { + "auxiliary_loss_clip": 0.01269444, + "auxiliary_loss_mlp": 0.0104653, + "balance_loss_clip": 1.06056929, + "balance_loss_mlp": 1.0354495, + "epoch": 0.09751698430830277, + "flos": 15743059188480.0, + "grad_norm": 2.2795011088167434, + "language_loss": 0.7362777, + "learning_rate": 3.952446720811004e-06, + "loss": 0.7594375, + "num_input_tokens_seen": 17329245, + "step": 811, + "time_per_iteration": 2.687995195388794 + }, + { + "auxiliary_loss_clip": 0.01157801, + "auxiliary_loss_mlp": 0.01001591, + "balance_loss_clip": 1.02446008, + "balance_loss_mlp": 0.99856341, + "epoch": 0.09763722719894186, + "flos": 63716806800000.0, + "grad_norm": 0.8371833132609219, + "language_loss": 0.63621366, + "learning_rate": 3.952277717342995e-06, + "loss": 0.65780759, + "num_input_tokens_seen": 17395680, + "step": 812, + "time_per_iteration": 3.3632519245147705 + }, + { + "auxiliary_loss_clip": 0.01273646, + "auxiliary_loss_mlp": 0.01040538, + "balance_loss_clip": 1.06792736, + "balance_loss_mlp": 1.02943349, + "epoch": 0.09775747008958095, + "flos": 22090916275200.0, + "grad_norm": 1.9223033806859926, + "language_loss": 0.85391176, + "learning_rate": 3.952108417716009e-06, + "loss": 0.87705362, + "num_input_tokens_seen": 17415135, + "step": 813, + "time_per_iteration": 2.714770793914795 + }, + { + "auxiliary_loss_clip": 0.01260065, + "auxiliary_loss_mlp": 0.01039475, + "balance_loss_clip": 1.07266688, + "balance_loss_mlp": 1.0279479, + "epoch": 0.09787771298022005, + "flos": 21286050272640.0, + "grad_norm": 2.1242013133456266, + "language_loss": 0.84707773, + "learning_rate": 3.951938821955727e-06, + "loss": 0.87007314, + "num_input_tokens_seen": 17434535, + "step": 814, + "time_per_iteration": 2.640007972717285 + }, + { + "auxiliary_loss_clip": 0.01263621, + "auxiliary_loss_mlp": 0.01035073, + "balance_loss_clip": 1.06779623, + "balance_loss_mlp": 1.02384365, + "epoch": 0.09799795587085913, + "flos": 22054574689920.0, + "grad_norm": 1.6099935551515854, + "language_loss": 0.7646513, + "learning_rate": 3.9517689300878786e-06, + "loss": 0.78763831, + "num_input_tokens_seen": 17454270, + "step": 815, + "time_per_iteration": 2.757657051086426 + }, + { + "auxiliary_loss_clip": 0.01253997, + "auxiliary_loss_mlp": 0.01047022, + "balance_loss_clip": 1.07209718, + "balance_loss_mlp": 1.03615582, + "epoch": 0.09811819876149823, + "flos": 22163743100160.0, + "grad_norm": 1.7919273383562389, + "language_loss": 0.78586268, + "learning_rate": 3.951598742138236e-06, + "loss": 0.80887288, + "num_input_tokens_seen": 17472995, + "step": 816, + "time_per_iteration": 2.6263532638549805 + }, + { + "auxiliary_loss_clip": 0.01269019, + "auxiliary_loss_mlp": 0.01035567, + "balance_loss_clip": 1.06389797, + "balance_loss_mlp": 1.02483785, + "epoch": 0.09823844165213731, + "flos": 22231111057920.0, + "grad_norm": 2.1801973253635922, + "language_loss": 0.79556644, + "learning_rate": 3.951428258132615e-06, + "loss": 0.81861228, + "num_input_tokens_seen": 17491115, + "step": 817, + "time_per_iteration": 2.7390787601470947 + }, + { + "auxiliary_loss_clip": 0.01273045, + "auxiliary_loss_mlp": 0.01040082, + "balance_loss_clip": 1.06994796, + "balance_loss_mlp": 1.02877498, + "epoch": 0.09835868454277641, + "flos": 22487728798080.0, + "grad_norm": 1.776477416037337, + "language_loss": 0.84253293, + "learning_rate": 3.951257478096879e-06, + "loss": 0.86566412, + "num_input_tokens_seen": 17509480, + "step": 818, + "time_per_iteration": 2.706880807876587 + }, + { + "auxiliary_loss_clip": 0.01272568, + "auxiliary_loss_mlp": 0.01382081, + "balance_loss_clip": 1.06921041, + "balance_loss_mlp": 1.0001204, + "epoch": 0.0984789274334155, + "flos": 16362554077440.0, + "grad_norm": 2.8762324977174516, + "language_loss": 0.68911278, + "learning_rate": 3.951086402056936e-06, + "loss": 0.71565932, + "num_input_tokens_seen": 17524080, + "step": 819, + "time_per_iteration": 2.718003034591675 + }, + { + "auxiliary_loss_clip": 0.01290994, + "auxiliary_loss_mlp": 0.01381842, + "balance_loss_clip": 1.06191313, + "balance_loss_mlp": 1.00009346, + "epoch": 0.09859917032405459, + "flos": 24243545416320.0, + "grad_norm": 2.149533985725929, + "language_loss": 0.83631968, + "learning_rate": 3.950915030038735e-06, + "loss": 0.86304808, + "num_input_tokens_seen": 17543875, + "step": 820, + "time_per_iteration": 2.893578052520752 + }, + { + "auxiliary_loss_clip": 0.01259869, + "auxiliary_loss_mlp": 0.01045854, + "balance_loss_clip": 1.07076788, + "balance_loss_mlp": 1.03500032, + "epoch": 0.09871941321469369, + "flos": 17420195064960.0, + "grad_norm": 2.2767423395215824, + "language_loss": 0.83619344, + "learning_rate": 3.9507433620682765e-06, + "loss": 0.85925066, + "num_input_tokens_seen": 17560810, + "step": 821, + "time_per_iteration": 2.8442482948303223 + }, + { + "auxiliary_loss_clip": 0.01273508, + "auxiliary_loss_mlp": 0.01045431, + "balance_loss_clip": 1.06415713, + "balance_loss_mlp": 1.03399861, + "epoch": 0.09883965610533277, + "flos": 28477341590400.0, + "grad_norm": 2.0182639344351903, + "language_loss": 0.88345504, + "learning_rate": 3.9505713981716e-06, + "loss": 0.90664446, + "num_input_tokens_seen": 17583640, + "step": 822, + "time_per_iteration": 3.7563037872314453 + }, + { + "auxiliary_loss_clip": 0.01269132, + "auxiliary_loss_mlp": 0.01038656, + "balance_loss_clip": 1.0717659, + "balance_loss_mlp": 1.02776599, + "epoch": 0.09895989899597187, + "flos": 23693932437120.0, + "grad_norm": 3.4748297216865724, + "language_loss": 0.81292576, + "learning_rate": 3.950399138374795e-06, + "loss": 0.8360036, + "num_input_tokens_seen": 17602720, + "step": 823, + "time_per_iteration": 2.6902894973754883 + }, + { + "auxiliary_loss_clip": 0.01259359, + "auxiliary_loss_mlp": 0.01043401, + "balance_loss_clip": 1.06915379, + "balance_loss_mlp": 1.03282106, + "epoch": 0.09908014188661095, + "flos": 24679608526080.0, + "grad_norm": 1.7405262255222995, + "language_loss": 0.74390352, + "learning_rate": 3.95022658270399e-06, + "loss": 0.76693112, + "num_input_tokens_seen": 17623085, + "step": 824, + "time_per_iteration": 3.6179234981536865 + }, + { + "auxiliary_loss_clip": 0.01261567, + "auxiliary_loss_mlp": 0.01043394, + "balance_loss_clip": 1.06702232, + "balance_loss_mlp": 1.03184831, + "epoch": 0.09920038477725004, + "flos": 14064307200000.0, + "grad_norm": 1.836535596919007, + "language_loss": 0.78320646, + "learning_rate": 3.9500537311853635e-06, + "loss": 0.80625606, + "num_input_tokens_seen": 17641040, + "step": 825, + "time_per_iteration": 3.658647060394287 + }, + { + "auxiliary_loss_clip": 0.01258175, + "auxiliary_loss_mlp": 0.01046281, + "balance_loss_clip": 1.06711459, + "balance_loss_mlp": 1.03518307, + "epoch": 0.09932062766788914, + "flos": 13407070095360.0, + "grad_norm": 6.204833895853051, + "language_loss": 0.83622974, + "learning_rate": 3.949880583845136e-06, + "loss": 0.85927427, + "num_input_tokens_seen": 17659115, + "step": 826, + "time_per_iteration": 2.626492977142334 + }, + { + "auxiliary_loss_clip": 0.01266906, + "auxiliary_loss_mlp": 0.01031161, + "balance_loss_clip": 1.06944168, + "balance_loss_mlp": 1.01996756, + "epoch": 0.09944087055852822, + "flos": 19500751566720.0, + "grad_norm": 1.856257042347427, + "language_loss": 0.81288731, + "learning_rate": 3.949707140709575e-06, + "loss": 0.835868, + "num_input_tokens_seen": 17678845, + "step": 827, + "time_per_iteration": 2.6937332153320312 + }, + { + "auxiliary_loss_clip": 0.01260806, + "auxiliary_loss_mlp": 0.01052682, + "balance_loss_clip": 1.06809485, + "balance_loss_mlp": 1.04046869, + "epoch": 0.09956111344916732, + "flos": 17749100926080.0, + "grad_norm": 4.798361168953554, + "language_loss": 0.83570206, + "learning_rate": 3.949533401804991e-06, + "loss": 0.85883695, + "num_input_tokens_seen": 17695750, + "step": 828, + "time_per_iteration": 2.6031384468078613 + }, + { + "auxiliary_loss_clip": 0.01260318, + "auxiliary_loss_mlp": 0.01381784, + "balance_loss_clip": 1.07032692, + "balance_loss_mlp": 1.00008214, + "epoch": 0.0996813563398064, + "flos": 17967581400960.0, + "grad_norm": 1.9155407551141033, + "language_loss": 0.90691471, + "learning_rate": 3.949359367157739e-06, + "loss": 0.93333578, + "num_input_tokens_seen": 17714445, + "step": 829, + "time_per_iteration": 2.6590731143951416 + }, + { + "auxiliary_loss_clip": 0.01268514, + "auxiliary_loss_mlp": 0.01044666, + "balance_loss_clip": 1.07104897, + "balance_loss_mlp": 1.03300738, + "epoch": 0.0998015992304455, + "flos": 17457039440640.0, + "grad_norm": 2.9700342022937543, + "language_loss": 0.7538172, + "learning_rate": 3.949185036794222e-06, + "loss": 0.77694905, + "num_input_tokens_seen": 17732455, + "step": 830, + "time_per_iteration": 2.6086459159851074 + }, + { + "auxiliary_loss_clip": 0.01249574, + "auxiliary_loss_mlp": 0.01044004, + "balance_loss_clip": 1.07008517, + "balance_loss_mlp": 1.03250647, + "epoch": 0.0999218421210846, + "flos": 25888757080320.0, + "grad_norm": 1.559974405846854, + "language_loss": 0.78980774, + "learning_rate": 3.949010410740884e-06, + "loss": 0.81274354, + "num_input_tokens_seen": 17755280, + "step": 831, + "time_per_iteration": 2.72861647605896 + }, + { + "auxiliary_loss_clip": 0.01252524, + "auxiliary_loss_mlp": 0.01381927, + "balance_loss_clip": 1.06290185, + "balance_loss_mlp": 1.00005341, + "epoch": 0.10004208501172368, + "flos": 21215916967680.0, + "grad_norm": 1.6966446869041927, + "language_loss": 0.86192918, + "learning_rate": 3.948835489024216e-06, + "loss": 0.88827366, + "num_input_tokens_seen": 17775015, + "step": 832, + "time_per_iteration": 2.674177885055542 + }, + { + "auxiliary_loss_clip": 0.01259499, + "auxiliary_loss_mlp": 0.01035332, + "balance_loss_clip": 1.06742716, + "balance_loss_mlp": 1.0247401, + "epoch": 0.10016232790236278, + "flos": 17348409734400.0, + "grad_norm": 2.36440150990516, + "language_loss": 0.90472925, + "learning_rate": 3.948660271670755e-06, + "loss": 0.92767751, + "num_input_tokens_seen": 17792165, + "step": 833, + "time_per_iteration": 2.666325330734253 + }, + { + "auxiliary_loss_clip": 0.01261901, + "auxiliary_loss_mlp": 0.01043088, + "balance_loss_clip": 1.06690061, + "balance_loss_mlp": 1.03161454, + "epoch": 0.10028257079300186, + "flos": 25666541591040.0, + "grad_norm": 2.542371289097284, + "language_loss": 0.84309924, + "learning_rate": 3.948484758707079e-06, + "loss": 0.86614907, + "num_input_tokens_seen": 17811765, + "step": 834, + "time_per_iteration": 2.7097980976104736 + }, + { + "auxiliary_loss_clip": 0.01262331, + "auxiliary_loss_mlp": 0.01044889, + "balance_loss_clip": 1.06081676, + "balance_loss_mlp": 1.03348625, + "epoch": 0.10040281368364096, + "flos": 25156035544320.0, + "grad_norm": 2.1045040790614453, + "language_loss": 0.83565807, + "learning_rate": 3.948308950159815e-06, + "loss": 0.85873032, + "num_input_tokens_seen": 17830445, + "step": 835, + "time_per_iteration": 2.700568199157715 + }, + { + "auxiliary_loss_clip": 0.01265548, + "auxiliary_loss_mlp": 0.01042235, + "balance_loss_clip": 1.06084323, + "balance_loss_mlp": 1.02951598, + "epoch": 0.10052305657428004, + "flos": 17603303621760.0, + "grad_norm": 2.5213307957377458, + "language_loss": 0.75880218, + "learning_rate": 3.9481328460556326e-06, + "loss": 0.78188008, + "num_input_tokens_seen": 17847665, + "step": 836, + "time_per_iteration": 2.7336771488189697 + }, + { + "auxiliary_loss_clip": 0.01256537, + "auxiliary_loss_mlp": 0.01047063, + "balance_loss_clip": 1.0636363, + "balance_loss_mlp": 1.035393, + "epoch": 0.10064329946491914, + "flos": 18660154510080.0, + "grad_norm": 2.3997821760626685, + "language_loss": 0.89358771, + "learning_rate": 3.9479564464212455e-06, + "loss": 0.91662371, + "num_input_tokens_seen": 17866825, + "step": 837, + "time_per_iteration": 2.6653454303741455 + }, + { + "auxiliary_loss_clip": 0.01256748, + "auxiliary_loss_mlp": 0.01043638, + "balance_loss_clip": 1.07066655, + "balance_loss_mlp": 1.03233743, + "epoch": 0.10076354235555823, + "flos": 17199056983680.0, + "grad_norm": 2.479100710546662, + "language_loss": 0.76479101, + "learning_rate": 3.947779751283414e-06, + "loss": 0.78779483, + "num_input_tokens_seen": 17883995, + "step": 838, + "time_per_iteration": 2.660522937774658 + }, + { + "auxiliary_loss_clip": 0.0126152, + "auxiliary_loss_mlp": 0.01382152, + "balance_loss_clip": 1.07413936, + "balance_loss_mlp": 1.00007176, + "epoch": 0.10088378524619732, + "flos": 22962252395520.0, + "grad_norm": 1.8272565519858386, + "language_loss": 0.75924706, + "learning_rate": 3.947602760668944e-06, + "loss": 0.78568387, + "num_input_tokens_seen": 17903785, + "step": 839, + "time_per_iteration": 2.6775152683258057 + }, + { + "auxiliary_loss_clip": 0.01260579, + "auxiliary_loss_mlp": 0.01046013, + "balance_loss_clip": 1.07313669, + "balance_loss_mlp": 1.03535581, + "epoch": 0.10100402813683641, + "flos": 37885828746240.0, + "grad_norm": 1.8394781476876665, + "language_loss": 0.71493125, + "learning_rate": 3.947425474604684e-06, + "loss": 0.73799711, + "num_input_tokens_seen": 17927720, + "step": 840, + "time_per_iteration": 2.8059773445129395 + }, + { + "auxiliary_loss_clip": 0.01265822, + "auxiliary_loss_mlp": 0.01045617, + "balance_loss_clip": 1.06611538, + "balance_loss_mlp": 1.03446507, + "epoch": 0.1011242710274755, + "flos": 21543458112000.0, + "grad_norm": 2.805282851197344, + "language_loss": 0.92293715, + "learning_rate": 3.947247893117528e-06, + "loss": 0.94605154, + "num_input_tokens_seen": 17946225, + "step": 841, + "time_per_iteration": 2.660198211669922 + }, + { + "auxiliary_loss_clip": 0.01250261, + "auxiliary_loss_mlp": 0.01045229, + "balance_loss_clip": 1.06554389, + "balance_loss_mlp": 1.03330171, + "epoch": 0.10124451391811459, + "flos": 13621456419840.0, + "grad_norm": 3.495060560517996, + "language_loss": 0.69298029, + "learning_rate": 3.947070016234413e-06, + "loss": 0.71593511, + "num_input_tokens_seen": 17962015, + "step": 842, + "time_per_iteration": 2.659862518310547 + }, + { + "auxiliary_loss_clip": 0.0127696, + "auxiliary_loss_mlp": 0.01046639, + "balance_loss_clip": 1.06928921, + "balance_loss_mlp": 1.03533792, + "epoch": 0.10136475680875369, + "flos": 16649228522880.0, + "grad_norm": 2.2885297790208425, + "language_loss": 0.7480973, + "learning_rate": 3.946891843982326e-06, + "loss": 0.77133334, + "num_input_tokens_seen": 17979680, + "step": 843, + "time_per_iteration": 2.6850411891937256 + }, + { + "auxiliary_loss_clip": 0.01258342, + "auxiliary_loss_mlp": 0.01046414, + "balance_loss_clip": 1.06913555, + "balance_loss_mlp": 1.0340755, + "epoch": 0.10148499969939277, + "flos": 19461034103040.0, + "grad_norm": 1.964106711209104, + "language_loss": 0.74405771, + "learning_rate": 3.9467133763882935e-06, + "loss": 0.76710528, + "num_input_tokens_seen": 17998145, + "step": 844, + "time_per_iteration": 2.6363353729248047 + }, + { + "auxiliary_loss_clip": 0.01244059, + "auxiliary_loss_mlp": 0.01044794, + "balance_loss_clip": 1.06625843, + "balance_loss_mlp": 1.03295052, + "epoch": 0.10160524259003187, + "flos": 21104988791040.0, + "grad_norm": 2.9522221130623714, + "language_loss": 0.86132932, + "learning_rate": 3.9465346134793905e-06, + "loss": 0.88421786, + "num_input_tokens_seen": 18017955, + "step": 845, + "time_per_iteration": 2.6545145511627197 + }, + { + "auxiliary_loss_clip": 0.01276987, + "auxiliary_loss_mlp": 0.01048115, + "balance_loss_clip": 1.06849754, + "balance_loss_mlp": 1.03735042, + "epoch": 0.10172548548067095, + "flos": 17712687513600.0, + "grad_norm": 2.0932352823446925, + "language_loss": 0.79480708, + "learning_rate": 3.9463555552827335e-06, + "loss": 0.81805813, + "num_input_tokens_seen": 18035125, + "step": 846, + "time_per_iteration": 2.8514316082000732 + }, + { + "auxiliary_loss_clip": 0.01245859, + "auxiliary_loss_mlp": 0.01043467, + "balance_loss_clip": 1.06764674, + "balance_loss_mlp": 1.03317928, + "epoch": 0.10184572837131005, + "flos": 21104845136640.0, + "grad_norm": 3.098912815080973, + "language_loss": 0.86222124, + "learning_rate": 3.946176201825487e-06, + "loss": 0.88511455, + "num_input_tokens_seen": 18053160, + "step": 847, + "time_per_iteration": 2.6724343299865723 + }, + { + "auxiliary_loss_clip": 0.01268965, + "auxiliary_loss_mlp": 0.01040261, + "balance_loss_clip": 1.06960464, + "balance_loss_mlp": 1.02870953, + "epoch": 0.10196597126194913, + "flos": 26067591918720.0, + "grad_norm": 1.8513949015506337, + "language_loss": 0.83607209, + "learning_rate": 3.9459965531348575e-06, + "loss": 0.8591643, + "num_input_tokens_seen": 18072815, + "step": 848, + "time_per_iteration": 2.716082811355591 + }, + { + "auxiliary_loss_clip": 0.01271563, + "auxiliary_loss_mlp": 0.01381918, + "balance_loss_clip": 1.06843591, + "balance_loss_mlp": 1.0001955, + "epoch": 0.10208621415258823, + "flos": 29314634595840.0, + "grad_norm": 2.1345290351789252, + "language_loss": 0.85291898, + "learning_rate": 3.945816609238098e-06, + "loss": 0.87945378, + "num_input_tokens_seen": 18092225, + "step": 849, + "time_per_iteration": 3.68448543548584 + }, + { + "auxiliary_loss_clip": 0.01270229, + "auxiliary_loss_mlp": 0.01048822, + "balance_loss_clip": 1.06310606, + "balance_loss_mlp": 1.03644216, + "epoch": 0.10220645704322733, + "flos": 23805794367360.0, + "grad_norm": 2.3858700822845758, + "language_loss": 0.85309047, + "learning_rate": 3.945636370162507e-06, + "loss": 0.87628102, + "num_input_tokens_seen": 18112335, + "step": 850, + "time_per_iteration": 3.7249515056610107 + }, + { + "auxiliary_loss_clip": 0.01255389, + "auxiliary_loss_mlp": 0.01035725, + "balance_loss_clip": 1.06751108, + "balance_loss_mlp": 1.02477562, + "epoch": 0.10232669993386641, + "flos": 23218546913280.0, + "grad_norm": 1.722928240621308, + "language_loss": 0.79065198, + "learning_rate": 3.945455835935425e-06, + "loss": 0.81356311, + "num_input_tokens_seen": 18131520, + "step": 851, + "time_per_iteration": 3.6169731616973877 + }, + { + "auxiliary_loss_clip": 0.01264791, + "auxiliary_loss_mlp": 0.01039102, + "balance_loss_clip": 1.06474185, + "balance_loss_mlp": 1.0275563, + "epoch": 0.1024469428245055, + "flos": 22922929981440.0, + "grad_norm": 2.490852296560856, + "language_loss": 0.75026762, + "learning_rate": 3.94527500658424e-06, + "loss": 0.77330661, + "num_input_tokens_seen": 18149185, + "step": 852, + "time_per_iteration": 2.7478106021881104 + }, + { + "auxiliary_loss_clip": 0.01280517, + "auxiliary_loss_mlp": 0.01039768, + "balance_loss_clip": 1.06256223, + "balance_loss_mlp": 1.02896214, + "epoch": 0.10256718571514459, + "flos": 31359495957120.0, + "grad_norm": 2.731478306272976, + "language_loss": 0.81097466, + "learning_rate": 3.945093882136382e-06, + "loss": 0.83417749, + "num_input_tokens_seen": 18172960, + "step": 853, + "time_per_iteration": 2.8127009868621826 + }, + { + "auxiliary_loss_clip": 0.01264696, + "auxiliary_loss_mlp": 0.01381682, + "balance_loss_clip": 1.06744969, + "balance_loss_mlp": 1.00021458, + "epoch": 0.10268742860578368, + "flos": 23474877344640.0, + "grad_norm": 1.9303570663127871, + "language_loss": 0.8468017, + "learning_rate": 3.944912462619329e-06, + "loss": 0.87326545, + "num_input_tokens_seen": 18191925, + "step": 854, + "time_per_iteration": 2.7793755531311035 + }, + { + "auxiliary_loss_clip": 0.01266423, + "auxiliary_loss_mlp": 0.01044391, + "balance_loss_clip": 1.06545341, + "balance_loss_mlp": 1.03329253, + "epoch": 0.10280767149642277, + "flos": 25520313323520.0, + "grad_norm": 1.9030385883114762, + "language_loss": 0.80513036, + "learning_rate": 3.9447307480606025e-06, + "loss": 0.82823849, + "num_input_tokens_seen": 18212010, + "step": 855, + "time_per_iteration": 2.7391369342803955 + }, + { + "auxiliary_loss_clip": 0.01255727, + "auxiliary_loss_mlp": 0.0104189, + "balance_loss_clip": 1.06555831, + "balance_loss_mlp": 1.0300405, + "epoch": 0.10292791438706186, + "flos": 17347691462400.0, + "grad_norm": 1.940753126979139, + "language_loss": 0.90023172, + "learning_rate": 3.944548738487767e-06, + "loss": 0.92320788, + "num_input_tokens_seen": 18229525, + "step": 856, + "time_per_iteration": 2.7022886276245117 + }, + { + "auxiliary_loss_clip": 0.01255503, + "auxiliary_loss_mlp": 0.01045918, + "balance_loss_clip": 1.07266712, + "balance_loss_mlp": 1.03522539, + "epoch": 0.10304815727770096, + "flos": 27052693390080.0, + "grad_norm": 1.9653499699995784, + "language_loss": 0.90910232, + "learning_rate": 3.944366433928434e-06, + "loss": 0.93211651, + "num_input_tokens_seen": 18249505, + "step": 857, + "time_per_iteration": 2.68328857421875 + }, + { + "auxiliary_loss_clip": 0.01257104, + "auxiliary_loss_mlp": 0.01046756, + "balance_loss_clip": 1.06396198, + "balance_loss_mlp": 1.03649223, + "epoch": 0.10316840016834004, + "flos": 22782591544320.0, + "grad_norm": 1.475390875823502, + "language_loss": 0.8367635, + "learning_rate": 3.9441838344102594e-06, + "loss": 0.85980207, + "num_input_tokens_seen": 18269230, + "step": 858, + "time_per_iteration": 2.699615478515625 + }, + { + "auxiliary_loss_clip": 0.0126862, + "auxiliary_loss_mlp": 0.01043831, + "balance_loss_clip": 1.06742454, + "balance_loss_mlp": 1.03297114, + "epoch": 0.10328864305897914, + "flos": 20704584908160.0, + "grad_norm": 2.3069720942225396, + "language_loss": 0.67142051, + "learning_rate": 3.944000939960943e-06, + "loss": 0.69454503, + "num_input_tokens_seen": 18287955, + "step": 859, + "time_per_iteration": 2.7029082775115967 + }, + { + "auxiliary_loss_clip": 0.01258006, + "auxiliary_loss_mlp": 0.01038151, + "balance_loss_clip": 1.06629217, + "balance_loss_mlp": 1.02774954, + "epoch": 0.10340888594961822, + "flos": 28478814048000.0, + "grad_norm": 1.791838040555011, + "language_loss": 0.79669595, + "learning_rate": 3.943817750608229e-06, + "loss": 0.81965756, + "num_input_tokens_seen": 18310505, + "step": 860, + "time_per_iteration": 2.7225420475006104 + }, + { + "auxiliary_loss_clip": 0.01264519, + "auxiliary_loss_mlp": 0.010383, + "balance_loss_clip": 1.07047236, + "balance_loss_mlp": 1.02745795, + "epoch": 0.10352912884025732, + "flos": 13370333460480.0, + "grad_norm": 2.599571455022866, + "language_loss": 0.82258701, + "learning_rate": 3.943634266379908e-06, + "loss": 0.84561527, + "num_input_tokens_seen": 18327400, + "step": 861, + "time_per_iteration": 2.666307210922241 + }, + { + "auxiliary_loss_clip": 0.01257965, + "auxiliary_loss_mlp": 0.01045831, + "balance_loss_clip": 1.06655717, + "balance_loss_mlp": 1.03443491, + "epoch": 0.10364937173089642, + "flos": 25558558329600.0, + "grad_norm": 2.108935773463216, + "language_loss": 0.84808969, + "learning_rate": 3.943450487303815e-06, + "loss": 0.87112761, + "num_input_tokens_seen": 18347895, + "step": 862, + "time_per_iteration": 2.6944217681884766 + }, + { + "auxiliary_loss_clip": 0.01253132, + "auxiliary_loss_mlp": 0.01035798, + "balance_loss_clip": 1.0664103, + "balance_loss_mlp": 1.02490246, + "epoch": 0.1037696146215355, + "flos": 21215486004480.0, + "grad_norm": 2.0357794123360926, + "language_loss": 0.85396701, + "learning_rate": 3.943266413407827e-06, + "loss": 0.87685633, + "num_input_tokens_seen": 18367170, + "step": 863, + "time_per_iteration": 2.70969820022583 + }, + { + "auxiliary_loss_clip": 0.01260747, + "auxiliary_loss_mlp": 0.01045225, + "balance_loss_clip": 1.07013011, + "balance_loss_mlp": 1.03407288, + "epoch": 0.1038898575121746, + "flos": 25807382818560.0, + "grad_norm": 2.111924961033968, + "language_loss": 0.84791493, + "learning_rate": 3.94308204471987e-06, + "loss": 0.87097454, + "num_input_tokens_seen": 18386185, + "step": 864, + "time_per_iteration": 2.6796107292175293 + }, + { + "auxiliary_loss_clip": 0.01269975, + "auxiliary_loss_mlp": 0.01031289, + "balance_loss_clip": 1.06406879, + "balance_loss_mlp": 1.02120972, + "epoch": 0.10401010040281368, + "flos": 19062425900160.0, + "grad_norm": 2.4466474752070346, + "language_loss": 0.74872047, + "learning_rate": 3.942897381267912e-06, + "loss": 0.77173311, + "num_input_tokens_seen": 18402550, + "step": 865, + "time_per_iteration": 2.70306658744812 + }, + { + "auxiliary_loss_clip": 0.0125903, + "auxiliary_loss_mlp": 0.01041101, + "balance_loss_clip": 1.06780374, + "balance_loss_mlp": 1.03013968, + "epoch": 0.10413034329345278, + "flos": 16355119962240.0, + "grad_norm": 3.107924121936569, + "language_loss": 0.66178596, + "learning_rate": 3.942712423079965e-06, + "loss": 0.68478727, + "num_input_tokens_seen": 18418940, + "step": 866, + "time_per_iteration": 2.682680606842041 + }, + { + "auxiliary_loss_clip": 0.01252387, + "auxiliary_loss_mlp": 0.01037561, + "balance_loss_clip": 1.05810463, + "balance_loss_mlp": 1.02705836, + "epoch": 0.10425058618409186, + "flos": 17236511890560.0, + "grad_norm": 2.671738392383511, + "language_loss": 0.8997426, + "learning_rate": 3.942527170184088e-06, + "loss": 0.92264205, + "num_input_tokens_seen": 18435560, + "step": 867, + "time_per_iteration": 2.69118595123291 + }, + { + "auxiliary_loss_clip": 0.01250888, + "auxiliary_loss_mlp": 0.01035108, + "balance_loss_clip": 1.07078004, + "balance_loss_mlp": 1.02430153, + "epoch": 0.10437082907473096, + "flos": 17967365919360.0, + "grad_norm": 2.6317408435989527, + "language_loss": 0.77306128, + "learning_rate": 3.942341622608385e-06, + "loss": 0.79592121, + "num_input_tokens_seen": 18452590, + "step": 868, + "time_per_iteration": 2.569462537765503 + }, + { + "auxiliary_loss_clip": 0.01268129, + "auxiliary_loss_mlp": 0.0103986, + "balance_loss_clip": 1.0700047, + "balance_loss_mlp": 1.02912557, + "epoch": 0.10449107196537005, + "flos": 36283315374720.0, + "grad_norm": 2.379888273193229, + "language_loss": 0.77901298, + "learning_rate": 3.942155780381001e-06, + "loss": 0.80209291, + "num_input_tokens_seen": 18476325, + "step": 869, + "time_per_iteration": 2.87731671333313 + }, + { + "auxiliary_loss_clip": 0.01266105, + "auxiliary_loss_mlp": 0.01042071, + "balance_loss_clip": 1.06598639, + "balance_loss_mlp": 1.03066838, + "epoch": 0.10461131485600914, + "flos": 23802095266560.0, + "grad_norm": 1.9129734527461169, + "language_loss": 0.75481039, + "learning_rate": 3.94196964353013e-06, + "loss": 0.77789211, + "num_input_tokens_seen": 18495775, + "step": 870, + "time_per_iteration": 2.75899076461792 + }, + { + "auxiliary_loss_clip": 0.0126211, + "auxiliary_loss_mlp": 0.01381705, + "balance_loss_clip": 1.06525838, + "balance_loss_mlp": 1.00019145, + "epoch": 0.10473155774664823, + "flos": 18405476104320.0, + "grad_norm": 1.9511657565657032, + "language_loss": 0.8066026, + "learning_rate": 3.941783212084008e-06, + "loss": 0.83304071, + "num_input_tokens_seen": 18513530, + "step": 871, + "time_per_iteration": 2.7645883560180664 + }, + { + "auxiliary_loss_clip": 0.01244718, + "auxiliary_loss_mlp": 0.01043774, + "balance_loss_clip": 1.06445193, + "balance_loss_mlp": 1.03278327, + "epoch": 0.10485180063728732, + "flos": 25592637358080.0, + "grad_norm": 2.9207471880796603, + "language_loss": 0.79496908, + "learning_rate": 3.941596486070916e-06, + "loss": 0.81785393, + "num_input_tokens_seen": 18531575, + "step": 872, + "time_per_iteration": 2.855783462524414 + }, + { + "auxiliary_loss_clip": 0.01276164, + "auxiliary_loss_mlp": 0.01045037, + "balance_loss_clip": 1.06617725, + "balance_loss_mlp": 1.03364694, + "epoch": 0.10497204352792641, + "flos": 27088747666560.0, + "grad_norm": 2.575069332289133, + "language_loss": 0.58593208, + "learning_rate": 3.941409465519182e-06, + "loss": 0.60914409, + "num_input_tokens_seen": 18552100, + "step": 873, + "time_per_iteration": 2.8510470390319824 + }, + { + "auxiliary_loss_clip": 0.01248008, + "auxiliary_loss_mlp": 0.01037064, + "balance_loss_clip": 1.06544554, + "balance_loss_mlp": 1.02522671, + "epoch": 0.10509228641856551, + "flos": 32858479353600.0, + "grad_norm": 1.6436924120612624, + "language_loss": 0.85346377, + "learning_rate": 3.941222150457176e-06, + "loss": 0.87631452, + "num_input_tokens_seen": 18575355, + "step": 874, + "time_per_iteration": 3.78497052192688 + }, + { + "auxiliary_loss_clip": 0.01263022, + "auxiliary_loss_mlp": 0.01045155, + "balance_loss_clip": 1.06840658, + "balance_loss_mlp": 1.03394318, + "epoch": 0.10521252930920459, + "flos": 14319165173760.0, + "grad_norm": 2.8902071140314773, + "language_loss": 0.71593601, + "learning_rate": 3.941034540913311e-06, + "loss": 0.73901778, + "num_input_tokens_seen": 18592885, + "step": 875, + "time_per_iteration": 2.789520740509033 + }, + { + "auxiliary_loss_clip": 0.01260907, + "auxiliary_loss_mlp": 0.013821, + "balance_loss_clip": 1.0695014, + "balance_loss_mlp": 1.00016618, + "epoch": 0.10533277219984369, + "flos": 21687028773120.0, + "grad_norm": 1.6434241962178044, + "language_loss": 0.82445335, + "learning_rate": 3.940846636916051e-06, + "loss": 0.85088336, + "num_input_tokens_seen": 18612920, + "step": 876, + "time_per_iteration": 3.5382983684539795 + }, + { + "auxiliary_loss_clip": 0.01268219, + "auxiliary_loss_mlp": 0.01046157, + "balance_loss_clip": 1.07170248, + "balance_loss_mlp": 1.03533864, + "epoch": 0.10545301509048277, + "flos": 22269787027200.0, + "grad_norm": 2.007526584642881, + "language_loss": 0.86416364, + "learning_rate": 3.940658438493899e-06, + "loss": 0.88730741, + "num_input_tokens_seen": 18630765, + "step": 877, + "time_per_iteration": 3.7470953464508057 + }, + { + "auxiliary_loss_clip": 0.01249824, + "auxiliary_loss_mlp": 0.01036353, + "balance_loss_clip": 1.06700838, + "balance_loss_mlp": 1.02397346, + "epoch": 0.10557325798112187, + "flos": 22199725549440.0, + "grad_norm": 2.2280650977427174, + "language_loss": 0.76226217, + "learning_rate": 3.940469945675405e-06, + "loss": 0.78512394, + "num_input_tokens_seen": 18649150, + "step": 878, + "time_per_iteration": 2.6858861446380615 + }, + { + "auxiliary_loss_clip": 0.01250619, + "auxiliary_loss_mlp": 0.01034963, + "balance_loss_clip": 1.05748737, + "balance_loss_mlp": 1.02480626, + "epoch": 0.10569350087176095, + "flos": 25775889569280.0, + "grad_norm": 2.0110930555062376, + "language_loss": 0.91437984, + "learning_rate": 3.940281158489163e-06, + "loss": 0.93723565, + "num_input_tokens_seen": 18668380, + "step": 879, + "time_per_iteration": 2.823310136795044 + }, + { + "auxiliary_loss_clip": 0.01281527, + "auxiliary_loss_mlp": 0.01048103, + "balance_loss_clip": 1.05724597, + "balance_loss_mlp": 1.03743327, + "epoch": 0.10581374376240005, + "flos": 17311385790720.0, + "grad_norm": 1.8358511144809808, + "language_loss": 0.83063149, + "learning_rate": 3.940092076963812e-06, + "loss": 0.85392785, + "num_input_tokens_seen": 18685875, + "step": 880, + "time_per_iteration": 2.8135986328125 + }, + { + "auxiliary_loss_clip": 0.01257153, + "auxiliary_loss_mlp": 0.01035533, + "balance_loss_clip": 1.06206191, + "balance_loss_mlp": 1.02379119, + "epoch": 0.10593398665303914, + "flos": 34349454017280.0, + "grad_norm": 3.182283861497178, + "language_loss": 0.78996629, + "learning_rate": 3.9399027011280355e-06, + "loss": 0.81289321, + "num_input_tokens_seen": 18707970, + "step": 881, + "time_per_iteration": 2.8193700313568115 + }, + { + "auxiliary_loss_clip": 0.01265751, + "auxiliary_loss_mlp": 0.01040288, + "balance_loss_clip": 1.07007766, + "balance_loss_mlp": 1.0281291, + "epoch": 0.10605422954367823, + "flos": 23257977068160.0, + "grad_norm": 2.1958751395867493, + "language_loss": 0.77490067, + "learning_rate": 3.939713031010561e-06, + "loss": 0.79796106, + "num_input_tokens_seen": 18726335, + "step": 882, + "time_per_iteration": 2.677690267562866 + }, + { + "auxiliary_loss_clip": 0.01274378, + "auxiliary_loss_mlp": 0.01041783, + "balance_loss_clip": 1.0667882, + "balance_loss_mlp": 1.03064895, + "epoch": 0.10617447243431732, + "flos": 22820118278400.0, + "grad_norm": 2.2194052786499956, + "language_loss": 0.77644861, + "learning_rate": 3.939523066640163e-06, + "loss": 0.79961014, + "num_input_tokens_seen": 18745230, + "step": 883, + "time_per_iteration": 2.719846248626709 + }, + { + "auxiliary_loss_clip": 0.01260941, + "auxiliary_loss_mlp": 0.01047763, + "balance_loss_clip": 1.07066214, + "balance_loss_mlp": 1.03630662, + "epoch": 0.10629471532495641, + "flos": 24386577373440.0, + "grad_norm": 1.8055486725496042, + "language_loss": 0.81131744, + "learning_rate": 3.939332808045657e-06, + "loss": 0.83440447, + "num_input_tokens_seen": 18764880, + "step": 884, + "time_per_iteration": 2.709035873413086 + }, + { + "auxiliary_loss_clip": 0.01273729, + "auxiliary_loss_mlp": 0.01041915, + "balance_loss_clip": 1.06384766, + "balance_loss_mlp": 1.03094149, + "epoch": 0.1064149582155955, + "flos": 21105491581440.0, + "grad_norm": 1.6743351220918528, + "language_loss": 0.84684628, + "learning_rate": 3.939142255255906e-06, + "loss": 0.87000275, + "num_input_tokens_seen": 18785765, + "step": 885, + "time_per_iteration": 2.703964948654175 + }, + { + "auxiliary_loss_clip": 0.01258735, + "auxiliary_loss_mlp": 0.01039349, + "balance_loss_clip": 1.07068038, + "balance_loss_mlp": 1.02823317, + "epoch": 0.1065352011062346, + "flos": 20702035042560.0, + "grad_norm": 2.3756822110946536, + "language_loss": 0.86630481, + "learning_rate": 3.938951408299817e-06, + "loss": 0.88928562, + "num_input_tokens_seen": 18804605, + "step": 886, + "time_per_iteration": 2.678270101547241 + }, + { + "auxiliary_loss_clip": 0.01177298, + "auxiliary_loss_mlp": 0.01010453, + "balance_loss_clip": 1.03695369, + "balance_loss_mlp": 1.00779462, + "epoch": 0.10665544399687368, + "flos": 62659632689280.0, + "grad_norm": 0.7963270058250194, + "language_loss": 0.54453242, + "learning_rate": 3.938760267206342e-06, + "loss": 0.56640995, + "num_input_tokens_seen": 18866425, + "step": 887, + "time_per_iteration": 3.212491512298584 + }, + { + "auxiliary_loss_clip": 0.01250819, + "auxiliary_loss_mlp": 0.0104072, + "balance_loss_clip": 1.07135606, + "balance_loss_mlp": 1.03043234, + "epoch": 0.10677568688751278, + "flos": 26140382830080.0, + "grad_norm": 2.233841178099758, + "language_loss": 0.7839132, + "learning_rate": 3.938568832004475e-06, + "loss": 0.80682856, + "num_input_tokens_seen": 18885130, + "step": 888, + "time_per_iteration": 2.6243391036987305 + }, + { + "auxiliary_loss_clip": 0.01250627, + "auxiliary_loss_mlp": 0.01040212, + "balance_loss_clip": 1.0621469, + "balance_loss_mlp": 1.02872562, + "epoch": 0.10689592977815186, + "flos": 12786533712000.0, + "grad_norm": 1.855418871181123, + "language_loss": 0.75663328, + "learning_rate": 3.938377102723257e-06, + "loss": 0.77954167, + "num_input_tokens_seen": 18902265, + "step": 889, + "time_per_iteration": 2.70369553565979 + }, + { + "auxiliary_loss_clip": 0.01258106, + "auxiliary_loss_mlp": 0.01042178, + "balance_loss_clip": 1.06050479, + "balance_loss_mlp": 1.03016138, + "epoch": 0.10701617266879096, + "flos": 22126683242880.0, + "grad_norm": 2.2612576415277363, + "language_loss": 0.83662558, + "learning_rate": 3.938185079391774e-06, + "loss": 0.85962844, + "num_input_tokens_seen": 18919310, + "step": 890, + "time_per_iteration": 2.7525975704193115 + }, + { + "auxiliary_loss_clip": 0.01249807, + "auxiliary_loss_mlp": 0.0105018, + "balance_loss_clip": 1.07018018, + "balance_loss_mlp": 1.0381639, + "epoch": 0.10713641555943004, + "flos": 19745625559680.0, + "grad_norm": 2.356995775294376, + "language_loss": 1.05718982, + "learning_rate": 3.937992762039157e-06, + "loss": 1.0801897, + "num_input_tokens_seen": 18932635, + "step": 891, + "time_per_iteration": 2.621098279953003 + }, + { + "auxiliary_loss_clip": 0.01256118, + "auxiliary_loss_mlp": 0.01045239, + "balance_loss_clip": 1.07088995, + "balance_loss_mlp": 1.03415227, + "epoch": 0.10725665845006914, + "flos": 23952992302080.0, + "grad_norm": 1.7935162776726818, + "language_loss": 0.80686468, + "learning_rate": 3.937800150694577e-06, + "loss": 0.82987827, + "num_input_tokens_seen": 18953810, + "step": 892, + "time_per_iteration": 2.687190055847168 + }, + { + "auxiliary_loss_clip": 0.01285414, + "auxiliary_loss_mlp": 0.01042285, + "balance_loss_clip": 1.06699622, + "balance_loss_mlp": 1.03049529, + "epoch": 0.10737690134070824, + "flos": 18551704371840.0, + "grad_norm": 2.493448143779079, + "language_loss": 0.75993669, + "learning_rate": 3.937607245387255e-06, + "loss": 0.78321373, + "num_input_tokens_seen": 18973175, + "step": 893, + "time_per_iteration": 2.734126329421997 + }, + { + "auxiliary_loss_clip": 0.01268466, + "auxiliary_loss_mlp": 0.01034965, + "balance_loss_clip": 1.06447673, + "balance_loss_mlp": 1.02504623, + "epoch": 0.10749714423134732, + "flos": 22707609903360.0, + "grad_norm": 2.0522613188088212, + "language_loss": 0.72133893, + "learning_rate": 3.937414046146455e-06, + "loss": 0.7443732, + "num_input_tokens_seen": 18991130, + "step": 894, + "time_per_iteration": 2.686249017715454 + }, + { + "auxiliary_loss_clip": 0.01252497, + "auxiliary_loss_mlp": 0.01041053, + "balance_loss_clip": 1.07249689, + "balance_loss_mlp": 1.02993059, + "epoch": 0.10761738712198642, + "flos": 21106066199040.0, + "grad_norm": 2.47412290145592, + "language_loss": 0.75590986, + "learning_rate": 3.9372205530014845e-06, + "loss": 0.77884531, + "num_input_tokens_seen": 19009610, + "step": 895, + "time_per_iteration": 2.6026995182037354 + }, + { + "auxiliary_loss_clip": 0.01250245, + "auxiliary_loss_mlp": 0.01043559, + "balance_loss_clip": 1.06969285, + "balance_loss_mlp": 1.03201938, + "epoch": 0.1077376300126255, + "flos": 23766723348480.0, + "grad_norm": 2.3353313991398545, + "language_loss": 0.7125814, + "learning_rate": 3.937026765981696e-06, + "loss": 0.73551941, + "num_input_tokens_seen": 19029680, + "step": 896, + "time_per_iteration": 2.7144899368286133 + }, + { + "auxiliary_loss_clip": 0.0127871, + "auxiliary_loss_mlp": 0.01049938, + "balance_loss_clip": 1.06953394, + "balance_loss_mlp": 1.03885722, + "epoch": 0.1078578729032646, + "flos": 20919581763840.0, + "grad_norm": 1.8725535135756692, + "language_loss": 0.79285502, + "learning_rate": 3.936832685116488e-06, + "loss": 0.81614149, + "num_input_tokens_seen": 19047775, + "step": 897, + "time_per_iteration": 2.744060754776001 + }, + { + "auxiliary_loss_clip": 0.01252711, + "auxiliary_loss_mlp": 0.01045292, + "balance_loss_clip": 1.07283175, + "balance_loss_mlp": 1.03365183, + "epoch": 0.10797811579390369, + "flos": 14829886702080.0, + "grad_norm": 2.0441680924431527, + "language_loss": 0.90109956, + "learning_rate": 3.936638310435301e-06, + "loss": 0.92407954, + "num_input_tokens_seen": 19065640, + "step": 898, + "time_per_iteration": 2.7598519325256348 + }, + { + "auxiliary_loss_clip": 0.01264839, + "auxiliary_loss_mlp": 0.01041072, + "balance_loss_clip": 1.07256269, + "balance_loss_mlp": 1.03050447, + "epoch": 0.10809835868454278, + "flos": 19536985411200.0, + "grad_norm": 1.8713397544002135, + "language_loss": 0.81756335, + "learning_rate": 3.936443641967623e-06, + "loss": 0.84062243, + "num_input_tokens_seen": 19084470, + "step": 899, + "time_per_iteration": 2.691602945327759 + }, + { + "auxiliary_loss_clip": 0.01275284, + "auxiliary_loss_mlp": 0.0105049, + "balance_loss_clip": 1.07169783, + "balance_loss_mlp": 1.03970742, + "epoch": 0.10821860157518187, + "flos": 18442320480000.0, + "grad_norm": 1.9207239134550436, + "language_loss": 0.82889408, + "learning_rate": 3.936248679742983e-06, + "loss": 0.85215181, + "num_input_tokens_seen": 19102965, + "step": 900, + "time_per_iteration": 2.758188247680664 + }, + { + "auxiliary_loss_clip": 0.01160344, + "auxiliary_loss_mlp": 0.01012599, + "balance_loss_clip": 1.02963257, + "balance_loss_mlp": 1.00955892, + "epoch": 0.10833884446582095, + "flos": 49359468447360.0, + "grad_norm": 1.0470680812671083, + "language_loss": 0.70233166, + "learning_rate": 3.936053423790959e-06, + "loss": 0.72406107, + "num_input_tokens_seen": 19151285, + "step": 901, + "time_per_iteration": 3.9609711170196533 + }, + { + "auxiliary_loss_clip": 0.01258092, + "auxiliary_loss_mlp": 0.0103613, + "balance_loss_clip": 1.07854223, + "balance_loss_mlp": 1.02526999, + "epoch": 0.10845908735646005, + "flos": 20411912891520.0, + "grad_norm": 1.761246016668262, + "language_loss": 0.77316707, + "learning_rate": 3.935857874141168e-06, + "loss": 0.79610932, + "num_input_tokens_seen": 19170120, + "step": 902, + "time_per_iteration": 2.665623188018799 + }, + { + "auxiliary_loss_clip": 0.01261344, + "auxiliary_loss_mlp": 0.01043404, + "balance_loss_clip": 1.06833744, + "balance_loss_mlp": 1.03191781, + "epoch": 0.10857933024709913, + "flos": 14027750133120.0, + "grad_norm": 2.1767026287817277, + "language_loss": 0.83297706, + "learning_rate": 3.935662030823279e-06, + "loss": 0.8560245, + "num_input_tokens_seen": 19186305, + "step": 903, + "time_per_iteration": 4.605750560760498 + }, + { + "auxiliary_loss_clip": 0.01262569, + "auxiliary_loss_mlp": 0.01045662, + "balance_loss_clip": 1.07055402, + "balance_loss_mlp": 1.03393161, + "epoch": 0.10869957313773823, + "flos": 13369004657280.0, + "grad_norm": 2.103525651812255, + "language_loss": 0.72393703, + "learning_rate": 3.935465893866998e-06, + "loss": 0.74701941, + "num_input_tokens_seen": 19204530, + "step": 904, + "time_per_iteration": 2.7376036643981934 + }, + { + "auxiliary_loss_clip": 0.01267927, + "auxiliary_loss_mlp": 0.01039494, + "balance_loss_clip": 1.07157874, + "balance_loss_mlp": 1.02838933, + "epoch": 0.10881981602837733, + "flos": 25807095509760.0, + "grad_norm": 1.9626365414844038, + "language_loss": 0.80021077, + "learning_rate": 3.935269463302079e-06, + "loss": 0.82328498, + "num_input_tokens_seen": 19222735, + "step": 905, + "time_per_iteration": 2.7859182357788086 + }, + { + "auxiliary_loss_clip": 0.01268816, + "auxiliary_loss_mlp": 0.01051267, + "balance_loss_clip": 1.07488263, + "balance_loss_mlp": 1.040133, + "epoch": 0.10894005891901641, + "flos": 20777555387520.0, + "grad_norm": 1.8941542401529212, + "language_loss": 0.76913893, + "learning_rate": 3.935072739158322e-06, + "loss": 0.7923398, + "num_input_tokens_seen": 19242445, + "step": 906, + "time_per_iteration": 2.688711404800415 + }, + { + "auxiliary_loss_clip": 0.01266289, + "auxiliary_loss_mlp": 0.01042746, + "balance_loss_clip": 1.06824613, + "balance_loss_mlp": 1.03141499, + "epoch": 0.10906030180965551, + "flos": 26649883296000.0, + "grad_norm": 1.6746426477236334, + "language_loss": 0.80102372, + "learning_rate": 3.934875721465569e-06, + "loss": 0.82411408, + "num_input_tokens_seen": 19262865, + "step": 907, + "time_per_iteration": 2.802865505218506 + }, + { + "auxiliary_loss_clip": 0.01265617, + "auxiliary_loss_mlp": 0.01043178, + "balance_loss_clip": 1.06842554, + "balance_loss_mlp": 1.03142357, + "epoch": 0.10918054470029459, + "flos": 36534402420480.0, + "grad_norm": 2.3734972772823344, + "language_loss": 0.71725315, + "learning_rate": 3.9346784102537076e-06, + "loss": 0.74034107, + "num_input_tokens_seen": 19285000, + "step": 908, + "time_per_iteration": 3.189213991165161 + }, + { + "auxiliary_loss_clip": 0.01252843, + "auxiliary_loss_mlp": 0.01042426, + "balance_loss_clip": 1.07223034, + "balance_loss_mlp": 1.03101778, + "epoch": 0.10930078759093369, + "flos": 21762549118080.0, + "grad_norm": 2.009888035260768, + "language_loss": 0.78078622, + "learning_rate": 3.934480805552669e-06, + "loss": 0.80373889, + "num_input_tokens_seen": 19306010, + "step": 909, + "time_per_iteration": 2.7977118492126465 + }, + { + "auxiliary_loss_clip": 0.0125204, + "auxiliary_loss_mlp": 0.01382634, + "balance_loss_clip": 1.07244384, + "balance_loss_mlp": 1.00045538, + "epoch": 0.10942103048157277, + "flos": 22601781457920.0, + "grad_norm": 2.286795119223008, + "language_loss": 0.87977993, + "learning_rate": 3.93428290739243e-06, + "loss": 0.90612674, + "num_input_tokens_seen": 19325380, + "step": 910, + "time_per_iteration": 2.7387771606445312 + }, + { + "auxiliary_loss_clip": 0.01273777, + "auxiliary_loss_mlp": 0.01042652, + "balance_loss_clip": 1.07268357, + "balance_loss_mlp": 1.03099942, + "epoch": 0.10954127337221187, + "flos": 15045781397760.0, + "grad_norm": 2.299600388352568, + "language_loss": 0.80370098, + "learning_rate": 3.9340847158030125e-06, + "loss": 0.82686532, + "num_input_tokens_seen": 19338960, + "step": 911, + "time_per_iteration": 2.6804039478302 + }, + { + "auxiliary_loss_clip": 0.01266316, + "auxiliary_loss_mlp": 0.01048264, + "balance_loss_clip": 1.07204175, + "balance_loss_mlp": 1.03636074, + "epoch": 0.10966151626285096, + "flos": 21650974496640.0, + "grad_norm": 2.226493293747816, + "language_loss": 0.75695527, + "learning_rate": 3.9338862308144814e-06, + "loss": 0.78010112, + "num_input_tokens_seen": 19357780, + "step": 912, + "time_per_iteration": 2.738926649093628 + }, + { + "auxiliary_loss_clip": 0.01255168, + "auxiliary_loss_mlp": 0.01041675, + "balance_loss_clip": 1.07564974, + "balance_loss_mlp": 1.03070164, + "epoch": 0.10978175915349005, + "flos": 20121359777280.0, + "grad_norm": 1.7093659640786059, + "language_loss": 0.84579712, + "learning_rate": 3.933687452456946e-06, + "loss": 0.86876559, + "num_input_tokens_seen": 19377680, + "step": 913, + "time_per_iteration": 2.607318878173828 + }, + { + "auxiliary_loss_clip": 0.0127002, + "auxiliary_loss_mlp": 0.01040031, + "balance_loss_clip": 1.06219518, + "balance_loss_mlp": 1.02746594, + "epoch": 0.10990200204412914, + "flos": 20412667077120.0, + "grad_norm": 2.0525239093133565, + "language_loss": 0.86321342, + "learning_rate": 3.933488380760562e-06, + "loss": 0.88631392, + "num_input_tokens_seen": 19397040, + "step": 914, + "time_per_iteration": 2.721909284591675 + }, + { + "auxiliary_loss_clip": 0.01254, + "auxiliary_loss_mlp": 0.01382422, + "balance_loss_clip": 1.07299256, + "balance_loss_mlp": 1.00046194, + "epoch": 0.11002224493476823, + "flos": 17530117660800.0, + "grad_norm": 2.8669807118367747, + "language_loss": 0.87220287, + "learning_rate": 3.9332890157555286e-06, + "loss": 0.89856708, + "num_input_tokens_seen": 19413975, + "step": 915, + "time_per_iteration": 2.621328830718994 + }, + { + "auxiliary_loss_clip": 0.01272241, + "auxiliary_loss_mlp": 0.01047759, + "balance_loss_clip": 1.06972957, + "balance_loss_mlp": 1.0352838, + "epoch": 0.11014248782540732, + "flos": 12203093099520.0, + "grad_norm": 2.7266219794786, + "language_loss": 0.76774287, + "learning_rate": 3.933089357472088e-06, + "loss": 0.79094291, + "num_input_tokens_seen": 19432005, + "step": 916, + "time_per_iteration": 2.676187515258789 + }, + { + "auxiliary_loss_clip": 0.01255482, + "auxiliary_loss_mlp": 0.01051457, + "balance_loss_clip": 1.07690895, + "balance_loss_mlp": 1.04030466, + "epoch": 0.11026273071604642, + "flos": 22382977760640.0, + "grad_norm": 1.9191469596592963, + "language_loss": 0.85736167, + "learning_rate": 3.932889405940529e-06, + "loss": 0.88043112, + "num_input_tokens_seen": 19450100, + "step": 917, + "time_per_iteration": 2.707909107208252 + }, + { + "auxiliary_loss_clip": 0.01271471, + "auxiliary_loss_mlp": 0.01046947, + "balance_loss_clip": 1.07191455, + "balance_loss_mlp": 1.03536558, + "epoch": 0.1103829736066855, + "flos": 19829046896640.0, + "grad_norm": 2.5202272047860723, + "language_loss": 0.79716563, + "learning_rate": 3.932689161191184e-06, + "loss": 0.82034981, + "num_input_tokens_seen": 19467805, + "step": 918, + "time_per_iteration": 2.7675201892852783 + }, + { + "auxiliary_loss_clip": 0.01260767, + "auxiliary_loss_mlp": 0.01042481, + "balance_loss_clip": 1.07063842, + "balance_loss_mlp": 1.03100753, + "epoch": 0.1105032164973246, + "flos": 22669616292480.0, + "grad_norm": 2.0203036703392443, + "language_loss": 0.8797676, + "learning_rate": 3.93248862325443e-06, + "loss": 0.90280008, + "num_input_tokens_seen": 19486710, + "step": 919, + "time_per_iteration": 2.6753127574920654 + }, + { + "auxiliary_loss_clip": 0.01140647, + "auxiliary_loss_mlp": 0.01005667, + "balance_loss_clip": 1.02790403, + "balance_loss_mlp": 1.00230491, + "epoch": 0.11062345938796368, + "flos": 66483507876480.0, + "grad_norm": 0.9281503827801143, + "language_loss": 0.64535522, + "learning_rate": 3.932287792160688e-06, + "loss": 0.66681838, + "num_input_tokens_seen": 19545170, + "step": 920, + "time_per_iteration": 3.107619285583496 + }, + { + "auxiliary_loss_clip": 0.01263309, + "auxiliary_loss_mlp": 0.01048294, + "balance_loss_clip": 1.07134461, + "balance_loss_mlp": 1.03629518, + "epoch": 0.11074370227860278, + "flos": 21907771804800.0, + "grad_norm": 3.0250638212926115, + "language_loss": 0.8031292, + "learning_rate": 3.932086667940424e-06, + "loss": 0.82624525, + "num_input_tokens_seen": 19561875, + "step": 921, + "time_per_iteration": 2.7069458961486816 + }, + { + "auxiliary_loss_clip": 0.01258075, + "auxiliary_loss_mlp": 0.01382149, + "balance_loss_clip": 1.07221222, + "balance_loss_mlp": 1.00046182, + "epoch": 0.11086394516924186, + "flos": 28658115763200.0, + "grad_norm": 1.7925961636668861, + "language_loss": 0.8161695, + "learning_rate": 3.93188525062415e-06, + "loss": 0.84257174, + "num_input_tokens_seen": 19582340, + "step": 922, + "time_per_iteration": 2.7770235538482666 + }, + { + "auxiliary_loss_clip": 0.01264077, + "auxiliary_loss_mlp": 0.01045478, + "balance_loss_clip": 1.0723387, + "balance_loss_mlp": 1.03340769, + "epoch": 0.11098418805988096, + "flos": 24535247765760.0, + "grad_norm": 1.8206342164088907, + "language_loss": 0.86270773, + "learning_rate": 3.931683540242418e-06, + "loss": 0.88580334, + "num_input_tokens_seen": 19603405, + "step": 923, + "time_per_iteration": 2.7651541233062744 + }, + { + "auxiliary_loss_clip": 0.01250278, + "auxiliary_loss_mlp": 0.0104184, + "balance_loss_clip": 1.06780112, + "balance_loss_mlp": 1.03066397, + "epoch": 0.11110443095052006, + "flos": 22960384888320.0, + "grad_norm": 2.7149736680861496, + "language_loss": 0.90689552, + "learning_rate": 3.9314815368258295e-06, + "loss": 0.9298166, + "num_input_tokens_seen": 19619885, + "step": 924, + "time_per_iteration": 2.651249647140503 + }, + { + "auxiliary_loss_clip": 0.01264194, + "auxiliary_loss_mlp": 0.01037088, + "balance_loss_clip": 1.07548141, + "balance_loss_mlp": 1.02669334, + "epoch": 0.11122467384115914, + "flos": 18950025265920.0, + "grad_norm": 3.0595571261208145, + "language_loss": 0.78736335, + "learning_rate": 3.9312792404050275e-06, + "loss": 0.81037617, + "num_input_tokens_seen": 19637940, + "step": 925, + "time_per_iteration": 2.6792051792144775 + }, + { + "auxiliary_loss_clip": 0.01250168, + "auxiliary_loss_mlp": 0.01042501, + "balance_loss_clip": 1.07440996, + "balance_loss_mlp": 1.03148651, + "epoch": 0.11134491673179824, + "flos": 25082957324160.0, + "grad_norm": 1.8151253695741367, + "language_loss": 0.77313733, + "learning_rate": 3.9310766510107e-06, + "loss": 0.79606402, + "num_input_tokens_seen": 19657115, + "step": 926, + "time_per_iteration": 2.686643362045288 + }, + { + "auxiliary_loss_clip": 0.01274763, + "auxiliary_loss_mlp": 0.01046855, + "balance_loss_clip": 1.06539142, + "balance_loss_mlp": 1.03488076, + "epoch": 0.11146515962243732, + "flos": 24499121662080.0, + "grad_norm": 1.8944811588644515, + "language_loss": 0.92147005, + "learning_rate": 3.9308737686735806e-06, + "loss": 0.94468617, + "num_input_tokens_seen": 19677075, + "step": 927, + "time_per_iteration": 3.6707701683044434 + }, + { + "auxiliary_loss_clip": 0.01252187, + "auxiliary_loss_mlp": 0.01049272, + "balance_loss_clip": 1.0726161, + "balance_loss_mlp": 1.03727913, + "epoch": 0.11158540251307641, + "flos": 22343763087360.0, + "grad_norm": 1.913613472900898, + "language_loss": 0.82941681, + "learning_rate": 3.9306705934244455e-06, + "loss": 0.85243142, + "num_input_tokens_seen": 19697155, + "step": 928, + "time_per_iteration": 2.761420726776123 + }, + { + "auxiliary_loss_clip": 0.0125232, + "auxiliary_loss_mlp": 0.01040163, + "balance_loss_clip": 1.06595325, + "balance_loss_mlp": 1.02948189, + "epoch": 0.11170564540371551, + "flos": 19902304684800.0, + "grad_norm": 1.6596004129665347, + "language_loss": 0.88196868, + "learning_rate": 3.930467125294116e-06, + "loss": 0.90489352, + "num_input_tokens_seen": 19716705, + "step": 929, + "time_per_iteration": 4.618546962738037 + }, + { + "auxiliary_loss_clip": 0.01173924, + "auxiliary_loss_mlp": 0.0100441, + "balance_loss_clip": 1.02249527, + "balance_loss_mlp": 1.00109577, + "epoch": 0.1118258882943546, + "flos": 64586239499520.0, + "grad_norm": 0.9248569360289243, + "language_loss": 0.60427558, + "learning_rate": 3.930263364313458e-06, + "loss": 0.62605894, + "num_input_tokens_seen": 19767275, + "step": 930, + "time_per_iteration": 3.107739210128784 + }, + { + "auxiliary_loss_clip": 0.01270502, + "auxiliary_loss_mlp": 0.01042068, + "balance_loss_clip": 1.06551003, + "balance_loss_mlp": 1.03063583, + "epoch": 0.11194613118499369, + "flos": 17201965985280.0, + "grad_norm": 1.9416618922383388, + "language_loss": 0.82761413, + "learning_rate": 3.930059310513384e-06, + "loss": 0.85073984, + "num_input_tokens_seen": 19786315, + "step": 931, + "time_per_iteration": 2.8286333084106445 + }, + { + "auxiliary_loss_clip": 0.01256924, + "auxiliary_loss_mlp": 0.01382297, + "balance_loss_clip": 1.0648402, + "balance_loss_mlp": 1.00043344, + "epoch": 0.11206637407563277, + "flos": 31863465728640.0, + "grad_norm": 1.765135944332004, + "language_loss": 0.84034216, + "learning_rate": 3.929854963924846e-06, + "loss": 0.86673439, + "num_input_tokens_seen": 19806580, + "step": 932, + "time_per_iteration": 2.8296828269958496 + }, + { + "auxiliary_loss_clip": 0.01267411, + "auxiliary_loss_mlp": 0.01040831, + "balance_loss_clip": 1.06237602, + "balance_loss_mlp": 1.02927351, + "epoch": 0.11218661696627187, + "flos": 21945621761280.0, + "grad_norm": 1.7780314376460093, + "language_loss": 0.77072966, + "learning_rate": 3.929650324578845e-06, + "loss": 0.79381204, + "num_input_tokens_seen": 19826045, + "step": 933, + "time_per_iteration": 2.7179458141326904 + }, + { + "auxiliary_loss_clip": 0.01267488, + "auxiliary_loss_mlp": 0.01033048, + "balance_loss_clip": 1.06625795, + "balance_loss_mlp": 1.02160406, + "epoch": 0.11230685985691095, + "flos": 25878198481920.0, + "grad_norm": 2.254191802776774, + "language_loss": 0.82149041, + "learning_rate": 3.929445392506423e-06, + "loss": 0.84449577, + "num_input_tokens_seen": 19843985, + "step": 934, + "time_per_iteration": 2.825714588165283 + }, + { + "auxiliary_loss_clip": 0.01258068, + "auxiliary_loss_mlp": 0.01043747, + "balance_loss_clip": 1.07408094, + "balance_loss_mlp": 1.03263056, + "epoch": 0.11242710274755005, + "flos": 22231506107520.0, + "grad_norm": 2.0109679194584738, + "language_loss": 0.75787854, + "learning_rate": 3.92924016773867e-06, + "loss": 0.78089666, + "num_input_tokens_seen": 19860480, + "step": 935, + "time_per_iteration": 2.7335453033447266 + }, + { + "auxiliary_loss_clip": 0.01261748, + "auxiliary_loss_mlp": 0.01382018, + "balance_loss_clip": 1.06488883, + "balance_loss_mlp": 1.00040269, + "epoch": 0.11254734563818915, + "flos": 17712184723200.0, + "grad_norm": 3.0252738868208686, + "language_loss": 0.73666382, + "learning_rate": 3.9290346503067175e-06, + "loss": 0.76310146, + "num_input_tokens_seen": 19877145, + "step": 936, + "time_per_iteration": 2.7060370445251465 + }, + { + "auxiliary_loss_clip": 0.01256876, + "auxiliary_loss_mlp": 0.01051713, + "balance_loss_clip": 1.0682627, + "balance_loss_mlp": 1.04068005, + "epoch": 0.11266758852882823, + "flos": 54930397334400.0, + "grad_norm": 1.745836548462731, + "language_loss": 0.789186, + "learning_rate": 3.9288288402417415e-06, + "loss": 0.81227195, + "num_input_tokens_seen": 19903405, + "step": 937, + "time_per_iteration": 2.9909377098083496 + }, + { + "auxiliary_loss_clip": 0.01257128, + "auxiliary_loss_mlp": 0.01049483, + "balance_loss_clip": 1.07142127, + "balance_loss_mlp": 1.03774703, + "epoch": 0.11278783141946733, + "flos": 18878132194560.0, + "grad_norm": 2.311455126671567, + "language_loss": 0.7055366, + "learning_rate": 3.928622737574964e-06, + "loss": 0.72860277, + "num_input_tokens_seen": 19918740, + "step": 938, + "time_per_iteration": 2.68542218208313 + }, + { + "auxiliary_loss_clip": 0.01260406, + "auxiliary_loss_mlp": 0.01040799, + "balance_loss_clip": 1.06521714, + "balance_loss_mlp": 1.02912235, + "epoch": 0.11290807431010641, + "flos": 26469252777600.0, + "grad_norm": 1.8129446821851505, + "language_loss": 0.90834045, + "learning_rate": 3.928416342337652e-06, + "loss": 0.9313525, + "num_input_tokens_seen": 19938475, + "step": 939, + "time_per_iteration": 2.749116897583008 + }, + { + "auxiliary_loss_clip": 0.01264932, + "auxiliary_loss_mlp": 0.01035746, + "balance_loss_clip": 1.06895161, + "balance_loss_mlp": 1.02484465, + "epoch": 0.1130283172007455, + "flos": 22710590732160.0, + "grad_norm": 1.7531050213465509, + "language_loss": 0.82701194, + "learning_rate": 3.928209654561113e-06, + "loss": 0.85001862, + "num_input_tokens_seen": 19959310, + "step": 940, + "time_per_iteration": 3.2237794399261475 + }, + { + "auxiliary_loss_clip": 0.01259021, + "auxiliary_loss_mlp": 0.010414, + "balance_loss_clip": 1.06874204, + "balance_loss_mlp": 1.03083158, + "epoch": 0.1131485600913846, + "flos": 23219911630080.0, + "grad_norm": 2.1338433555313787, + "language_loss": 0.81525755, + "learning_rate": 3.928002674276703e-06, + "loss": 0.83826172, + "num_input_tokens_seen": 19978700, + "step": 941, + "time_per_iteration": 2.738011121749878 + }, + { + "auxiliary_loss_clip": 0.01248098, + "auxiliary_loss_mlp": 0.0104075, + "balance_loss_clip": 1.05377173, + "balance_loss_mlp": 1.0291748, + "epoch": 0.11326880298202369, + "flos": 14064271286400.0, + "grad_norm": 2.1289118507247684, + "language_loss": 0.75117046, + "learning_rate": 3.92779540151582e-06, + "loss": 0.77405894, + "num_input_tokens_seen": 19995785, + "step": 942, + "time_per_iteration": 2.745229721069336 + }, + { + "auxiliary_loss_clip": 0.01265053, + "auxiliary_loss_mlp": 0.01043537, + "balance_loss_clip": 1.06938291, + "balance_loss_mlp": 1.0327301, + "epoch": 0.11338904587266278, + "flos": 16325386479360.0, + "grad_norm": 2.171388102717847, + "language_loss": 0.85725194, + "learning_rate": 3.927587836309907e-06, + "loss": 0.88033783, + "num_input_tokens_seen": 20013615, + "step": 943, + "time_per_iteration": 2.7218596935272217 + }, + { + "auxiliary_loss_clip": 0.012526, + "auxiliary_loss_mlp": 0.01034377, + "balance_loss_clip": 1.06275797, + "balance_loss_mlp": 1.02312374, + "epoch": 0.11350928876330187, + "flos": 24426258923520.0, + "grad_norm": 1.769158218487718, + "language_loss": 0.78347206, + "learning_rate": 3.927379978690452e-06, + "loss": 0.80634189, + "num_input_tokens_seen": 20032880, + "step": 944, + "time_per_iteration": 2.7028706073760986 + }, + { + "auxiliary_loss_clip": 0.01248054, + "auxiliary_loss_mlp": 0.0103867, + "balance_loss_clip": 1.05494761, + "balance_loss_mlp": 1.02796531, + "epoch": 0.11362953165394096, + "flos": 24497074586880.0, + "grad_norm": 2.1444722320512892, + "language_loss": 0.87252355, + "learning_rate": 3.927171828688987e-06, + "loss": 0.89539075, + "num_input_tokens_seen": 20052405, + "step": 945, + "time_per_iteration": 2.7429747581481934 + }, + { + "auxiliary_loss_clip": 0.01246732, + "auxiliary_loss_mlp": 0.01036058, + "balance_loss_clip": 1.070158, + "balance_loss_mlp": 1.02453637, + "epoch": 0.11374977454458005, + "flos": 24060831909120.0, + "grad_norm": 3.740580085403855, + "language_loss": 0.82191432, + "learning_rate": 3.926963386337088e-06, + "loss": 0.84474218, + "num_input_tokens_seen": 20070635, + "step": 946, + "time_per_iteration": 2.629817247390747 + }, + { + "auxiliary_loss_clip": 0.01249058, + "auxiliary_loss_mlp": 0.01032021, + "balance_loss_clip": 1.0706718, + "balance_loss_mlp": 1.02042818, + "epoch": 0.11387001743521914, + "flos": 39457638967680.0, + "grad_norm": 8.573573138917114, + "language_loss": 0.70317316, + "learning_rate": 3.926754651666375e-06, + "loss": 0.72598398, + "num_input_tokens_seen": 20091195, + "step": 947, + "time_per_iteration": 2.775566577911377 + }, + { + "auxiliary_loss_clip": 0.01269882, + "auxiliary_loss_mlp": 0.01041676, + "balance_loss_clip": 1.065534, + "balance_loss_mlp": 1.03098309, + "epoch": 0.11399026032585824, + "flos": 25082454533760.0, + "grad_norm": 3.376713169298172, + "language_loss": 0.78208661, + "learning_rate": 3.926545624708513e-06, + "loss": 0.80520213, + "num_input_tokens_seen": 20110435, + "step": 948, + "time_per_iteration": 2.7008090019226074 + }, + { + "auxiliary_loss_clip": 0.01267179, + "auxiliary_loss_mlp": 0.01037699, + "balance_loss_clip": 1.06331241, + "balance_loss_mlp": 1.02761972, + "epoch": 0.11411050321649732, + "flos": 17961835224960.0, + "grad_norm": 1.8025637367904075, + "language_loss": 0.85417438, + "learning_rate": 3.926336305495213e-06, + "loss": 0.87722313, + "num_input_tokens_seen": 20128995, + "step": 949, + "time_per_iteration": 2.6891064643859863 + }, + { + "auxiliary_loss_clip": 0.01252384, + "auxiliary_loss_mlp": 0.01035125, + "balance_loss_clip": 1.06090379, + "balance_loss_mlp": 1.02345419, + "epoch": 0.11423074610713642, + "flos": 22455409536000.0, + "grad_norm": 2.3704380582937947, + "language_loss": 0.88638484, + "learning_rate": 3.926126694058226e-06, + "loss": 0.90925997, + "num_input_tokens_seen": 20148145, + "step": 950, + "time_per_iteration": 2.699345350265503 + }, + { + "auxiliary_loss_clip": 0.01274477, + "auxiliary_loss_mlp": 0.01041021, + "balance_loss_clip": 1.06666279, + "balance_loss_mlp": 1.03090024, + "epoch": 0.1143509889977755, + "flos": 19717687756800.0, + "grad_norm": 1.3972898666147555, + "language_loss": 0.82153946, + "learning_rate": 3.92591679042935e-06, + "loss": 0.84469444, + "num_input_tokens_seen": 20168035, + "step": 951, + "time_per_iteration": 2.761415719985962 + }, + { + "auxiliary_loss_clip": 0.01255027, + "auxiliary_loss_mlp": 0.01043906, + "balance_loss_clip": 1.07083154, + "balance_loss_mlp": 1.03279567, + "epoch": 0.1144712318884146, + "flos": 19822869757440.0, + "grad_norm": 2.1815735577556725, + "language_loss": 0.82202899, + "learning_rate": 3.92570659464043e-06, + "loss": 0.84501839, + "num_input_tokens_seen": 20186095, + "step": 952, + "time_per_iteration": 2.6541032791137695 + }, + { + "auxiliary_loss_clip": 0.01253428, + "auxiliary_loss_mlp": 0.01381817, + "balance_loss_clip": 1.07241273, + "balance_loss_mlp": 1.00032902, + "epoch": 0.1145914747790537, + "flos": 14939198766720.0, + "grad_norm": 1.9348452401183467, + "language_loss": 0.79639405, + "learning_rate": 3.925496106723349e-06, + "loss": 0.82274652, + "num_input_tokens_seen": 20203535, + "step": 953, + "time_per_iteration": 3.606539726257324 + }, + { + "auxiliary_loss_clip": 0.01257596, + "auxiliary_loss_mlp": 0.01034593, + "balance_loss_clip": 1.06996846, + "balance_loss_mlp": 1.02364957, + "epoch": 0.11471171766969278, + "flos": 19865029345920.0, + "grad_norm": 1.9044487029695087, + "language_loss": 0.83626378, + "learning_rate": 3.9252853267100405e-06, + "loss": 0.85918564, + "num_input_tokens_seen": 20222780, + "step": 954, + "time_per_iteration": 2.7937750816345215 + }, + { + "auxiliary_loss_clip": 0.01260725, + "auxiliary_loss_mlp": 0.0104598, + "balance_loss_clip": 1.06646287, + "balance_loss_mlp": 1.0345602, + "epoch": 0.11483196056033187, + "flos": 22526476594560.0, + "grad_norm": 2.931115142913785, + "language_loss": 0.83696008, + "learning_rate": 3.9250742546324786e-06, + "loss": 0.86002707, + "num_input_tokens_seen": 20243015, + "step": 955, + "time_per_iteration": 4.5071635246276855 + }, + { + "auxiliary_loss_clip": 0.01258612, + "auxiliary_loss_mlp": 0.0103476, + "balance_loss_clip": 1.06466079, + "balance_loss_mlp": 1.02407289, + "epoch": 0.11495220345097096, + "flos": 28220292887040.0, + "grad_norm": 1.797071800973235, + "language_loss": 0.86991513, + "learning_rate": 3.924862890522683e-06, + "loss": 0.89284885, + "num_input_tokens_seen": 20263025, + "step": 956, + "time_per_iteration": 2.7354049682617188 + }, + { + "auxiliary_loss_clip": 0.01253214, + "auxiliary_loss_mlp": 0.01042597, + "balance_loss_clip": 1.06668806, + "balance_loss_mlp": 1.03103399, + "epoch": 0.11507244634161005, + "flos": 17492267704320.0, + "grad_norm": 2.0549642835110276, + "language_loss": 0.86311513, + "learning_rate": 3.9246512344127174e-06, + "loss": 0.88607323, + "num_input_tokens_seen": 20280685, + "step": 957, + "time_per_iteration": 2.6974925994873047 + }, + { + "auxiliary_loss_clip": 0.01261028, + "auxiliary_loss_mlp": 0.01038952, + "balance_loss_clip": 1.05706692, + "balance_loss_mlp": 1.02762759, + "epoch": 0.11519268923224914, + "flos": 22564937082240.0, + "grad_norm": 1.8728837764638229, + "language_loss": 0.81946671, + "learning_rate": 3.9244392863346895e-06, + "loss": 0.84246647, + "num_input_tokens_seen": 20300090, + "step": 958, + "time_per_iteration": 2.84867787361145 + }, + { + "auxiliary_loss_clip": 0.01267284, + "auxiliary_loss_mlp": 0.0104766, + "balance_loss_clip": 1.07245135, + "balance_loss_mlp": 1.03619194, + "epoch": 0.11531293212288823, + "flos": 16982839065600.0, + "grad_norm": 1.735677910081735, + "language_loss": 0.92428637, + "learning_rate": 3.9242270463207524e-06, + "loss": 0.9474358, + "num_input_tokens_seen": 20318480, + "step": 959, + "time_per_iteration": 2.7102365493774414 + }, + { + "auxiliary_loss_clip": 0.01257619, + "auxiliary_loss_mlp": 0.01037881, + "balance_loss_clip": 1.05864167, + "balance_loss_mlp": 1.02748585, + "epoch": 0.11543317501352733, + "flos": 12422004537600.0, + "grad_norm": 2.8552104257980737, + "language_loss": 0.85733342, + "learning_rate": 3.924014514403102e-06, + "loss": 0.88028842, + "num_input_tokens_seen": 20334635, + "step": 960, + "time_per_iteration": 2.7144618034362793 + }, + { + "auxiliary_loss_clip": 0.01260968, + "auxiliary_loss_mlp": 0.0104395, + "balance_loss_clip": 1.0581336, + "balance_loss_mlp": 1.03202939, + "epoch": 0.11555341790416641, + "flos": 19821648695040.0, + "grad_norm": 2.281597683960962, + "language_loss": 0.91095233, + "learning_rate": 3.92380169061398e-06, + "loss": 0.93400151, + "num_input_tokens_seen": 20352415, + "step": 961, + "time_per_iteration": 2.7422850131988525 + }, + { + "auxiliary_loss_clip": 0.01254499, + "auxiliary_loss_mlp": 0.01381675, + "balance_loss_clip": 1.05830324, + "balance_loss_mlp": 1.00030684, + "epoch": 0.11567366079480551, + "flos": 25738865625600.0, + "grad_norm": 2.081211076131072, + "language_loss": 0.84037817, + "learning_rate": 3.9235885749856705e-06, + "loss": 0.86673993, + "num_input_tokens_seen": 20371095, + "step": 962, + "time_per_iteration": 2.8004918098449707 + }, + { + "auxiliary_loss_clip": 0.01259679, + "auxiliary_loss_mlp": 0.01039574, + "balance_loss_clip": 1.0704, + "balance_loss_mlp": 1.02936423, + "epoch": 0.1157939036854446, + "flos": 18223301301120.0, + "grad_norm": 1.843101850625895, + "language_loss": 0.82616925, + "learning_rate": 3.9233751675505035e-06, + "loss": 0.84916174, + "num_input_tokens_seen": 20389805, + "step": 963, + "time_per_iteration": 2.8320541381835938 + }, + { + "auxiliary_loss_clip": 0.01256691, + "auxiliary_loss_mlp": 0.01044049, + "balance_loss_clip": 1.06817508, + "balance_loss_mlp": 1.03326607, + "epoch": 0.11591414657608369, + "flos": 23073755189760.0, + "grad_norm": 2.393935770865973, + "language_loss": 0.84810448, + "learning_rate": 3.923161468340853e-06, + "loss": 0.87111181, + "num_input_tokens_seen": 20409640, + "step": 964, + "time_per_iteration": 2.819632053375244 + }, + { + "auxiliary_loss_clip": 0.01251958, + "auxiliary_loss_mlp": 0.01039365, + "balance_loss_clip": 1.05537438, + "balance_loss_mlp": 1.0283618, + "epoch": 0.11603438946672277, + "flos": 19461716461440.0, + "grad_norm": 1.7430931493527728, + "language_loss": 0.81685883, + "learning_rate": 3.9229474773891374e-06, + "loss": 0.83977205, + "num_input_tokens_seen": 20428180, + "step": 965, + "time_per_iteration": 2.771916151046753 + }, + { + "auxiliary_loss_clip": 0.01272938, + "auxiliary_loss_mlp": 0.01029757, + "balance_loss_clip": 1.06064105, + "balance_loss_mlp": 1.0184921, + "epoch": 0.11615463235736187, + "flos": 26831986272000.0, + "grad_norm": 2.9073417411628104, + "language_loss": 0.83824188, + "learning_rate": 3.922733194727818e-06, + "loss": 0.86126876, + "num_input_tokens_seen": 20447975, + "step": 966, + "time_per_iteration": 2.7792046070098877 + }, + { + "auxiliary_loss_clip": 0.01259977, + "auxiliary_loss_mlp": 0.01043884, + "balance_loss_clip": 1.07080352, + "balance_loss_mlp": 1.03329253, + "epoch": 0.11627487524800097, + "flos": 18580324533120.0, + "grad_norm": 2.1157535898229543, + "language_loss": 0.87652731, + "learning_rate": 3.922518620389402e-06, + "loss": 0.89956594, + "num_input_tokens_seen": 20464840, + "step": 967, + "time_per_iteration": 2.628938674926758 + }, + { + "auxiliary_loss_clip": 0.01257453, + "auxiliary_loss_mlp": 0.0104989, + "balance_loss_clip": 1.05469728, + "balance_loss_mlp": 1.03908968, + "epoch": 0.11639511813864005, + "flos": 18150474476160.0, + "grad_norm": 1.717019913592421, + "language_loss": 0.89475131, + "learning_rate": 3.922303754406439e-06, + "loss": 0.91782475, + "num_input_tokens_seen": 20482680, + "step": 968, + "time_per_iteration": 2.8222908973693848 + }, + { + "auxiliary_loss_clip": 0.01241193, + "auxiliary_loss_mlp": 0.01038685, + "balance_loss_clip": 1.05790567, + "balance_loss_mlp": 1.02749145, + "epoch": 0.11651536102927915, + "flos": 20922023888640.0, + "grad_norm": 1.6803239743493972, + "language_loss": 0.78908831, + "learning_rate": 3.922088596811526e-06, + "loss": 0.81188703, + "num_input_tokens_seen": 20501810, + "step": 969, + "time_per_iteration": 2.659844160079956 + }, + { + "auxiliary_loss_clip": 0.01236284, + "auxiliary_loss_mlp": 0.01050075, + "balance_loss_clip": 1.06523776, + "balance_loss_mlp": 1.03753424, + "epoch": 0.11663560391991823, + "flos": 16508602776960.0, + "grad_norm": 2.4347891397588266, + "language_loss": 0.87103999, + "learning_rate": 3.9218731476373e-06, + "loss": 0.89390355, + "num_input_tokens_seen": 20517995, + "step": 970, + "time_per_iteration": 2.5871691703796387 + }, + { + "auxiliary_loss_clip": 0.01257276, + "auxiliary_loss_mlp": 0.01039936, + "balance_loss_clip": 1.07282782, + "balance_loss_mlp": 1.02706707, + "epoch": 0.11675584681055733, + "flos": 19865029345920.0, + "grad_norm": 1.9327789543336868, + "language_loss": 0.84411049, + "learning_rate": 3.9216574069164455e-06, + "loss": 0.86708254, + "num_input_tokens_seen": 20536970, + "step": 971, + "time_per_iteration": 2.5867161750793457 + }, + { + "auxiliary_loss_clip": 0.01241298, + "auxiliary_loss_mlp": 0.01038061, + "balance_loss_clip": 1.06931615, + "balance_loss_mlp": 1.02810669, + "epoch": 0.11687608970119642, + "flos": 21944364785280.0, + "grad_norm": 1.594112193821458, + "language_loss": 0.79998446, + "learning_rate": 3.921441374681691e-06, + "loss": 0.82277799, + "num_input_tokens_seen": 20557030, + "step": 972, + "time_per_iteration": 2.6475210189819336 + }, + { + "auxiliary_loss_clip": 0.0125355, + "auxiliary_loss_mlp": 0.010391, + "balance_loss_clip": 1.06523752, + "balance_loss_mlp": 1.02857375, + "epoch": 0.1169963325918355, + "flos": 24061155131520.0, + "grad_norm": 2.1750214543724233, + "language_loss": 0.64969939, + "learning_rate": 3.921225050965808e-06, + "loss": 0.6726259, + "num_input_tokens_seen": 20576915, + "step": 973, + "time_per_iteration": 2.6946330070495605 + }, + { + "auxiliary_loss_clip": 0.01266665, + "auxiliary_loss_mlp": 0.01036477, + "balance_loss_clip": 1.06396818, + "balance_loss_mlp": 1.02552152, + "epoch": 0.1171165754824746, + "flos": 23368151059200.0, + "grad_norm": 2.144333443838165, + "language_loss": 0.75103784, + "learning_rate": 3.921008435801612e-06, + "loss": 0.77406931, + "num_input_tokens_seen": 20596000, + "step": 974, + "time_per_iteration": 2.6955461502075195 + }, + { + "auxiliary_loss_clip": 0.01236199, + "auxiliary_loss_mlp": 0.01043149, + "balance_loss_clip": 1.06436801, + "balance_loss_mlp": 1.03234255, + "epoch": 0.11723681837311369, + "flos": 18552243075840.0, + "grad_norm": 4.970465985806789, + "language_loss": 0.75138056, + "learning_rate": 3.920791529221963e-06, + "loss": 0.77417409, + "num_input_tokens_seen": 20614675, + "step": 975, + "time_per_iteration": 2.6369242668151855 + }, + { + "auxiliary_loss_clip": 0.01260089, + "auxiliary_loss_mlp": 0.01381912, + "balance_loss_clip": 1.06619883, + "balance_loss_mlp": 1.00040138, + "epoch": 0.11735706126375278, + "flos": 23550541344000.0, + "grad_norm": 1.9340134384690897, + "language_loss": 0.76736104, + "learning_rate": 3.920574331259768e-06, + "loss": 0.79378104, + "num_input_tokens_seen": 20635875, + "step": 976, + "time_per_iteration": 2.881483316421509 + }, + { + "auxiliary_loss_clip": 0.01247947, + "auxiliary_loss_mlp": 0.01034587, + "balance_loss_clip": 1.06514072, + "balance_loss_mlp": 1.0237329, + "epoch": 0.11747730415439187, + "flos": 22381541216640.0, + "grad_norm": 4.578445800976889, + "language_loss": 0.79467684, + "learning_rate": 3.9203568419479716e-06, + "loss": 0.81750214, + "num_input_tokens_seen": 20656430, + "step": 977, + "time_per_iteration": 2.7102839946746826 + }, + { + "auxiliary_loss_clip": 0.01261179, + "auxiliary_loss_mlp": 0.0103338, + "balance_loss_clip": 1.06928086, + "balance_loss_mlp": 1.02306795, + "epoch": 0.11759754704503096, + "flos": 22200731130240.0, + "grad_norm": 1.8326622480535348, + "language_loss": 0.75265723, + "learning_rate": 3.92013906131957e-06, + "loss": 0.77560282, + "num_input_tokens_seen": 20675360, + "step": 978, + "time_per_iteration": 3.5355618000030518 + }, + { + "auxiliary_loss_clip": 0.01266877, + "auxiliary_loss_mlp": 0.01038796, + "balance_loss_clip": 1.06545997, + "balance_loss_mlp": 1.02821612, + "epoch": 0.11771778993567006, + "flos": 22309755886080.0, + "grad_norm": 1.5875094015343088, + "language_loss": 0.82741523, + "learning_rate": 3.9199209894076e-06, + "loss": 0.85047197, + "num_input_tokens_seen": 20695675, + "step": 979, + "time_per_iteration": 2.7133824825286865 + }, + { + "auxiliary_loss_clip": 0.01244653, + "auxiliary_loss_mlp": 0.01042942, + "balance_loss_clip": 1.06790829, + "balance_loss_mlp": 1.03080618, + "epoch": 0.11783803282630914, + "flos": 21288169175040.0, + "grad_norm": 1.867795951890497, + "language_loss": 0.90433061, + "learning_rate": 3.919702626245142e-06, + "loss": 0.92720658, + "num_input_tokens_seen": 20715330, + "step": 980, + "time_per_iteration": 2.65293025970459 + }, + { + "auxiliary_loss_clip": 0.01239776, + "auxiliary_loss_mlp": 0.01041059, + "balance_loss_clip": 1.0605942, + "balance_loss_mlp": 1.02884626, + "epoch": 0.11795827571694824, + "flos": 25371535190400.0, + "grad_norm": 2.24387416000974, + "language_loss": 0.66119772, + "learning_rate": 3.919483971865322e-06, + "loss": 0.68400609, + "num_input_tokens_seen": 20735325, + "step": 981, + "time_per_iteration": 4.570155620574951 + }, + { + "auxiliary_loss_clip": 0.01257752, + "auxiliary_loss_mlp": 0.01033325, + "balance_loss_clip": 1.06822157, + "balance_loss_mlp": 1.02334738, + "epoch": 0.11807851860758732, + "flos": 23622218933760.0, + "grad_norm": 3.5622269108884566, + "language_loss": 0.87871659, + "learning_rate": 3.91926502630131e-06, + "loss": 0.90162736, + "num_input_tokens_seen": 20755940, + "step": 982, + "time_per_iteration": 2.732025384902954 + }, + { + "auxiliary_loss_clip": 0.01259311, + "auxiliary_loss_mlp": 0.01046822, + "balance_loss_clip": 1.07288849, + "balance_loss_mlp": 1.03683233, + "epoch": 0.11819876149822642, + "flos": 24972496024320.0, + "grad_norm": 2.4345907902529715, + "language_loss": 0.72278106, + "learning_rate": 3.91904578958632e-06, + "loss": 0.7458424, + "num_input_tokens_seen": 20775355, + "step": 983, + "time_per_iteration": 2.6983113288879395 + }, + { + "auxiliary_loss_clip": 0.01242887, + "auxiliary_loss_mlp": 0.01040735, + "balance_loss_clip": 1.06850111, + "balance_loss_mlp": 1.03005433, + "epoch": 0.11831900438886551, + "flos": 23003226835200.0, + "grad_norm": 2.384813293158484, + "language_loss": 0.84374386, + "learning_rate": 3.918826261753608e-06, + "loss": 0.86658007, + "num_input_tokens_seen": 20794935, + "step": 984, + "time_per_iteration": 2.626990556716919 + }, + { + "auxiliary_loss_clip": 0.01260963, + "auxiliary_loss_mlp": 0.01043858, + "balance_loss_clip": 1.06688404, + "balance_loss_mlp": 1.03400517, + "epoch": 0.1184392472795046, + "flos": 27965147604480.0, + "grad_norm": 2.5295290632724328, + "language_loss": 0.71088582, + "learning_rate": 3.918606442836478e-06, + "loss": 0.73393404, + "num_input_tokens_seen": 20817155, + "step": 985, + "time_per_iteration": 2.732293128967285 + }, + { + "auxiliary_loss_clip": 0.01252481, + "auxiliary_loss_mlp": 0.01040885, + "balance_loss_clip": 1.07299614, + "balance_loss_mlp": 1.03036463, + "epoch": 0.1185594901701437, + "flos": 19898497843200.0, + "grad_norm": 1.8085706341585597, + "language_loss": 0.77590984, + "learning_rate": 3.918386332868277e-06, + "loss": 0.79884344, + "num_input_tokens_seen": 20835125, + "step": 986, + "time_per_iteration": 2.6542441844940186 + }, + { + "auxiliary_loss_clip": 0.01240091, + "auxiliary_loss_mlp": 0.01040305, + "balance_loss_clip": 1.06842208, + "balance_loss_mlp": 1.02912331, + "epoch": 0.11867973306078278, + "flos": 18912354877440.0, + "grad_norm": 1.7180215189488153, + "language_loss": 0.94582039, + "learning_rate": 3.918165931882394e-06, + "loss": 0.96862435, + "num_input_tokens_seen": 20853525, + "step": 987, + "time_per_iteration": 2.710793972015381 + }, + { + "auxiliary_loss_clip": 0.01268969, + "auxiliary_loss_mlp": 0.01043943, + "balance_loss_clip": 1.05759549, + "balance_loss_mlp": 1.03278458, + "epoch": 0.11879997595142187, + "flos": 16982803152000.0, + "grad_norm": 2.490451901620442, + "language_loss": 0.75365424, + "learning_rate": 3.917945239912264e-06, + "loss": 0.77678347, + "num_input_tokens_seen": 20871000, + "step": 988, + "time_per_iteration": 2.753286123275757 + }, + { + "auxiliary_loss_clip": 0.01275718, + "auxiliary_loss_mlp": 0.01041207, + "balance_loss_clip": 1.0635829, + "balance_loss_mlp": 1.03016794, + "epoch": 0.11892021884206096, + "flos": 17530369056000.0, + "grad_norm": 2.135268782138956, + "language_loss": 0.75993025, + "learning_rate": 3.917724256991367e-06, + "loss": 0.78309953, + "num_input_tokens_seen": 20889745, + "step": 989, + "time_per_iteration": 2.8192312717437744 + }, + { + "auxiliary_loss_clip": 0.01247019, + "auxiliary_loss_mlp": 0.01046741, + "balance_loss_clip": 1.06658506, + "balance_loss_mlp": 1.0350287, + "epoch": 0.11904046173270005, + "flos": 30955895763840.0, + "grad_norm": 2.0251057558465595, + "language_loss": 0.81475025, + "learning_rate": 3.9175029831532245e-06, + "loss": 0.83768785, + "num_input_tokens_seen": 20909260, + "step": 990, + "time_per_iteration": 2.787214994430542 + }, + { + "auxiliary_loss_clip": 0.01270343, + "auxiliary_loss_mlp": 0.01041123, + "balance_loss_clip": 1.06592345, + "balance_loss_mlp": 1.03037, + "epoch": 0.11916070462333915, + "flos": 20157234485760.0, + "grad_norm": 3.805562579542918, + "language_loss": 0.88623649, + "learning_rate": 3.917281418431404e-06, + "loss": 0.90935111, + "num_input_tokens_seen": 20928305, + "step": 991, + "time_per_iteration": 2.7224411964416504 + }, + { + "auxiliary_loss_clip": 0.01263134, + "auxiliary_loss_mlp": 0.01044144, + "balance_loss_clip": 1.0707314, + "balance_loss_mlp": 1.03253269, + "epoch": 0.11928094751397823, + "flos": 23551115961600.0, + "grad_norm": 1.8400563238226604, + "language_loss": 0.77099454, + "learning_rate": 3.917059562859516e-06, + "loss": 0.79406732, + "num_input_tokens_seen": 20947630, + "step": 992, + "time_per_iteration": 2.731330394744873 + }, + { + "auxiliary_loss_clip": 0.01251092, + "auxiliary_loss_mlp": 0.01040301, + "balance_loss_clip": 1.06800961, + "balance_loss_mlp": 1.0299592, + "epoch": 0.11940119040461733, + "flos": 23908426502400.0, + "grad_norm": 2.1215113933064043, + "language_loss": 0.88819939, + "learning_rate": 3.916837416471218e-06, + "loss": 0.91111332, + "num_input_tokens_seen": 20964250, + "step": 993, + "time_per_iteration": 2.7523977756500244 + }, + { + "auxiliary_loss_clip": 0.01239789, + "auxiliary_loss_mlp": 0.01048915, + "balance_loss_clip": 1.06191039, + "balance_loss_mlp": 1.03871632, + "epoch": 0.11952143329525641, + "flos": 13844533835520.0, + "grad_norm": 2.6334712357328884, + "language_loss": 0.72580922, + "learning_rate": 3.916614979300207e-06, + "loss": 0.74869627, + "num_input_tokens_seen": 20979095, + "step": 994, + "time_per_iteration": 2.6339502334594727 + }, + { + "auxiliary_loss_clip": 0.0126468, + "auxiliary_loss_mlp": 0.01032568, + "balance_loss_clip": 1.06449127, + "balance_loss_mlp": 1.02195215, + "epoch": 0.11964167618589551, + "flos": 27015525792000.0, + "grad_norm": 1.7049495509753887, + "language_loss": 0.78464818, + "learning_rate": 3.9163922513802274e-06, + "loss": 0.80762064, + "num_input_tokens_seen": 21001430, + "step": 995, + "time_per_iteration": 2.8908121585845947 + }, + { + "auxiliary_loss_clip": 0.01246607, + "auxiliary_loss_mlp": 0.01042832, + "balance_loss_clip": 1.06929731, + "balance_loss_mlp": 1.03122103, + "epoch": 0.1197619190765346, + "flos": 12567622273920.0, + "grad_norm": 2.836324253737529, + "language_loss": 0.82410961, + "learning_rate": 3.916169232745067e-06, + "loss": 0.84700406, + "num_input_tokens_seen": 21019105, + "step": 996, + "time_per_iteration": 2.6525423526763916 + }, + { + "auxiliary_loss_clip": 0.01244376, + "auxiliary_loss_mlp": 0.01043065, + "balance_loss_clip": 1.06308722, + "balance_loss_mlp": 1.03047013, + "epoch": 0.11988216196717369, + "flos": 16909437623040.0, + "grad_norm": 2.983121021670007, + "language_loss": 0.91840494, + "learning_rate": 3.915945923428559e-06, + "loss": 0.94127929, + "num_input_tokens_seen": 21035630, + "step": 997, + "time_per_iteration": 2.7078707218170166 + }, + { + "auxiliary_loss_clip": 0.01245555, + "auxiliary_loss_mlp": 0.01042971, + "balance_loss_clip": 1.0660069, + "balance_loss_mlp": 1.03212321, + "epoch": 0.12000240485781279, + "flos": 16216577205120.0, + "grad_norm": 2.10789243083597, + "language_loss": 0.83265424, + "learning_rate": 3.915722323464577e-06, + "loss": 0.85553956, + "num_input_tokens_seen": 21054235, + "step": 998, + "time_per_iteration": 2.770740032196045 + }, + { + "auxiliary_loss_clip": 0.01250939, + "auxiliary_loss_mlp": 0.01044544, + "balance_loss_clip": 1.06891418, + "balance_loss_mlp": 1.03314805, + "epoch": 0.12012264774845187, + "flos": 49344887525760.0, + "grad_norm": 2.25269204928778, + "language_loss": 0.7077893, + "learning_rate": 3.91549843288704e-06, + "loss": 0.73074412, + "num_input_tokens_seen": 21077915, + "step": 999, + "time_per_iteration": 2.9267094135284424 + }, + { + "auxiliary_loss_clip": 0.01264758, + "auxiliary_loss_mlp": 0.01382027, + "balance_loss_clip": 1.0614363, + "balance_loss_mlp": 1.0004375, + "epoch": 0.12024289063909097, + "flos": 26979435601920.0, + "grad_norm": 3.012533816121702, + "language_loss": 0.79268068, + "learning_rate": 3.915274251729916e-06, + "loss": 0.81914854, + "num_input_tokens_seen": 21099205, + "step": 1000, + "time_per_iteration": 2.802319049835205 + }, + { + "auxiliary_loss_clip": 0.01271503, + "auxiliary_loss_mlp": 0.01040894, + "balance_loss_clip": 1.06661308, + "balance_loss_mlp": 1.02909803, + "epoch": 0.12036313352973005, + "flos": 19537308633600.0, + "grad_norm": 2.1954062198830466, + "language_loss": 0.90018594, + "learning_rate": 3.91504978002721e-06, + "loss": 0.92330992, + "num_input_tokens_seen": 21118260, + "step": 1001, + "time_per_iteration": 2.7023653984069824 + }, + { + "auxiliary_loss_clip": 0.01259576, + "auxiliary_loss_mlp": 0.01381544, + "balance_loss_clip": 1.06380343, + "balance_loss_mlp": 1.00034833, + "epoch": 0.12048337642036915, + "flos": 17268256535040.0, + "grad_norm": 2.4856589624211183, + "language_loss": 0.76057434, + "learning_rate": 3.914825017812974e-06, + "loss": 0.78698552, + "num_input_tokens_seen": 21134910, + "step": 1002, + "time_per_iteration": 2.674286127090454 + }, + { + "auxiliary_loss_clip": 0.01259776, + "auxiliary_loss_mlp": 0.010359, + "balance_loss_clip": 1.06846738, + "balance_loss_mlp": 1.02495706, + "epoch": 0.12060361931100824, + "flos": 22856962654080.0, + "grad_norm": 2.8520398835387883, + "language_loss": 0.72675395, + "learning_rate": 3.9145999651213065e-06, + "loss": 0.74971074, + "num_input_tokens_seen": 21154150, + "step": 1003, + "time_per_iteration": 2.7072904109954834 + }, + { + "auxiliary_loss_clip": 0.01254949, + "auxiliary_loss_mlp": 0.01035464, + "balance_loss_clip": 1.0702312, + "balance_loss_mlp": 1.02387071, + "epoch": 0.12072386220164733, + "flos": 16726795943040.0, + "grad_norm": 2.141847365585141, + "language_loss": 0.88607681, + "learning_rate": 3.9143746219863465e-06, + "loss": 0.90898097, + "num_input_tokens_seen": 21171255, + "step": 1004, + "time_per_iteration": 3.7659332752227783 + }, + { + "auxiliary_loss_clip": 0.01162193, + "auxiliary_loss_mlp": 0.01003719, + "balance_loss_clip": 1.03444195, + "balance_loss_mlp": 1.00021386, + "epoch": 0.12084410509228642, + "flos": 55144176105600.0, + "grad_norm": 0.9496470467698884, + "language_loss": 0.64798599, + "learning_rate": 3.914148988442278e-06, + "loss": 0.66964513, + "num_input_tokens_seen": 21227045, + "step": 1005, + "time_per_iteration": 3.2765560150146484 + }, + { + "auxiliary_loss_clip": 0.01241507, + "auxiliary_loss_mlp": 0.01038024, + "balance_loss_clip": 1.06304789, + "balance_loss_mlp": 1.02713418, + "epoch": 0.1209643479829255, + "flos": 26760236855040.0, + "grad_norm": 2.9835520420396513, + "language_loss": 0.95065486, + "learning_rate": 3.91392306452333e-06, + "loss": 0.97345018, + "num_input_tokens_seen": 21244120, + "step": 1006, + "time_per_iteration": 2.7051568031311035 + }, + { + "auxiliary_loss_clip": 0.0124806, + "auxiliary_loss_mlp": 0.01043527, + "balance_loss_clip": 1.07036996, + "balance_loss_mlp": 1.03068209, + "epoch": 0.1210845908735646, + "flos": 11035026725760.0, + "grad_norm": 2.8632976578768066, + "language_loss": 0.65888321, + "learning_rate": 3.913696850263774e-06, + "loss": 0.68179905, + "num_input_tokens_seen": 21258485, + "step": 1007, + "time_per_iteration": 5.303524732589722 + }, + { + "auxiliary_loss_clip": 0.01247224, + "auxiliary_loss_mlp": 0.01039838, + "balance_loss_clip": 1.06603444, + "balance_loss_mlp": 1.02924013, + "epoch": 0.1212048337642037, + "flos": 20484631975680.0, + "grad_norm": 2.153808700575404, + "language_loss": 0.79259628, + "learning_rate": 3.913470345697929e-06, + "loss": 0.81546688, + "num_input_tokens_seen": 21277115, + "step": 1008, + "time_per_iteration": 2.9463326930999756 + }, + { + "auxiliary_loss_clip": 0.0127482, + "auxiliary_loss_mlp": 0.01038574, + "balance_loss_clip": 1.06279361, + "balance_loss_mlp": 1.02756548, + "epoch": 0.12132507665484278, + "flos": 22346061557760.0, + "grad_norm": 2.1067377526806177, + "language_loss": 0.85601974, + "learning_rate": 3.913243550860153e-06, + "loss": 0.87915373, + "num_input_tokens_seen": 21294880, + "step": 1009, + "time_per_iteration": 2.885068893432617 + }, + { + "auxiliary_loss_clip": 0.01256986, + "auxiliary_loss_mlp": 0.01041516, + "balance_loss_clip": 1.07544994, + "balance_loss_mlp": 1.028916, + "epoch": 0.12144531954548188, + "flos": 29314957818240.0, + "grad_norm": 2.2151750651582405, + "language_loss": 0.75669014, + "learning_rate": 3.913016465784852e-06, + "loss": 0.77967519, + "num_input_tokens_seen": 21315555, + "step": 1010, + "time_per_iteration": 2.72279953956604 + }, + { + "auxiliary_loss_clip": 0.01270923, + "auxiliary_loss_mlp": 0.01051552, + "balance_loss_clip": 1.06071758, + "balance_loss_mlp": 1.04031634, + "epoch": 0.12156556243612096, + "flos": 20485242506880.0, + "grad_norm": 3.232138503726439, + "language_loss": 0.72298682, + "learning_rate": 3.912789090506474e-06, + "loss": 0.74621165, + "num_input_tokens_seen": 21334815, + "step": 1011, + "time_per_iteration": 2.7878201007843018 + }, + { + "auxiliary_loss_clip": 0.01269317, + "auxiliary_loss_mlp": 0.01037918, + "balance_loss_clip": 1.06159258, + "balance_loss_mlp": 1.02671218, + "epoch": 0.12168580532676006, + "flos": 16472009796480.0, + "grad_norm": 2.6445816984953336, + "language_loss": 0.71555525, + "learning_rate": 3.9125614250595114e-06, + "loss": 0.73862761, + "num_input_tokens_seen": 21351025, + "step": 1012, + "time_per_iteration": 2.6739535331726074 + }, + { + "auxiliary_loss_clip": 0.01253992, + "auxiliary_loss_mlp": 0.01041618, + "balance_loss_clip": 1.06745315, + "balance_loss_mlp": 1.03041792, + "epoch": 0.12180604821739914, + "flos": 15341290588800.0, + "grad_norm": 2.8786634052987146, + "language_loss": 0.89356101, + "learning_rate": 3.912333469478502e-06, + "loss": 0.91651714, + "num_input_tokens_seen": 21368990, + "step": 1013, + "time_per_iteration": 2.6147091388702393 + }, + { + "auxiliary_loss_clip": 0.01253842, + "auxiliary_loss_mlp": 0.01035646, + "balance_loss_clip": 1.06389809, + "balance_loss_mlp": 1.02406502, + "epoch": 0.12192629110803824, + "flos": 19318038059520.0, + "grad_norm": 2.1061501560117035, + "language_loss": 0.7815547, + "learning_rate": 3.912105223798025e-06, + "loss": 0.80444956, + "num_input_tokens_seen": 21388410, + "step": 1014, + "time_per_iteration": 2.6781227588653564 + }, + { + "auxiliary_loss_clip": 0.01165564, + "auxiliary_loss_mlp": 0.01004448, + "balance_loss_clip": 1.02805483, + "balance_loss_mlp": 1.00111032, + "epoch": 0.12204653399867733, + "flos": 47725354085760.0, + "grad_norm": 0.9926705067009104, + "language_loss": 0.6768719, + "learning_rate": 3.9118766880527065e-06, + "loss": 0.69857204, + "num_input_tokens_seen": 21442845, + "step": 1015, + "time_per_iteration": 3.1683809757232666 + }, + { + "auxiliary_loss_clip": 0.01263416, + "auxiliary_loss_mlp": 0.01037143, + "balance_loss_clip": 1.05915976, + "balance_loss_mlp": 1.02659285, + "epoch": 0.12216677688931642, + "flos": 18221936584320.0, + "grad_norm": 1.7162499545496328, + "language_loss": 0.73562384, + "learning_rate": 3.9116478622772145e-06, + "loss": 0.75862944, + "num_input_tokens_seen": 21461420, + "step": 1016, + "time_per_iteration": 2.7718567848205566 + }, + { + "auxiliary_loss_clip": 0.01248918, + "auxiliary_loss_mlp": 0.01046654, + "balance_loss_clip": 1.06813049, + "balance_loss_mlp": 1.03574657, + "epoch": 0.12228701977995551, + "flos": 27525636789120.0, + "grad_norm": 1.9289746967869454, + "language_loss": 0.87995374, + "learning_rate": 3.911418746506261e-06, + "loss": 0.90290952, + "num_input_tokens_seen": 21481550, + "step": 1017, + "time_per_iteration": 2.7258951663970947 + }, + { + "auxiliary_loss_clip": 0.01256981, + "auxiliary_loss_mlp": 0.01042482, + "balance_loss_clip": 1.07536113, + "balance_loss_mlp": 1.03167593, + "epoch": 0.1224072626705946, + "flos": 21798136517760.0, + "grad_norm": 1.6740252426071747, + "language_loss": 0.78597128, + "learning_rate": 3.911189340774604e-06, + "loss": 0.80896592, + "num_input_tokens_seen": 21501680, + "step": 1018, + "time_per_iteration": 2.652629852294922 + }, + { + "auxiliary_loss_clip": 0.0126352, + "auxiliary_loss_mlp": 0.01039265, + "balance_loss_clip": 1.06376123, + "balance_loss_mlp": 1.02913201, + "epoch": 0.1225275055612337, + "flos": 20703758895360.0, + "grad_norm": 1.8777934882732357, + "language_loss": 0.79624879, + "learning_rate": 3.910959645117043e-06, + "loss": 0.81927669, + "num_input_tokens_seen": 21521015, + "step": 1019, + "time_per_iteration": 2.6377971172332764 + }, + { + "auxiliary_loss_clip": 0.01151869, + "auxiliary_loss_mlp": 0.01377487, + "balance_loss_clip": 1.031533, + "balance_loss_mlp": 1.00029373, + "epoch": 0.12264774845187278, + "flos": 57745294462080.0, + "grad_norm": 0.812240383496602, + "language_loss": 0.56725287, + "learning_rate": 3.910729659568423e-06, + "loss": 0.59254646, + "num_input_tokens_seen": 21578200, + "step": 1020, + "time_per_iteration": 3.2371087074279785 + }, + { + "auxiliary_loss_clip": 0.01262535, + "auxiliary_loss_mlp": 0.01034531, + "balance_loss_clip": 1.07040429, + "balance_loss_mlp": 1.0247736, + "epoch": 0.12276799134251187, + "flos": 26396282298240.0, + "grad_norm": 1.732924290925205, + "language_loss": 0.82176316, + "learning_rate": 3.9104993841636344e-06, + "loss": 0.84473383, + "num_input_tokens_seen": 21598770, + "step": 1021, + "time_per_iteration": 2.6943492889404297 + }, + { + "auxiliary_loss_clip": 0.01258043, + "auxiliary_loss_mlp": 0.01381449, + "balance_loss_clip": 1.07192135, + "balance_loss_mlp": 1.00020707, + "epoch": 0.12288823423315097, + "flos": 21064193919360.0, + "grad_norm": 1.8607977311080213, + "language_loss": 0.80802071, + "learning_rate": 3.910268818937608e-06, + "loss": 0.83441567, + "num_input_tokens_seen": 21616925, + "step": 1022, + "time_per_iteration": 2.734934091567993 + }, + { + "auxiliary_loss_clip": 0.01274935, + "auxiliary_loss_mlp": 0.010359, + "balance_loss_clip": 1.06590366, + "balance_loss_mlp": 1.02599335, + "epoch": 0.12300847712379005, + "flos": 12312441077760.0, + "grad_norm": 2.920337283467533, + "language_loss": 0.87456298, + "learning_rate": 3.9100379639253196e-06, + "loss": 0.89767122, + "num_input_tokens_seen": 21633645, + "step": 1023, + "time_per_iteration": 2.7210261821746826 + }, + { + "auxiliary_loss_clip": 0.01252746, + "auxiliary_loss_mlp": 0.01037505, + "balance_loss_clip": 1.06107235, + "balance_loss_mlp": 1.02569127, + "epoch": 0.12312872001442915, + "flos": 16762239688320.0, + "grad_norm": 2.690924365038752, + "language_loss": 0.86348379, + "learning_rate": 3.909806819161791e-06, + "loss": 0.88638628, + "num_input_tokens_seen": 21649120, + "step": 1024, + "time_per_iteration": 2.65145206451416 + }, + { + "auxiliary_loss_clip": 0.0126772, + "auxiliary_loss_mlp": 0.01037432, + "balance_loss_clip": 1.06205869, + "balance_loss_mlp": 1.02584469, + "epoch": 0.12324896290506823, + "flos": 18404937400320.0, + "grad_norm": 2.3437524409132053, + "language_loss": 0.86368853, + "learning_rate": 3.909575384682086e-06, + "loss": 0.88674003, + "num_input_tokens_seen": 21668000, + "step": 1025, + "time_per_iteration": 2.6937172412872314 + }, + { + "auxiliary_loss_clip": 0.01252159, + "auxiliary_loss_mlp": 0.01035594, + "balance_loss_clip": 1.06659126, + "balance_loss_mlp": 1.02586031, + "epoch": 0.12336920579570733, + "flos": 18915407533440.0, + "grad_norm": 1.7325301820826076, + "language_loss": 0.69307119, + "learning_rate": 3.9093436605213144e-06, + "loss": 0.7159487, + "num_input_tokens_seen": 21688500, + "step": 1026, + "time_per_iteration": 2.6763064861297607 + }, + { + "auxiliary_loss_clip": 0.01258492, + "auxiliary_loss_mlp": 0.01036079, + "balance_loss_clip": 1.0657835, + "balance_loss_mlp": 1.02613139, + "epoch": 0.12348944868634643, + "flos": 23878369797120.0, + "grad_norm": 1.777024992846186, + "language_loss": 0.79524541, + "learning_rate": 3.909111646714627e-06, + "loss": 0.81819111, + "num_input_tokens_seen": 21709345, + "step": 1027, + "time_per_iteration": 2.7150442600250244 + }, + { + "auxiliary_loss_clip": 0.01238508, + "auxiliary_loss_mlp": 0.01034519, + "balance_loss_clip": 1.06655765, + "balance_loss_mlp": 1.02445173, + "epoch": 0.12360969157698551, + "flos": 19026084314880.0, + "grad_norm": 2.16416444486176, + "language_loss": 0.72452223, + "learning_rate": 3.9088793432972206e-06, + "loss": 0.74725252, + "num_input_tokens_seen": 21728165, + "step": 1028, + "time_per_iteration": 2.7618846893310547 + }, + { + "auxiliary_loss_clip": 0.01275959, + "auxiliary_loss_mlp": 0.0104271, + "balance_loss_clip": 1.06614566, + "balance_loss_mlp": 1.03282762, + "epoch": 0.1237299344676246, + "flos": 13224607983360.0, + "grad_norm": 2.1612981027493845, + "language_loss": 0.82351637, + "learning_rate": 3.908646750304336e-06, + "loss": 0.84670311, + "num_input_tokens_seen": 21745850, + "step": 1029, + "time_per_iteration": 2.7115659713745117 + }, + { + "auxiliary_loss_clip": 0.01264065, + "auxiliary_loss_mlp": 0.01046144, + "balance_loss_clip": 1.07017326, + "balance_loss_mlp": 1.03542697, + "epoch": 0.12385017735826369, + "flos": 20485673470080.0, + "grad_norm": 1.5849648489286297, + "language_loss": 0.87282151, + "learning_rate": 3.908413867771257e-06, + "loss": 0.89592355, + "num_input_tokens_seen": 21764760, + "step": 1030, + "time_per_iteration": 3.7637252807617188 + }, + { + "auxiliary_loss_clip": 0.01250986, + "auxiliary_loss_mlp": 0.01044307, + "balance_loss_clip": 1.06937814, + "balance_loss_mlp": 1.03302968, + "epoch": 0.12397042024890279, + "flos": 17347835116800.0, + "grad_norm": 1.850193369082683, + "language_loss": 0.80396831, + "learning_rate": 3.908180695733311e-06, + "loss": 0.82692122, + "num_input_tokens_seen": 21784250, + "step": 1031, + "time_per_iteration": 2.683992385864258 + }, + { + "auxiliary_loss_clip": 0.01241191, + "auxiliary_loss_mlp": 0.01044247, + "balance_loss_clip": 1.05303001, + "balance_loss_mlp": 1.03341031, + "epoch": 0.12409066313954187, + "flos": 20412343854720.0, + "grad_norm": 1.958307600408251, + "language_loss": 0.83217406, + "learning_rate": 3.907947234225871e-06, + "loss": 0.85502839, + "num_input_tokens_seen": 21803260, + "step": 1032, + "time_per_iteration": 3.652454137802124 + }, + { + "auxiliary_loss_clip": 0.01272213, + "auxiliary_loss_mlp": 0.01038491, + "balance_loss_clip": 1.05878532, + "balance_loss_mlp": 1.02820897, + "epoch": 0.12421090603018096, + "flos": 20736688688640.0, + "grad_norm": 1.9206510442628126, + "language_loss": 0.87118924, + "learning_rate": 3.907713483284352e-06, + "loss": 0.89429629, + "num_input_tokens_seen": 21822735, + "step": 1033, + "time_per_iteration": 2.7427637577056885 + }, + { + "auxiliary_loss_clip": 0.01270709, + "auxiliary_loss_mlp": 0.01042238, + "balance_loss_clip": 1.05246687, + "balance_loss_mlp": 1.03068662, + "epoch": 0.12433114892082006, + "flos": 24498834353280.0, + "grad_norm": 2.4295507159543237, + "language_loss": 0.97064245, + "learning_rate": 3.907479442944216e-06, + "loss": 0.99377191, + "num_input_tokens_seen": 21841140, + "step": 1034, + "time_per_iteration": 3.7658567428588867 + }, + { + "auxiliary_loss_clip": 0.0125071, + "auxiliary_loss_mlp": 0.01035193, + "balance_loss_clip": 1.06999516, + "balance_loss_mlp": 1.02501261, + "epoch": 0.12445139181145914, + "flos": 19682315838720.0, + "grad_norm": 3.1657211380225596, + "language_loss": 0.92698133, + "learning_rate": 3.907245113240963e-06, + "loss": 0.94984043, + "num_input_tokens_seen": 21859260, + "step": 1035, + "time_per_iteration": 2.601335287094116 + }, + { + "auxiliary_loss_clip": 0.01260762, + "auxiliary_loss_mlp": 0.01042702, + "balance_loss_clip": 1.05936241, + "balance_loss_mlp": 1.03115082, + "epoch": 0.12457163470209824, + "flos": 46423087522560.0, + "grad_norm": 2.025905843778734, + "language_loss": 0.73677337, + "learning_rate": 3.907010494210144e-06, + "loss": 0.759808, + "num_input_tokens_seen": 21881920, + "step": 1036, + "time_per_iteration": 2.97920298576355 + }, + { + "auxiliary_loss_clip": 0.01254218, + "auxiliary_loss_mlp": 0.01039099, + "balance_loss_clip": 1.06884098, + "balance_loss_mlp": 1.02805996, + "epoch": 0.12469187759273732, + "flos": 20376289578240.0, + "grad_norm": 1.9485910768637458, + "language_loss": 0.91991949, + "learning_rate": 3.9067755858873495e-06, + "loss": 0.94285268, + "num_input_tokens_seen": 21898720, + "step": 1037, + "time_per_iteration": 2.6125948429107666 + }, + { + "auxiliary_loss_clip": 0.01159944, + "auxiliary_loss_mlp": 0.01010437, + "balance_loss_clip": 1.02844787, + "balance_loss_mlp": 1.0076592, + "epoch": 0.12481212048337642, + "flos": 69224641447680.0, + "grad_norm": 0.9087933320732766, + "language_loss": 0.62866735, + "learning_rate": 3.906540388308214e-06, + "loss": 0.65037119, + "num_input_tokens_seen": 21958305, + "step": 1038, + "time_per_iteration": 3.265657663345337 + }, + { + "auxiliary_loss_clip": 0.01255997, + "auxiliary_loss_mlp": 0.010402, + "balance_loss_clip": 1.06081891, + "balance_loss_mlp": 1.02951908, + "epoch": 0.12493236337401552, + "flos": 18223696350720.0, + "grad_norm": 1.7825836925517353, + "language_loss": 0.81465566, + "learning_rate": 3.906304901508417e-06, + "loss": 0.83761764, + "num_input_tokens_seen": 21977205, + "step": 1039, + "time_per_iteration": 2.7658417224884033 + }, + { + "auxiliary_loss_clip": 0.01257753, + "auxiliary_loss_mlp": 0.01038096, + "balance_loss_clip": 1.07357621, + "balance_loss_mlp": 1.02707529, + "epoch": 0.12505260626465461, + "flos": 30044375303040.0, + "grad_norm": 2.7792747318257867, + "language_loss": 0.75514758, + "learning_rate": 3.9060691255236835e-06, + "loss": 0.77810603, + "num_input_tokens_seen": 21997770, + "step": 1040, + "time_per_iteration": 2.6775660514831543 + }, + { + "auxiliary_loss_clip": 0.01243744, + "auxiliary_loss_mlp": 0.01040653, + "balance_loss_clip": 1.06305289, + "balance_loss_mlp": 1.02951336, + "epoch": 0.1251728491552937, + "flos": 24433980347520.0, + "grad_norm": 2.018545497336916, + "language_loss": 0.80885226, + "learning_rate": 3.905833060389778e-06, + "loss": 0.83169627, + "num_input_tokens_seen": 22021890, + "step": 1041, + "time_per_iteration": 2.744475841522217 + }, + { + "auxiliary_loss_clip": 0.01245069, + "auxiliary_loss_mlp": 0.01381352, + "balance_loss_clip": 1.07101142, + "balance_loss_mlp": 1.0000391, + "epoch": 0.12529309204593278, + "flos": 27119809952640.0, + "grad_norm": 3.2814283488167746, + "language_loss": 0.78110188, + "learning_rate": 3.905596706142513e-06, + "loss": 0.80736613, + "num_input_tokens_seen": 22043300, + "step": 1042, + "time_per_iteration": 2.7112302780151367 + }, + { + "auxiliary_loss_clip": 0.01254136, + "auxiliary_loss_mlp": 0.01038074, + "balance_loss_clip": 1.059286, + "balance_loss_mlp": 1.0268445, + "epoch": 0.12541333493657186, + "flos": 30774151923840.0, + "grad_norm": 1.9863843583257372, + "language_loss": 0.86271465, + "learning_rate": 3.9053600628177435e-06, + "loss": 0.88563681, + "num_input_tokens_seen": 22062910, + "step": 1043, + "time_per_iteration": 2.7979841232299805 + }, + { + "auxiliary_loss_clip": 0.01240889, + "auxiliary_loss_mlp": 0.01038531, + "balance_loss_clip": 1.06789422, + "balance_loss_mlp": 1.02732527, + "epoch": 0.12553357782721097, + "flos": 23659566099840.0, + "grad_norm": 3.5603822986855977, + "language_loss": 0.84371603, + "learning_rate": 3.905123130451367e-06, + "loss": 0.86651027, + "num_input_tokens_seen": 22084010, + "step": 1044, + "time_per_iteration": 2.650315284729004 + }, + { + "auxiliary_loss_clip": 0.012418, + "auxiliary_loss_mlp": 0.01038117, + "balance_loss_clip": 1.06865954, + "balance_loss_mlp": 1.02698255, + "epoch": 0.12565382071785006, + "flos": 24863758577280.0, + "grad_norm": 1.6978993206145099, + "language_loss": 0.79006171, + "learning_rate": 3.904885909079326e-06, + "loss": 0.81286091, + "num_input_tokens_seen": 22102795, + "step": 1045, + "time_per_iteration": 2.652986764907837 + }, + { + "auxiliary_loss_clip": 0.01250575, + "auxiliary_loss_mlp": 0.01040085, + "balance_loss_clip": 1.06681728, + "balance_loss_mlp": 1.02868271, + "epoch": 0.12577406360848914, + "flos": 21360780518400.0, + "grad_norm": 3.095605380104854, + "language_loss": 0.77936149, + "learning_rate": 3.904648398737607e-06, + "loss": 0.80226815, + "num_input_tokens_seen": 22121360, + "step": 1046, + "time_per_iteration": 2.674622058868408 + }, + { + "auxiliary_loss_clip": 0.01239146, + "auxiliary_loss_mlp": 0.01042022, + "balance_loss_clip": 1.06644082, + "balance_loss_mlp": 1.03135312, + "epoch": 0.12589430649912825, + "flos": 36138056774400.0, + "grad_norm": 2.7280755598456725, + "language_loss": 0.78088975, + "learning_rate": 3.9044105994622406e-06, + "loss": 0.8037014, + "num_input_tokens_seen": 22142505, + "step": 1047, + "time_per_iteration": 2.704763174057007 + }, + { + "auxiliary_loss_clip": 0.01260045, + "auxiliary_loss_mlp": 0.01381559, + "balance_loss_clip": 1.06532764, + "balance_loss_mlp": 1.00003219, + "epoch": 0.12601454938976733, + "flos": 25337671643520.0, + "grad_norm": 2.010626230615255, + "language_loss": 0.81658304, + "learning_rate": 3.9041725112893005e-06, + "loss": 0.8429991, + "num_input_tokens_seen": 22163730, + "step": 1048, + "time_per_iteration": 2.744621515274048 + }, + { + "auxiliary_loss_clip": 0.01261352, + "auxiliary_loss_mlp": 0.01037598, + "balance_loss_clip": 1.06546056, + "balance_loss_mlp": 1.02615976, + "epoch": 0.12613479228040642, + "flos": 15560094286080.0, + "grad_norm": 1.774632066869547, + "language_loss": 0.75092864, + "learning_rate": 3.903934134254904e-06, + "loss": 0.77391815, + "num_input_tokens_seen": 22181520, + "step": 1049, + "time_per_iteration": 2.693035364151001 + }, + { + "auxiliary_loss_clip": 0.01254155, + "auxiliary_loss_mlp": 0.01040964, + "balance_loss_clip": 1.06632757, + "balance_loss_mlp": 1.02944875, + "epoch": 0.1262550351710455, + "flos": 21470595373440.0, + "grad_norm": 2.13218498194485, + "language_loss": 0.84849739, + "learning_rate": 3.903695468395213e-06, + "loss": 0.87144858, + "num_input_tokens_seen": 22199390, + "step": 1050, + "time_per_iteration": 2.670666217803955 + }, + { + "auxiliary_loss_clip": 0.01260057, + "auxiliary_loss_mlp": 0.01033103, + "balance_loss_clip": 1.06296563, + "balance_loss_mlp": 1.02232063, + "epoch": 0.1263752780616846, + "flos": 31576719456000.0, + "grad_norm": 5.523616560033191, + "language_loss": 0.55776829, + "learning_rate": 3.903456513746434e-06, + "loss": 0.58069986, + "num_input_tokens_seen": 22220365, + "step": 1051, + "time_per_iteration": 2.7382044792175293 + }, + { + "auxiliary_loss_clip": 0.01242198, + "auxiliary_loss_mlp": 0.01042438, + "balance_loss_clip": 1.07002139, + "balance_loss_mlp": 1.03157842, + "epoch": 0.1264955209523237, + "flos": 28768217927040.0, + "grad_norm": 1.7918155237931457, + "language_loss": 0.87484384, + "learning_rate": 3.903217270344815e-06, + "loss": 0.89769018, + "num_input_tokens_seen": 22240615, + "step": 1052, + "time_per_iteration": 2.6897079944610596 + }, + { + "auxiliary_loss_clip": 0.01258259, + "auxiliary_loss_mlp": 0.01034987, + "balance_loss_clip": 1.06150258, + "balance_loss_mlp": 1.02493739, + "epoch": 0.12661576384296278, + "flos": 29241125412480.0, + "grad_norm": 1.9490286296726687, + "language_loss": 0.82389462, + "learning_rate": 3.902977738226648e-06, + "loss": 0.84682703, + "num_input_tokens_seen": 22261350, + "step": 1053, + "time_per_iteration": 2.7486908435821533 + }, + { + "auxiliary_loss_clip": 0.01249996, + "auxiliary_loss_mlp": 0.01040592, + "balance_loss_clip": 1.06818271, + "balance_loss_mlp": 1.02937412, + "epoch": 0.12673600673360189, + "flos": 20850346298880.0, + "grad_norm": 1.8418939454405658, + "language_loss": 0.91140544, + "learning_rate": 3.902737917428273e-06, + "loss": 0.93431133, + "num_input_tokens_seen": 22279515, + "step": 1054, + "time_per_iteration": 2.6559078693389893 + }, + { + "auxiliary_loss_clip": 0.01242106, + "auxiliary_loss_mlp": 0.01045412, + "balance_loss_clip": 1.06711364, + "balance_loss_mlp": 1.03504062, + "epoch": 0.12685624962424097, + "flos": 25263695583360.0, + "grad_norm": 2.179261487584238, + "language_loss": 0.83831549, + "learning_rate": 3.902497807986068e-06, + "loss": 0.86119062, + "num_input_tokens_seen": 22299535, + "step": 1055, + "time_per_iteration": 2.6946637630462646 + }, + { + "auxiliary_loss_clip": 0.01262883, + "auxiliary_loss_mlp": 0.01040081, + "balance_loss_clip": 1.05860209, + "balance_loss_mlp": 1.02876806, + "epoch": 0.12697649251488005, + "flos": 27527109246720.0, + "grad_norm": 1.8566360175185332, + "language_loss": 0.84015524, + "learning_rate": 3.902257409936458e-06, + "loss": 0.86318493, + "num_input_tokens_seen": 22320300, + "step": 1056, + "time_per_iteration": 3.824070453643799 + }, + { + "auxiliary_loss_clip": 0.01261634, + "auxiliary_loss_mlp": 0.01043122, + "balance_loss_clip": 1.0696919, + "balance_loss_mlp": 1.03263116, + "epoch": 0.12709673540551916, + "flos": 21251863503360.0, + "grad_norm": 2.0718601261908267, + "language_loss": 0.83934903, + "learning_rate": 3.902016723315912e-06, + "loss": 0.8623966, + "num_input_tokens_seen": 22338240, + "step": 1057, + "time_per_iteration": 2.7346866130828857 + }, + { + "auxiliary_loss_clip": 0.01245651, + "auxiliary_loss_mlp": 0.01040834, + "balance_loss_clip": 1.06501889, + "balance_loss_mlp": 1.03042746, + "epoch": 0.12721697829615825, + "flos": 25337707557120.0, + "grad_norm": 2.689678430953966, + "language_loss": 0.69425046, + "learning_rate": 3.901775748160941e-06, + "loss": 0.71711528, + "num_input_tokens_seen": 22357420, + "step": 1058, + "time_per_iteration": 4.5397703647613525 + }, + { + "auxiliary_loss_clip": 0.01145942, + "auxiliary_loss_mlp": 0.01002972, + "balance_loss_clip": 1.03092599, + "balance_loss_mlp": 0.99993175, + "epoch": 0.12733722118679733, + "flos": 61943287754880.0, + "grad_norm": 0.8010948692554506, + "language_loss": 0.60861385, + "learning_rate": 3.901534484508101e-06, + "loss": 0.63010299, + "num_input_tokens_seen": 22420095, + "step": 1059, + "time_per_iteration": 4.070290565490723 + }, + { + "auxiliary_loss_clip": 0.01240635, + "auxiliary_loss_mlp": 0.01038085, + "balance_loss_clip": 1.0596354, + "balance_loss_mlp": 1.0274514, + "epoch": 0.1274574640774364, + "flos": 26976742081920.0, + "grad_norm": 1.9668266086030801, + "language_loss": 0.74714172, + "learning_rate": 3.901292932393991e-06, + "loss": 0.76992893, + "num_input_tokens_seen": 22438975, + "step": 1060, + "time_per_iteration": 2.721855878829956 + }, + { + "auxiliary_loss_clip": 0.01245789, + "auxiliary_loss_mlp": 0.01044303, + "balance_loss_clip": 1.07160342, + "balance_loss_mlp": 1.03356791, + "epoch": 0.12757770696807552, + "flos": 22236318529920.0, + "grad_norm": 2.6782606515239524, + "language_loss": 0.85384679, + "learning_rate": 3.9010510918552555e-06, + "loss": 0.87674773, + "num_input_tokens_seen": 22458050, + "step": 1061, + "time_per_iteration": 2.656593084335327 + }, + { + "auxiliary_loss_clip": 0.01249138, + "auxiliary_loss_mlp": 0.01041863, + "balance_loss_clip": 1.05943263, + "balance_loss_mlp": 1.02972722, + "epoch": 0.1276979498587146, + "flos": 28547905858560.0, + "grad_norm": 2.2296618315343864, + "language_loss": 0.74711204, + "learning_rate": 3.900808962928581e-06, + "loss": 0.77002203, + "num_input_tokens_seen": 22475665, + "step": 1062, + "time_per_iteration": 2.684951066970825 + }, + { + "auxiliary_loss_clip": 0.01241749, + "auxiliary_loss_mlp": 0.0105139, + "balance_loss_clip": 1.07077706, + "balance_loss_mlp": 1.0402323, + "epoch": 0.1278181927493537, + "flos": 17420338719360.0, + "grad_norm": 2.3623308535827565, + "language_loss": 0.89312118, + "learning_rate": 3.900566545650698e-06, + "loss": 0.91605258, + "num_input_tokens_seen": 22493335, + "step": 1063, + "time_per_iteration": 2.6266915798187256 + }, + { + "auxiliary_loss_clip": 0.01249052, + "auxiliary_loss_mlp": 0.01036881, + "balance_loss_clip": 1.06740189, + "balance_loss_mlp": 1.02640271, + "epoch": 0.1279384356399928, + "flos": 21138636856320.0, + "grad_norm": 2.4307782931190425, + "language_loss": 0.81856406, + "learning_rate": 3.900323840058381e-06, + "loss": 0.84142339, + "num_input_tokens_seen": 22511045, + "step": 1064, + "time_per_iteration": 2.628563404083252 + }, + { + "auxiliary_loss_clip": 0.0124727, + "auxiliary_loss_mlp": 0.01039501, + "balance_loss_clip": 1.06396294, + "balance_loss_mlp": 1.02952373, + "epoch": 0.12805867853063188, + "flos": 26576733248640.0, + "grad_norm": 2.048143087430973, + "language_loss": 0.81672794, + "learning_rate": 3.900080846188449e-06, + "loss": 0.83959568, + "num_input_tokens_seen": 22529635, + "step": 1065, + "time_per_iteration": 2.7164511680603027 + }, + { + "auxiliary_loss_clip": 0.01239712, + "auxiliary_loss_mlp": 0.01035038, + "balance_loss_clip": 1.06742859, + "balance_loss_mlp": 1.02440429, + "epoch": 0.12817892142127096, + "flos": 16436206915200.0, + "grad_norm": 1.8873979676283914, + "language_loss": 0.81495082, + "learning_rate": 3.8998375640777625e-06, + "loss": 0.83769834, + "num_input_tokens_seen": 22547505, + "step": 1066, + "time_per_iteration": 2.5683505535125732 + }, + { + "auxiliary_loss_clip": 0.01148217, + "auxiliary_loss_mlp": 0.01005099, + "balance_loss_clip": 1.03253245, + "balance_loss_mlp": 1.00233293, + "epoch": 0.12829916431191005, + "flos": 60757049099520.0, + "grad_norm": 0.7050025579724136, + "language_loss": 0.52638388, + "learning_rate": 3.899593993763229e-06, + "loss": 0.54791701, + "num_input_tokens_seen": 22608465, + "step": 1067, + "time_per_iteration": 3.222365140914917 + }, + { + "auxiliary_loss_clip": 0.01255608, + "auxiliary_loss_mlp": 0.01041477, + "balance_loss_clip": 1.06049645, + "balance_loss_mlp": 1.02903116, + "epoch": 0.12841940720254916, + "flos": 29786895636480.0, + "grad_norm": 2.3339432669671716, + "language_loss": 0.81232351, + "learning_rate": 3.899350135281796e-06, + "loss": 0.83529437, + "num_input_tokens_seen": 22629465, + "step": 1068, + "time_per_iteration": 2.7279200553894043 + }, + { + "auxiliary_loss_clip": 0.01261618, + "auxiliary_loss_mlp": 0.01044518, + "balance_loss_clip": 1.06163454, + "balance_loss_mlp": 1.03408146, + "epoch": 0.12853965009318824, + "flos": 25951851319680.0, + "grad_norm": 1.8778820140856642, + "language_loss": 0.79729533, + "learning_rate": 3.8991059886704585e-06, + "loss": 0.82035673, + "num_input_tokens_seen": 22648970, + "step": 1069, + "time_per_iteration": 2.780513048171997 + }, + { + "auxiliary_loss_clip": 0.01250504, + "auxiliary_loss_mlp": 0.01034159, + "balance_loss_clip": 1.05958104, + "balance_loss_mlp": 1.02357888, + "epoch": 0.12865989298382732, + "flos": 30846871008000.0, + "grad_norm": 2.3163369582284754, + "language_loss": 0.829759, + "learning_rate": 3.898861553966252e-06, + "loss": 0.85260558, + "num_input_tokens_seen": 22668620, + "step": 1070, + "time_per_iteration": 2.816542863845825 + }, + { + "auxiliary_loss_clip": 0.01278987, + "auxiliary_loss_mlp": 0.0104639, + "balance_loss_clip": 1.05264211, + "balance_loss_mlp": 1.03595972, + "epoch": 0.12878013587446643, + "flos": 25885776251520.0, + "grad_norm": 1.6644501570069516, + "language_loss": 0.88047826, + "learning_rate": 3.898616831206257e-06, + "loss": 0.90373206, + "num_input_tokens_seen": 22689045, + "step": 1071, + "time_per_iteration": 2.9690768718719482 + }, + { + "auxiliary_loss_clip": 0.0125506, + "auxiliary_loss_mlp": 0.01040749, + "balance_loss_clip": 1.05784154, + "balance_loss_mlp": 1.02928114, + "epoch": 0.12890037876510552, + "flos": 23333138277120.0, + "grad_norm": 3.3534735508869717, + "language_loss": 0.76991552, + "learning_rate": 3.8983718204276e-06, + "loss": 0.79287356, + "num_input_tokens_seen": 22711265, + "step": 1072, + "time_per_iteration": 3.1078381538391113 + }, + { + "auxiliary_loss_clip": 0.01255186, + "auxiliary_loss_mlp": 0.01040573, + "balance_loss_clip": 1.06367111, + "balance_loss_mlp": 1.0294683, + "epoch": 0.1290206216557446, + "flos": 23587242065280.0, + "grad_norm": 1.7174246749256088, + "language_loss": 0.82468009, + "learning_rate": 3.898126521667446e-06, + "loss": 0.84763765, + "num_input_tokens_seen": 22731420, + "step": 1073, + "time_per_iteration": 2.7290594577789307 + }, + { + "auxiliary_loss_clip": 0.01248401, + "auxiliary_loss_mlp": 0.01046857, + "balance_loss_clip": 1.06496537, + "balance_loss_mlp": 1.03642058, + "epoch": 0.12914086454638368, + "flos": 24170610850560.0, + "grad_norm": 1.5917538427708398, + "language_loss": 0.83291155, + "learning_rate": 3.897880934963007e-06, + "loss": 0.85586411, + "num_input_tokens_seen": 22750970, + "step": 1074, + "time_per_iteration": 2.7226545810699463 + }, + { + "auxiliary_loss_clip": 0.01250016, + "auxiliary_loss_mlp": 0.01036538, + "balance_loss_clip": 1.06068933, + "balance_loss_mlp": 1.0257498, + "epoch": 0.1292611074370228, + "flos": 20267157081600.0, + "grad_norm": 2.1305253708321352, + "language_loss": 0.7827825, + "learning_rate": 3.89763506035154e-06, + "loss": 0.80564803, + "num_input_tokens_seen": 22768820, + "step": 1075, + "time_per_iteration": 2.6786720752716064 + }, + { + "auxiliary_loss_clip": 0.01228287, + "auxiliary_loss_mlp": 0.01045443, + "balance_loss_clip": 1.06050944, + "balance_loss_mlp": 1.03493476, + "epoch": 0.12938135032766188, + "flos": 27377684668800.0, + "grad_norm": 1.6394456209943942, + "language_loss": 0.81073701, + "learning_rate": 3.897388897870343e-06, + "loss": 0.8334744, + "num_input_tokens_seen": 22789460, + "step": 1076, + "time_per_iteration": 2.7081656455993652 + }, + { + "auxiliary_loss_clip": 0.01262369, + "auxiliary_loss_mlp": 0.01039651, + "balance_loss_clip": 1.06082869, + "balance_loss_mlp": 1.02848148, + "epoch": 0.12950159321830096, + "flos": 29277107861760.0, + "grad_norm": 1.9642118031912503, + "language_loss": 0.74926645, + "learning_rate": 3.89714244755676e-06, + "loss": 0.77228665, + "num_input_tokens_seen": 22810820, + "step": 1077, + "time_per_iteration": 2.7223269939422607 + }, + { + "auxiliary_loss_clip": 0.01247325, + "auxiliary_loss_mlp": 0.01045068, + "balance_loss_clip": 1.05378914, + "balance_loss_mlp": 1.0332303, + "epoch": 0.12962183610894007, + "flos": 24534888629760.0, + "grad_norm": 2.4717221474661413, + "language_loss": 0.86306584, + "learning_rate": 3.896895709448175e-06, + "loss": 0.88598979, + "num_input_tokens_seen": 22830570, + "step": 1078, + "time_per_iteration": 2.771890640258789 + }, + { + "auxiliary_loss_clip": 0.01267007, + "auxiliary_loss_mlp": 0.01036227, + "balance_loss_clip": 1.0530473, + "balance_loss_mlp": 1.02503943, + "epoch": 0.12974207899957915, + "flos": 11215944552960.0, + "grad_norm": 2.6811793327145397, + "language_loss": 0.77058983, + "learning_rate": 3.896648683582019e-06, + "loss": 0.7936222, + "num_input_tokens_seen": 22845905, + "step": 1079, + "time_per_iteration": 2.740468978881836 + }, + { + "auxiliary_loss_clip": 0.01269991, + "auxiliary_loss_mlp": 0.01041583, + "balance_loss_clip": 1.06266546, + "balance_loss_mlp": 1.02965665, + "epoch": 0.12986232189021824, + "flos": 24717889445760.0, + "grad_norm": 2.0638117839628585, + "language_loss": 0.8099314, + "learning_rate": 3.896401369995766e-06, + "loss": 0.83304715, + "num_input_tokens_seen": 22865710, + "step": 1080, + "time_per_iteration": 2.7397656440734863 + }, + { + "auxiliary_loss_clip": 0.01241311, + "auxiliary_loss_mlp": 0.01043282, + "balance_loss_clip": 1.06835079, + "balance_loss_mlp": 1.03227901, + "epoch": 0.12998256478085732, + "flos": 23915357827200.0, + "grad_norm": 1.893460610616845, + "language_loss": 0.79606134, + "learning_rate": 3.896153768726932e-06, + "loss": 0.81890726, + "num_input_tokens_seen": 22886020, + "step": 1081, + "time_per_iteration": 2.6289498805999756 + }, + { + "auxiliary_loss_clip": 0.01247617, + "auxiliary_loss_mlp": 0.01039358, + "balance_loss_clip": 1.06780672, + "balance_loss_mlp": 1.02870679, + "epoch": 0.13010280767149643, + "flos": 18624207974400.0, + "grad_norm": 2.81498924249191, + "language_loss": 0.87831336, + "learning_rate": 3.8959058798130806e-06, + "loss": 0.90118313, + "num_input_tokens_seen": 22903995, + "step": 1082, + "time_per_iteration": 3.645328998565674 + }, + { + "auxiliary_loss_clip": 0.01251117, + "auxiliary_loss_mlp": 0.01381969, + "balance_loss_clip": 1.06268358, + "balance_loss_mlp": 1.00014639, + "epoch": 0.1302230505621355, + "flos": 22783992174720.0, + "grad_norm": 1.9608919410821124, + "language_loss": 0.75224304, + "learning_rate": 3.895657703291814e-06, + "loss": 0.77857393, + "num_input_tokens_seen": 22924100, + "step": 1083, + "time_per_iteration": 2.737985849380493 + }, + { + "auxiliary_loss_clip": 0.01257795, + "auxiliary_loss_mlp": 0.01035035, + "balance_loss_clip": 1.05915892, + "balance_loss_mlp": 1.02337027, + "epoch": 0.1303432934527746, + "flos": 21323612920320.0, + "grad_norm": 3.4726093021106164, + "language_loss": 0.79669976, + "learning_rate": 3.895409239200781e-06, + "loss": 0.81962806, + "num_input_tokens_seen": 22939985, + "step": 1084, + "time_per_iteration": 4.49918532371521 + }, + { + "auxiliary_loss_clip": 0.01235146, + "auxiliary_loss_mlp": 0.01046076, + "balance_loss_clip": 1.06048179, + "balance_loss_mlp": 1.03513861, + "epoch": 0.1304635363434137, + "flos": 20922490765440.0, + "grad_norm": 2.1899528908016204, + "language_loss": 0.91288996, + "learning_rate": 3.895160487577673e-06, + "loss": 0.93570215, + "num_input_tokens_seen": 22957555, + "step": 1085, + "time_per_iteration": 2.676694869995117 + }, + { + "auxiliary_loss_clip": 0.01136283, + "auxiliary_loss_mlp": 0.01004687, + "balance_loss_clip": 1.03089929, + "balance_loss_mlp": 1.0017904, + "epoch": 0.1305837792340528, + "flos": 63245659080960.0, + "grad_norm": 0.7871941823105998, + "language_loss": 0.60937297, + "learning_rate": 3.894911448460226e-06, + "loss": 0.63078272, + "num_input_tokens_seen": 23016870, + "step": 1086, + "time_per_iteration": 3.981031894683838 + }, + { + "auxiliary_loss_clip": 0.01259443, + "auxiliary_loss_mlp": 0.01039255, + "balance_loss_clip": 1.05343103, + "balance_loss_mlp": 1.02834153, + "epoch": 0.13070402212469187, + "flos": 26428852955520.0, + "grad_norm": 1.9064175363755604, + "language_loss": 0.72583717, + "learning_rate": 3.8946621218862195e-06, + "loss": 0.74882418, + "num_input_tokens_seen": 23037870, + "step": 1087, + "time_per_iteration": 2.919982433319092 + }, + { + "auxiliary_loss_clip": 0.01255411, + "auxiliary_loss_mlp": 0.01036712, + "balance_loss_clip": 1.06069458, + "balance_loss_mlp": 1.0257504, + "epoch": 0.13082426501533098, + "flos": 27673409341440.0, + "grad_norm": 1.8624358075143652, + "language_loss": 0.88764668, + "learning_rate": 3.894412507893475e-06, + "loss": 0.91056788, + "num_input_tokens_seen": 23058150, + "step": 1088, + "time_per_iteration": 2.744326114654541 + }, + { + "auxiliary_loss_clip": 0.01273178, + "auxiliary_loss_mlp": 0.01042026, + "balance_loss_clip": 1.05783105, + "balance_loss_mlp": 1.03124964, + "epoch": 0.13094450790597006, + "flos": 24826770547200.0, + "grad_norm": 2.1417756569305415, + "language_loss": 0.71681637, + "learning_rate": 3.894162606519859e-06, + "loss": 0.73996842, + "num_input_tokens_seen": 23077100, + "step": 1089, + "time_per_iteration": 2.7628602981567383 + }, + { + "auxiliary_loss_clip": 0.01263476, + "auxiliary_loss_mlp": 0.01035043, + "balance_loss_clip": 1.05912006, + "balance_loss_mlp": 1.02392125, + "epoch": 0.13106475079660915, + "flos": 19062605468160.0, + "grad_norm": 1.9258305916925866, + "language_loss": 0.77011156, + "learning_rate": 3.893912417803282e-06, + "loss": 0.79309672, + "num_input_tokens_seen": 23096815, + "step": 1090, + "time_per_iteration": 2.7168166637420654 + }, + { + "auxiliary_loss_clip": 0.0126439, + "auxiliary_loss_mlp": 0.01040789, + "balance_loss_clip": 1.05354548, + "balance_loss_mlp": 1.02956581, + "epoch": 0.13118499368724823, + "flos": 28913189218560.0, + "grad_norm": 4.019995928169794, + "language_loss": 0.76830465, + "learning_rate": 3.8936619417816975e-06, + "loss": 0.79135644, + "num_input_tokens_seen": 23117145, + "step": 1091, + "time_per_iteration": 2.779592514038086 + }, + { + "auxiliary_loss_clip": 0.0125828, + "auxiliary_loss_mlp": 0.01035632, + "balance_loss_clip": 1.06237078, + "balance_loss_mlp": 1.02476645, + "epoch": 0.13130523657788734, + "flos": 14283398206080.0, + "grad_norm": 1.7612985771616265, + "language_loss": 0.71502578, + "learning_rate": 3.8934111784931015e-06, + "loss": 0.73796493, + "num_input_tokens_seen": 23134595, + "step": 1092, + "time_per_iteration": 2.65523624420166 + }, + { + "auxiliary_loss_clip": 0.01145505, + "auxiliary_loss_mlp": 0.01005744, + "balance_loss_clip": 1.02846193, + "balance_loss_mlp": 1.00278807, + "epoch": 0.13142547946852642, + "flos": 70174155519360.0, + "grad_norm": 1.1142708663885825, + "language_loss": 0.59060597, + "learning_rate": 3.893160127975535e-06, + "loss": 0.61211848, + "num_input_tokens_seen": 23195285, + "step": 1093, + "time_per_iteration": 3.368644952774048 + }, + { + "auxiliary_loss_clip": 0.01274919, + "auxiliary_loss_mlp": 0.01037243, + "balance_loss_clip": 1.06046057, + "balance_loss_mlp": 1.02667475, + "epoch": 0.1315457223591655, + "flos": 45805998844800.0, + "grad_norm": 2.3593599270035073, + "language_loss": 0.80971265, + "learning_rate": 3.8929087902670826e-06, + "loss": 0.8328343, + "num_input_tokens_seen": 23216915, + "step": 1094, + "time_per_iteration": 3.033609628677368 + }, + { + "auxiliary_loss_clip": 0.01137105, + "auxiliary_loss_mlp": 0.01001514, + "balance_loss_clip": 1.02934384, + "balance_loss_mlp": 0.99854565, + "epoch": 0.13166596524980462, + "flos": 62881165820160.0, + "grad_norm": 0.9347662658066808, + "language_loss": 0.60724854, + "learning_rate": 3.8926571654058715e-06, + "loss": 0.62863481, + "num_input_tokens_seen": 23273560, + "step": 1095, + "time_per_iteration": 3.1847429275512695 + }, + { + "auxiliary_loss_clip": 0.01261103, + "auxiliary_loss_mlp": 0.01041654, + "balance_loss_clip": 1.06361938, + "balance_loss_mlp": 1.03081167, + "epoch": 0.1317862081404437, + "flos": 23586523793280.0, + "grad_norm": 2.419041875540747, + "language_loss": 0.77010465, + "learning_rate": 3.892405253430074e-06, + "loss": 0.79313219, + "num_input_tokens_seen": 23291080, + "step": 1096, + "time_per_iteration": 2.8020615577697754 + }, + { + "auxiliary_loss_clip": 0.01256181, + "auxiliary_loss_mlp": 0.01381669, + "balance_loss_clip": 1.06482935, + "balance_loss_mlp": 1.00016665, + "epoch": 0.13190645103108278, + "flos": 20260764460800.0, + "grad_norm": 1.7796868757476008, + "language_loss": 0.82406062, + "learning_rate": 3.892153054377904e-06, + "loss": 0.85043907, + "num_input_tokens_seen": 23308485, + "step": 1097, + "time_per_iteration": 2.7669522762298584 + }, + { + "auxiliary_loss_clip": 0.01164022, + "auxiliary_loss_mlp": 0.0099886, + "balance_loss_clip": 1.02267933, + "balance_loss_mlp": 0.99634439, + "epoch": 0.13202669392172187, + "flos": 53455440136320.0, + "grad_norm": 0.933424558655253, + "language_loss": 0.59362936, + "learning_rate": 3.891900568287619e-06, + "loss": 0.61525822, + "num_input_tokens_seen": 23360870, + "step": 1098, + "time_per_iteration": 3.1476972103118896 + }, + { + "auxiliary_loss_clip": 0.01265467, + "auxiliary_loss_mlp": 0.01044009, + "balance_loss_clip": 1.06117916, + "balance_loss_mlp": 1.0323745, + "epoch": 0.13214693681236098, + "flos": 15851293845120.0, + "grad_norm": 2.4239313489567444, + "language_loss": 0.72460866, + "learning_rate": 3.891647795197523e-06, + "loss": 0.74770343, + "num_input_tokens_seen": 23376910, + "step": 1099, + "time_per_iteration": 2.8111376762390137 + }, + { + "auxiliary_loss_clip": 0.01257861, + "auxiliary_loss_mlp": 0.01047423, + "balance_loss_clip": 1.05530357, + "balance_loss_mlp": 1.03547847, + "epoch": 0.13226717970300006, + "flos": 19353840940800.0, + "grad_norm": 2.0183394941981696, + "language_loss": 0.68616855, + "learning_rate": 3.8913947351459605e-06, + "loss": 0.70922136, + "num_input_tokens_seen": 23394450, + "step": 1100, + "time_per_iteration": 2.7816271781921387 + }, + { + "auxiliary_loss_clip": 0.01233398, + "auxiliary_loss_mlp": 0.01035476, + "balance_loss_clip": 1.06487894, + "balance_loss_mlp": 1.02523017, + "epoch": 0.13238742259363914, + "flos": 20698084546560.0, + "grad_norm": 1.821357391615246, + "language_loss": 0.67730755, + "learning_rate": 3.89114138817132e-06, + "loss": 0.69999623, + "num_input_tokens_seen": 23411115, + "step": 1101, + "time_per_iteration": 2.682507276535034 + }, + { + "auxiliary_loss_clip": 0.0124581, + "auxiliary_loss_mlp": 0.01035129, + "balance_loss_clip": 1.06900692, + "balance_loss_mlp": 1.02462697, + "epoch": 0.13250766548427825, + "flos": 21032449274880.0, + "grad_norm": 1.8526322263652122, + "language_loss": 0.84254199, + "learning_rate": 3.890887754312035e-06, + "loss": 0.86535138, + "num_input_tokens_seen": 23429360, + "step": 1102, + "time_per_iteration": 2.714142322540283 + }, + { + "auxiliary_loss_clip": 0.01245953, + "auxiliary_loss_mlp": 0.01035373, + "balance_loss_clip": 1.05984902, + "balance_loss_mlp": 1.02451944, + "epoch": 0.13262790837491734, + "flos": 22637871648000.0, + "grad_norm": 2.2553750858774486, + "language_loss": 0.87641829, + "learning_rate": 3.890633833606581e-06, + "loss": 0.89923155, + "num_input_tokens_seen": 23449050, + "step": 1103, + "time_per_iteration": 2.7205939292907715 + }, + { + "auxiliary_loss_clip": 0.01245014, + "auxiliary_loss_mlp": 0.0104014, + "balance_loss_clip": 1.06963491, + "balance_loss_mlp": 1.02883887, + "epoch": 0.13274815126555642, + "flos": 19683141851520.0, + "grad_norm": 1.9333748361002765, + "language_loss": 0.69777846, + "learning_rate": 3.890379626093477e-06, + "loss": 0.72062999, + "num_input_tokens_seen": 23468800, + "step": 1104, + "time_per_iteration": 2.626530885696411 + }, + { + "auxiliary_loss_clip": 0.0125053, + "auxiliary_loss_mlp": 0.01034085, + "balance_loss_clip": 1.05533338, + "balance_loss_mlp": 1.0230577, + "epoch": 0.1328683941561955, + "flos": 21317687176320.0, + "grad_norm": 1.991745292323274, + "language_loss": 0.92321271, + "learning_rate": 3.890125131811287e-06, + "loss": 0.94605887, + "num_input_tokens_seen": 23486850, + "step": 1105, + "time_per_iteration": 2.7432610988616943 + }, + { + "auxiliary_loss_clip": 0.01234541, + "auxiliary_loss_mlp": 0.01032924, + "balance_loss_clip": 1.05821776, + "balance_loss_mlp": 1.02296972, + "epoch": 0.1329886370468346, + "flos": 13699131580800.0, + "grad_norm": 2.2104675200752335, + "language_loss": 0.75170434, + "learning_rate": 3.889870350798618e-06, + "loss": 0.77437901, + "num_input_tokens_seen": 23504195, + "step": 1106, + "time_per_iteration": 2.6422767639160156 + }, + { + "auxiliary_loss_clip": 0.01235443, + "auxiliary_loss_mlp": 0.01036303, + "balance_loss_clip": 1.06669462, + "balance_loss_mlp": 1.02589607, + "epoch": 0.1331088799374737, + "flos": 21032413361280.0, + "grad_norm": 1.5764191034341466, + "language_loss": 0.78372175, + "learning_rate": 3.889615283094119e-06, + "loss": 0.80643922, + "num_input_tokens_seen": 23523385, + "step": 1107, + "time_per_iteration": 3.847628116607666 + }, + { + "auxiliary_loss_clip": 0.01240121, + "auxiliary_loss_mlp": 0.01041363, + "balance_loss_clip": 1.0657562, + "balance_loss_mlp": 1.02991891, + "epoch": 0.13322912282811278, + "flos": 18260432985600.0, + "grad_norm": 2.1377448464080637, + "language_loss": 0.84403342, + "learning_rate": 3.889359928736485e-06, + "loss": 0.86684823, + "num_input_tokens_seen": 23541330, + "step": 1108, + "time_per_iteration": 2.59261155128479 + }, + { + "auxiliary_loss_clip": 0.01245487, + "auxiliary_loss_mlp": 0.01381217, + "balance_loss_clip": 1.06284976, + "balance_loss_mlp": 1.00006163, + "epoch": 0.1333493657187519, + "flos": 24460876656000.0, + "grad_norm": 2.4568162478298228, + "language_loss": 0.91367543, + "learning_rate": 3.889104287764451e-06, + "loss": 0.93994248, + "num_input_tokens_seen": 23561705, + "step": 1109, + "time_per_iteration": 2.6910083293914795 + }, + { + "auxiliary_loss_clip": 0.01251198, + "auxiliary_loss_mlp": 0.0103739, + "balance_loss_clip": 1.06450558, + "balance_loss_mlp": 1.02736485, + "epoch": 0.13346960860939097, + "flos": 22158930677760.0, + "grad_norm": 2.1492498341454747, + "language_loss": 0.90885854, + "learning_rate": 3.888848360216798e-06, + "loss": 0.93174446, + "num_input_tokens_seen": 23579350, + "step": 1110, + "time_per_iteration": 4.472954750061035 + }, + { + "auxiliary_loss_clip": 0.01146967, + "auxiliary_loss_mlp": 0.01008738, + "balance_loss_clip": 1.02613473, + "balance_loss_mlp": 1.0059607, + "epoch": 0.13358985150003005, + "flos": 67931212608000.0, + "grad_norm": 0.8133789536125996, + "language_loss": 0.56671178, + "learning_rate": 3.888592146132351e-06, + "loss": 0.58826876, + "num_input_tokens_seen": 23640620, + "step": 1111, + "time_per_iteration": 3.410916328430176 + }, + { + "auxiliary_loss_clip": 0.01241689, + "auxiliary_loss_mlp": 0.01039773, + "balance_loss_clip": 1.06669331, + "balance_loss_mlp": 1.02894235, + "epoch": 0.13371009439066917, + "flos": 26834284742400.0, + "grad_norm": 1.7538707297695526, + "language_loss": 0.78321052, + "learning_rate": 3.888335645549978e-06, + "loss": 0.80602503, + "num_input_tokens_seen": 23661040, + "step": 1112, + "time_per_iteration": 3.664803981781006 + }, + { + "auxiliary_loss_clip": 0.01237717, + "auxiliary_loss_mlp": 0.01039611, + "balance_loss_clip": 1.06883669, + "balance_loss_mlp": 1.0285964, + "epoch": 0.13383033728130825, + "flos": 26322844942080.0, + "grad_norm": 2.5272753553366316, + "language_loss": 0.81386316, + "learning_rate": 3.888078858508588e-06, + "loss": 0.83663642, + "num_input_tokens_seen": 23680900, + "step": 1113, + "time_per_iteration": 2.658066749572754 + }, + { + "auxiliary_loss_clip": 0.01251753, + "auxiliary_loss_mlp": 0.01040631, + "balance_loss_clip": 1.06577182, + "balance_loss_mlp": 1.02960992, + "epoch": 0.13395058017194733, + "flos": 22563931501440.0, + "grad_norm": 2.442551236686218, + "language_loss": 0.84660196, + "learning_rate": 3.8878217850471365e-06, + "loss": 0.86952579, + "num_input_tokens_seen": 23700815, + "step": 1114, + "time_per_iteration": 2.8023741245269775 + }, + { + "auxiliary_loss_clip": 0.01241511, + "auxiliary_loss_mlp": 0.01035577, + "balance_loss_clip": 1.06950533, + "balance_loss_mlp": 1.02392435, + "epoch": 0.13407082306258641, + "flos": 25810938264960.0, + "grad_norm": 1.793958688940105, + "language_loss": 0.74115753, + "learning_rate": 3.887564425204621e-06, + "loss": 0.76392847, + "num_input_tokens_seen": 23722500, + "step": 1115, + "time_per_iteration": 2.627763509750366 + }, + { + "auxiliary_loss_clip": 0.01148976, + "auxiliary_loss_mlp": 0.01005702, + "balance_loss_clip": 1.02430749, + "balance_loss_mlp": 1.0031147, + "epoch": 0.13419106595322552, + "flos": 68338365269760.0, + "grad_norm": 0.8455399319102755, + "language_loss": 0.54641002, + "learning_rate": 3.887306779020083e-06, + "loss": 0.56795681, + "num_input_tokens_seen": 23777155, + "step": 1116, + "time_per_iteration": 3.26275634765625 + }, + { + "auxiliary_loss_clip": 0.01249493, + "auxiliary_loss_mlp": 0.01040517, + "balance_loss_clip": 1.0664556, + "balance_loss_mlp": 1.02989519, + "epoch": 0.1343113088438646, + "flos": 20449080489600.0, + "grad_norm": 2.350645530317215, + "language_loss": 0.70134759, + "learning_rate": 3.887048846532608e-06, + "loss": 0.72424769, + "num_input_tokens_seen": 23794130, + "step": 1117, + "time_per_iteration": 2.8127267360687256 + }, + { + "auxiliary_loss_clip": 0.0114777, + "auxiliary_loss_mlp": 0.01000915, + "balance_loss_clip": 1.02113652, + "balance_loss_mlp": 0.99812585, + "epoch": 0.1344315517345037, + "flos": 67389784951680.0, + "grad_norm": 0.7606460041274214, + "language_loss": 0.58135498, + "learning_rate": 3.8867906277813224e-06, + "loss": 0.60284185, + "num_input_tokens_seen": 23852285, + "step": 1118, + "time_per_iteration": 3.236849308013916 + }, + { + "auxiliary_loss_clip": 0.01249297, + "auxiliary_loss_mlp": 0.01381768, + "balance_loss_clip": 1.06627965, + "balance_loss_mlp": 1.00009513, + "epoch": 0.1345517946251428, + "flos": 40734442788480.0, + "grad_norm": 2.0720958774988163, + "language_loss": 0.73667115, + "learning_rate": 3.886532122805399e-06, + "loss": 0.76298177, + "num_input_tokens_seen": 23874765, + "step": 1119, + "time_per_iteration": 2.822547674179077 + }, + { + "auxiliary_loss_clip": 0.01256187, + "auxiliary_loss_mlp": 0.0104279, + "balance_loss_clip": 1.05290246, + "balance_loss_mlp": 1.03111994, + "epoch": 0.13467203751578188, + "flos": 22816850140800.0, + "grad_norm": 2.2382248828771787, + "language_loss": 0.89607102, + "learning_rate": 3.886273331644053e-06, + "loss": 0.91906077, + "num_input_tokens_seen": 23893635, + "step": 1120, + "time_per_iteration": 2.7933554649353027 + }, + { + "auxiliary_loss_clip": 0.01265784, + "auxiliary_loss_mlp": 0.01042933, + "balance_loss_clip": 1.06119728, + "balance_loss_mlp": 1.03272903, + "epoch": 0.13479228040642097, + "flos": 17091576512640.0, + "grad_norm": 2.1216772049811707, + "language_loss": 0.82272887, + "learning_rate": 3.886014254336542e-06, + "loss": 0.84581602, + "num_input_tokens_seen": 23910110, + "step": 1121, + "time_per_iteration": 2.7681357860565186 + }, + { + "auxiliary_loss_clip": 0.01245265, + "auxiliary_loss_mlp": 0.01035621, + "balance_loss_clip": 1.0655818, + "balance_loss_mlp": 1.0254581, + "epoch": 0.13491252329706005, + "flos": 23730525417600.0, + "grad_norm": 1.6304536026181926, + "language_loss": 0.92510879, + "learning_rate": 3.885754890922168e-06, + "loss": 0.94791764, + "num_input_tokens_seen": 23930440, + "step": 1122, + "time_per_iteration": 2.917173147201538 + }, + { + "auxiliary_loss_clip": 0.01266428, + "auxiliary_loss_mlp": 0.01038958, + "balance_loss_clip": 1.05487037, + "balance_loss_mlp": 1.02768636, + "epoch": 0.13503276618769916, + "flos": 34127058960000.0, + "grad_norm": 1.9045253237553887, + "language_loss": 0.78466642, + "learning_rate": 3.885495241440277e-06, + "loss": 0.80772024, + "num_input_tokens_seen": 23954535, + "step": 1123, + "time_per_iteration": 3.067135810852051 + }, + { + "auxiliary_loss_clip": 0.01236762, + "auxiliary_loss_mlp": 0.01038222, + "balance_loss_clip": 1.06541586, + "balance_loss_mlp": 1.02695704, + "epoch": 0.13515300907833824, + "flos": 17712328377600.0, + "grad_norm": 1.9774309423539393, + "language_loss": 0.74212289, + "learning_rate": 3.885235305930257e-06, + "loss": 0.76487273, + "num_input_tokens_seen": 23972735, + "step": 1124, + "time_per_iteration": 2.786196231842041 + }, + { + "auxiliary_loss_clip": 0.01252668, + "auxiliary_loss_mlp": 0.01040782, + "balance_loss_clip": 1.06369698, + "balance_loss_mlp": 1.02964187, + "epoch": 0.13527325196897733, + "flos": 20260872201600.0, + "grad_norm": 2.225961556457172, + "language_loss": 0.85654283, + "learning_rate": 3.884975084431539e-06, + "loss": 0.87947738, + "num_input_tokens_seen": 23987685, + "step": 1125, + "time_per_iteration": 2.7411487102508545 + }, + { + "auxiliary_loss_clip": 0.01231753, + "auxiliary_loss_mlp": 0.01381637, + "balance_loss_clip": 1.06266117, + "balance_loss_mlp": 1.00003219, + "epoch": 0.13539349485961644, + "flos": 18186492839040.0, + "grad_norm": 2.651354359364227, + "language_loss": 0.91971964, + "learning_rate": 3.8847145769836e-06, + "loss": 0.94585347, + "num_input_tokens_seen": 24004105, + "step": 1126, + "time_per_iteration": 2.668009042739868 + }, + { + "auxiliary_loss_clip": 0.01240883, + "auxiliary_loss_mlp": 0.01042016, + "balance_loss_clip": 1.06841874, + "balance_loss_mlp": 1.0306139, + "epoch": 0.13551373775025552, + "flos": 19317463441920.0, + "grad_norm": 2.6164795017983367, + "language_loss": 0.66418213, + "learning_rate": 3.884453783625959e-06, + "loss": 0.68701112, + "num_input_tokens_seen": 24021715, + "step": 1127, + "time_per_iteration": 2.5960521697998047 + }, + { + "auxiliary_loss_clip": 0.01248648, + "auxiliary_loss_mlp": 0.01041261, + "balance_loss_clip": 1.06488729, + "balance_loss_mlp": 1.03048408, + "epoch": 0.1356339806408946, + "flos": 20850813175680.0, + "grad_norm": 2.2253618365555683, + "language_loss": 0.85041124, + "learning_rate": 3.884192704398176e-06, + "loss": 0.87331033, + "num_input_tokens_seen": 24038915, + "step": 1128, + "time_per_iteration": 2.696666717529297 + }, + { + "auxiliary_loss_clip": 0.01244924, + "auxiliary_loss_mlp": 0.01044201, + "balance_loss_clip": 1.06413567, + "balance_loss_mlp": 1.03328753, + "epoch": 0.13575422353153369, + "flos": 50476037696640.0, + "grad_norm": 1.9310591863514102, + "language_loss": 0.74846041, + "learning_rate": 3.883931339339858e-06, + "loss": 0.7713517, + "num_input_tokens_seen": 24063300, + "step": 1129, + "time_per_iteration": 2.8970718383789062 + }, + { + "auxiliary_loss_clip": 0.01253116, + "auxiliary_loss_mlp": 0.0103904, + "balance_loss_clip": 1.06789982, + "balance_loss_mlp": 1.02590895, + "epoch": 0.1358744664221728, + "flos": 18150797698560.0, + "grad_norm": 1.8680392869850495, + "language_loss": 0.78759307, + "learning_rate": 3.883669688490654e-06, + "loss": 0.81051463, + "num_input_tokens_seen": 24081070, + "step": 1130, + "time_per_iteration": 2.630103588104248 + }, + { + "auxiliary_loss_clip": 0.01238045, + "auxiliary_loss_mlp": 0.01381715, + "balance_loss_clip": 1.05999827, + "balance_loss_mlp": 1.00002897, + "epoch": 0.13599470931281188, + "flos": 18442966924800.0, + "grad_norm": 2.0038060044686774, + "language_loss": 0.85442907, + "learning_rate": 3.883407751890256e-06, + "loss": 0.88062668, + "num_input_tokens_seen": 24099675, + "step": 1131, + "time_per_iteration": 2.639732837677002 + }, + { + "auxiliary_loss_clip": 0.01251327, + "auxiliary_loss_mlp": 0.01049802, + "balance_loss_clip": 1.05733263, + "balance_loss_mlp": 1.03901994, + "epoch": 0.13611495220345096, + "flos": 26680766014080.0, + "grad_norm": 2.1937209023372732, + "language_loss": 0.85786229, + "learning_rate": 3.8831455295783994e-06, + "loss": 0.88087362, + "num_input_tokens_seen": 24118925, + "step": 1132, + "time_per_iteration": 2.759542226791382 + }, + { + "auxiliary_loss_clip": 0.01244995, + "auxiliary_loss_mlp": 0.01040339, + "balance_loss_clip": 1.06055689, + "balance_loss_mlp": 1.0291872, + "epoch": 0.13623519509409007, + "flos": 21686238673920.0, + "grad_norm": 1.8401809446839417, + "language_loss": 0.74415708, + "learning_rate": 3.882883021594864e-06, + "loss": 0.76701045, + "num_input_tokens_seen": 24137065, + "step": 1133, + "time_per_iteration": 3.6458778381347656 + }, + { + "auxiliary_loss_clip": 0.0124675, + "auxiliary_loss_mlp": 0.01041887, + "balance_loss_clip": 1.05972612, + "balance_loss_mlp": 1.03094316, + "epoch": 0.13635543798472916, + "flos": 14830389492480.0, + "grad_norm": 3.2645191518685333, + "language_loss": 0.8700012, + "learning_rate": 3.8826202279794705e-06, + "loss": 0.89288759, + "num_input_tokens_seen": 24154125, + "step": 1134, + "time_per_iteration": 2.648160696029663 + }, + { + "auxiliary_loss_clip": 0.01237037, + "auxiliary_loss_mlp": 0.01048441, + "balance_loss_clip": 1.06695139, + "balance_loss_mlp": 1.0373069, + "epoch": 0.13647568087536824, + "flos": 22890323410560.0, + "grad_norm": 2.0885228267704212, + "language_loss": 0.70530081, + "learning_rate": 3.882357148772085e-06, + "loss": 0.72815561, + "num_input_tokens_seen": 24171550, + "step": 1135, + "time_per_iteration": 2.605268955230713 + }, + { + "auxiliary_loss_clip": 0.01237018, + "auxiliary_loss_mlp": 0.01046173, + "balance_loss_clip": 1.05732489, + "balance_loss_mlp": 1.03506327, + "epoch": 0.13659592376600732, + "flos": 19937927998080.0, + "grad_norm": 2.3745979808784305, + "language_loss": 0.84207469, + "learning_rate": 3.882093784012617e-06, + "loss": 0.86490655, + "num_input_tokens_seen": 24190190, + "step": 1136, + "time_per_iteration": 4.490791082382202 + }, + { + "auxiliary_loss_clip": 0.01244918, + "auxiliary_loss_mlp": 0.01041504, + "balance_loss_clip": 1.05950212, + "balance_loss_mlp": 1.03066182, + "epoch": 0.13671616665664643, + "flos": 21428579439360.0, + "grad_norm": 1.7335633498102818, + "language_loss": 0.8408348, + "learning_rate": 3.881830133741019e-06, + "loss": 0.86369896, + "num_input_tokens_seen": 24209055, + "step": 1137, + "time_per_iteration": 2.6946747303009033 + }, + { + "auxiliary_loss_clip": 0.01262498, + "auxiliary_loss_mlp": 0.01050707, + "balance_loss_clip": 1.06483245, + "balance_loss_mlp": 1.0399667, + "epoch": 0.13683640954728551, + "flos": 22778138257920.0, + "grad_norm": 2.532272598379333, + "language_loss": 0.7646898, + "learning_rate": 3.881566197997285e-06, + "loss": 0.78782183, + "num_input_tokens_seen": 24225490, + "step": 1138, + "time_per_iteration": 3.654323101043701 + }, + { + "auxiliary_loss_clip": 0.01248538, + "auxiliary_loss_mlp": 0.01034385, + "balance_loss_clip": 1.06420779, + "balance_loss_mlp": 1.02410924, + "epoch": 0.1369566524379246, + "flos": 21725884310400.0, + "grad_norm": 1.543576744862593, + "language_loss": 0.74864459, + "learning_rate": 3.881301976821456e-06, + "loss": 0.77147377, + "num_input_tokens_seen": 24245520, + "step": 1139, + "time_per_iteration": 2.6455841064453125 + }, + { + "auxiliary_loss_clip": 0.01240733, + "auxiliary_loss_mlp": 0.01038151, + "balance_loss_clip": 1.06487632, + "balance_loss_mlp": 1.027035, + "epoch": 0.1370768953285637, + "flos": 18624459369600.0, + "grad_norm": 2.3323851333430907, + "language_loss": 0.90665925, + "learning_rate": 3.881037470253612e-06, + "loss": 0.92944813, + "num_input_tokens_seen": 24265035, + "step": 1140, + "time_per_iteration": 2.6525275707244873 + }, + { + "auxiliary_loss_clip": 0.0126776, + "auxiliary_loss_mlp": 0.01041388, + "balance_loss_clip": 1.06186569, + "balance_loss_mlp": 1.03006876, + "epoch": 0.1371971382192028, + "flos": 14939521989120.0, + "grad_norm": 2.582801587799827, + "language_loss": 0.79263103, + "learning_rate": 3.88077267833388e-06, + "loss": 0.81572253, + "num_input_tokens_seen": 24281550, + "step": 1141, + "time_per_iteration": 2.680349826812744 + }, + { + "auxiliary_loss_clip": 0.01256065, + "auxiliary_loss_mlp": 0.01047945, + "balance_loss_clip": 1.05671525, + "balance_loss_mlp": 1.03747821, + "epoch": 0.13731738110984187, + "flos": 19023785844480.0, + "grad_norm": 1.9706837188563806, + "language_loss": 0.83969796, + "learning_rate": 3.880507601102427e-06, + "loss": 0.86273801, + "num_input_tokens_seen": 24299485, + "step": 1142, + "time_per_iteration": 2.740426778793335 + }, + { + "auxiliary_loss_clip": 0.01235215, + "auxiliary_loss_mlp": 0.01041695, + "balance_loss_clip": 1.06884933, + "balance_loss_mlp": 1.03120494, + "epoch": 0.13743762400048098, + "flos": 18187462506240.0, + "grad_norm": 1.8664744586441224, + "language_loss": 0.81767881, + "learning_rate": 3.880242238599467e-06, + "loss": 0.8404479, + "num_input_tokens_seen": 24316010, + "step": 1143, + "time_per_iteration": 2.5448668003082275 + }, + { + "auxiliary_loss_clip": 0.012326, + "auxiliary_loss_mlp": 0.01037351, + "balance_loss_clip": 1.06493163, + "balance_loss_mlp": 1.02645564, + "epoch": 0.13755786689112007, + "flos": 21031982398080.0, + "grad_norm": 1.6133503478999944, + "language_loss": 0.83113223, + "learning_rate": 3.879976590865254e-06, + "loss": 0.85383177, + "num_input_tokens_seen": 24335465, + "step": 1144, + "time_per_iteration": 2.6233975887298584 + }, + { + "auxiliary_loss_clip": 0.01252597, + "auxiliary_loss_mlp": 0.01034665, + "balance_loss_clip": 1.06584167, + "balance_loss_mlp": 1.02341795, + "epoch": 0.13767810978175915, + "flos": 21360636864000.0, + "grad_norm": 2.8971929545129678, + "language_loss": 0.87270659, + "learning_rate": 3.879710657940087e-06, + "loss": 0.89557922, + "num_input_tokens_seen": 24354415, + "step": 1145, + "time_per_iteration": 2.625638961791992 + }, + { + "auxiliary_loss_clip": 0.01245099, + "auxiliary_loss_mlp": 0.01036196, + "balance_loss_clip": 1.06591892, + "balance_loss_mlp": 1.02504349, + "epoch": 0.13779835267239823, + "flos": 30592084861440.0, + "grad_norm": 2.0191319038196105, + "language_loss": 0.70202237, + "learning_rate": 3.879444439864308e-06, + "loss": 0.72483528, + "num_input_tokens_seen": 24373990, + "step": 1146, + "time_per_iteration": 2.692530393600464 + }, + { + "auxiliary_loss_clip": 0.01242627, + "auxiliary_loss_mlp": 0.01381647, + "balance_loss_clip": 1.06439686, + "balance_loss_mlp": 1.00018311, + "epoch": 0.13791859556303734, + "flos": 22669867687680.0, + "grad_norm": 1.7871345275997832, + "language_loss": 0.85687315, + "learning_rate": 3.879177936678301e-06, + "loss": 0.88311589, + "num_input_tokens_seen": 24392995, + "step": 1147, + "time_per_iteration": 2.6792194843292236 + }, + { + "auxiliary_loss_clip": 0.01253997, + "auxiliary_loss_mlp": 0.01038391, + "balance_loss_clip": 1.06303501, + "balance_loss_mlp": 1.02765584, + "epoch": 0.13803883845367643, + "flos": 35224166016000.0, + "grad_norm": 1.8034825452914847, + "language_loss": 0.77119982, + "learning_rate": 3.878911148422496e-06, + "loss": 0.79412365, + "num_input_tokens_seen": 24414470, + "step": 1148, + "time_per_iteration": 2.7553648948669434 + }, + { + "auxiliary_loss_clip": 0.01244153, + "auxiliary_loss_mlp": 0.01037724, + "balance_loss_clip": 1.06361246, + "balance_loss_mlp": 1.02708435, + "epoch": 0.1381590813443155, + "flos": 32014542332160.0, + "grad_norm": 2.7334976078361795, + "language_loss": 0.70006508, + "learning_rate": 3.878644075137364e-06, + "loss": 0.72288382, + "num_input_tokens_seen": 24435120, + "step": 1149, + "time_per_iteration": 2.747785806655884 + }, + { + "auxiliary_loss_clip": 0.01227878, + "auxiliary_loss_mlp": 0.01046535, + "balance_loss_clip": 1.05479097, + "balance_loss_mlp": 1.03487003, + "epoch": 0.13827932423495462, + "flos": 17821855923840.0, + "grad_norm": 2.215449123021004, + "language_loss": 0.79569197, + "learning_rate": 3.878376716863418e-06, + "loss": 0.81843609, + "num_input_tokens_seen": 24451420, + "step": 1150, + "time_per_iteration": 2.702580213546753 + }, + { + "auxiliary_loss_clip": 0.01246523, + "auxiliary_loss_mlp": 0.01033447, + "balance_loss_clip": 1.06018257, + "balance_loss_mlp": 1.02202713, + "epoch": 0.1383995671255937, + "flos": 19427098728960.0, + "grad_norm": 2.282265237897816, + "language_loss": 0.71599144, + "learning_rate": 3.878109073641219e-06, + "loss": 0.73879117, + "num_input_tokens_seen": 24470450, + "step": 1151, + "time_per_iteration": 2.7038118839263916 + }, + { + "auxiliary_loss_clip": 0.01259361, + "auxiliary_loss_mlp": 0.01040405, + "balance_loss_clip": 1.05940175, + "balance_loss_mlp": 1.02892542, + "epoch": 0.13851981001623279, + "flos": 28296603331200.0, + "grad_norm": 1.5668789658673037, + "language_loss": 0.81246996, + "learning_rate": 3.877841145511366e-06, + "loss": 0.8354677, + "num_input_tokens_seen": 24493190, + "step": 1152, + "time_per_iteration": 2.798632860183716 + }, + { + "auxiliary_loss_clip": 0.01245261, + "auxiliary_loss_mlp": 0.01037268, + "balance_loss_clip": 1.06472659, + "balance_loss_mlp": 1.02659297, + "epoch": 0.13864005290687187, + "flos": 21213079793280.0, + "grad_norm": 1.6853619548374117, + "language_loss": 0.82762337, + "learning_rate": 3.8775729325145035e-06, + "loss": 0.85044861, + "num_input_tokens_seen": 24512425, + "step": 1153, + "time_per_iteration": 2.7476348876953125 + }, + { + "auxiliary_loss_clip": 0.01144516, + "auxiliary_loss_mlp": 0.0101063, + "balance_loss_clip": 1.02755165, + "balance_loss_mlp": 1.0081147, + "epoch": 0.13876029579751098, + "flos": 71653389413760.0, + "grad_norm": 0.8180683161535102, + "language_loss": 0.64722604, + "learning_rate": 3.877304434691321e-06, + "loss": 0.66877753, + "num_input_tokens_seen": 24579275, + "step": 1154, + "time_per_iteration": 3.414740562438965 + }, + { + "auxiliary_loss_clip": 0.0126577, + "auxiliary_loss_mlp": 0.01040164, + "balance_loss_clip": 1.06649494, + "balance_loss_mlp": 1.03003693, + "epoch": 0.13888053868815006, + "flos": 21941348042880.0, + "grad_norm": 1.883443095673576, + "language_loss": 0.79420537, + "learning_rate": 3.877035652082548e-06, + "loss": 0.81726468, + "num_input_tokens_seen": 24598720, + "step": 1155, + "time_per_iteration": 2.7434654235839844 + }, + { + "auxiliary_loss_clip": 0.0123954, + "auxiliary_loss_mlp": 0.01038824, + "balance_loss_clip": 1.06262302, + "balance_loss_mlp": 1.0270226, + "epoch": 0.13900078157878915, + "flos": 19608627087360.0, + "grad_norm": 1.7638197814134098, + "language_loss": 0.85513711, + "learning_rate": 3.87676658472896e-06, + "loss": 0.87792075, + "num_input_tokens_seen": 24617530, + "step": 1156, + "time_per_iteration": 2.7040135860443115 + }, + { + "auxiliary_loss_clip": 0.01240891, + "auxiliary_loss_mlp": 0.01041367, + "balance_loss_clip": 1.06098771, + "balance_loss_mlp": 1.02920771, + "epoch": 0.13912102446942826, + "flos": 22638051216000.0, + "grad_norm": 1.7862942963544235, + "language_loss": 0.85295594, + "learning_rate": 3.876497232671372e-06, + "loss": 0.87577856, + "num_input_tokens_seen": 24637485, + "step": 1157, + "time_per_iteration": 2.6666178703308105 + }, + { + "auxiliary_loss_clip": 0.01269853, + "auxiliary_loss_mlp": 0.0104029, + "balance_loss_clip": 1.05874705, + "balance_loss_mlp": 1.02830923, + "epoch": 0.13924126736006734, + "flos": 29643324975360.0, + "grad_norm": 3.472746456360893, + "language_loss": 0.83965409, + "learning_rate": 3.876227595950647e-06, + "loss": 0.86275554, + "num_input_tokens_seen": 24656915, + "step": 1158, + "time_per_iteration": 2.852586507797241 + }, + { + "auxiliary_loss_clip": 0.01235915, + "auxiliary_loss_mlp": 0.01033942, + "balance_loss_clip": 1.06720257, + "balance_loss_mlp": 1.02340376, + "epoch": 0.13936151025070642, + "flos": 27417653527680.0, + "grad_norm": 1.5659739974406361, + "language_loss": 0.78951836, + "learning_rate": 3.875957674607686e-06, + "loss": 0.812217, + "num_input_tokens_seen": 24679190, + "step": 1159, + "time_per_iteration": 3.5679187774658203 + }, + { + "auxiliary_loss_clip": 0.0123308, + "auxiliary_loss_mlp": 0.01382046, + "balance_loss_clip": 1.06041527, + "balance_loss_mlp": 1.00009346, + "epoch": 0.1394817531413455, + "flos": 16399326625920.0, + "grad_norm": 1.9842331847319352, + "language_loss": 0.88130808, + "learning_rate": 3.8756874686834386e-06, + "loss": 0.90745926, + "num_input_tokens_seen": 24697405, + "step": 1160, + "time_per_iteration": 2.714012861251831 + }, + { + "auxiliary_loss_clip": 0.01247077, + "auxiliary_loss_mlp": 0.01382075, + "balance_loss_clip": 1.06416726, + "balance_loss_mlp": 1.00018847, + "epoch": 0.13960199603198462, + "flos": 30922319525760.0, + "grad_norm": 1.7540965966904876, + "language_loss": 0.80111229, + "learning_rate": 3.875416978218893e-06, + "loss": 0.82740378, + "num_input_tokens_seen": 24720600, + "step": 1161, + "time_per_iteration": 2.779686450958252 + }, + { + "auxiliary_loss_clip": 0.01264938, + "auxiliary_loss_mlp": 0.01051927, + "balance_loss_clip": 1.05874884, + "balance_loss_mlp": 1.04046488, + "epoch": 0.1397222389226237, + "flos": 18113773754880.0, + "grad_norm": 2.3809063930785657, + "language_loss": 0.82800883, + "learning_rate": 3.8751462032550835e-06, + "loss": 0.85117745, + "num_input_tokens_seen": 24737605, + "step": 1162, + "time_per_iteration": 4.586500406265259 + }, + { + "auxiliary_loss_clip": 0.01250065, + "auxiliary_loss_mlp": 0.01036425, + "balance_loss_clip": 1.06856358, + "balance_loss_mlp": 1.02636981, + "epoch": 0.13984248181326278, + "flos": 16872772815360.0, + "grad_norm": 2.5363865335183085, + "language_loss": 0.82789409, + "learning_rate": 3.874875143833085e-06, + "loss": 0.85075891, + "num_input_tokens_seen": 24755845, + "step": 1163, + "time_per_iteration": 2.601254940032959 + }, + { + "auxiliary_loss_clip": 0.01246551, + "auxiliary_loss_mlp": 0.01043849, + "balance_loss_clip": 1.06595945, + "balance_loss_mlp": 1.03227401, + "epoch": 0.1399627247039019, + "flos": 54121401267840.0, + "grad_norm": 1.7973332386213867, + "language_loss": 0.68944901, + "learning_rate": 3.874603799994019e-06, + "loss": 0.71235299, + "num_input_tokens_seen": 24779380, + "step": 1164, + "time_per_iteration": 3.877577304840088 + }, + { + "auxiliary_loss_clip": 0.01245601, + "auxiliary_loss_mlp": 0.01045166, + "balance_loss_clip": 1.05814505, + "balance_loss_mlp": 1.03444362, + "epoch": 0.14008296759454097, + "flos": 11765521618560.0, + "grad_norm": 2.0718569628752856, + "language_loss": 0.86658013, + "learning_rate": 3.874332171779046e-06, + "loss": 0.8894878, + "num_input_tokens_seen": 24794260, + "step": 1165, + "time_per_iteration": 2.66451358795166 + }, + { + "auxiliary_loss_clip": 0.01253084, + "auxiliary_loss_mlp": 0.01039035, + "balance_loss_clip": 1.05805635, + "balance_loss_mlp": 1.02859187, + "epoch": 0.14020321048518006, + "flos": 22017514832640.0, + "grad_norm": 1.715576439813319, + "language_loss": 0.75665593, + "learning_rate": 3.874060259229373e-06, + "loss": 0.77957714, + "num_input_tokens_seen": 24815835, + "step": 1166, + "time_per_iteration": 2.7020602226257324 + }, + { + "auxiliary_loss_clip": 0.01251203, + "auxiliary_loss_mlp": 0.01049059, + "balance_loss_clip": 1.06983113, + "balance_loss_mlp": 1.03778207, + "epoch": 0.14032345337581917, + "flos": 23404313076480.0, + "grad_norm": 2.3445325269984534, + "language_loss": 0.93725777, + "learning_rate": 3.873788062386249e-06, + "loss": 0.96026039, + "num_input_tokens_seen": 24834095, + "step": 1167, + "time_per_iteration": 2.621399164199829 + }, + { + "auxiliary_loss_clip": 0.01266043, + "auxiliary_loss_mlp": 0.01047175, + "balance_loss_clip": 1.06463373, + "balance_loss_mlp": 1.03653002, + "epoch": 0.14044369626645825, + "flos": 29645767100160.0, + "grad_norm": 1.7611602567957068, + "language_loss": 0.81971192, + "learning_rate": 3.873515581290965e-06, + "loss": 0.84284413, + "num_input_tokens_seen": 24858900, + "step": 1168, + "time_per_iteration": 2.8820221424102783 + }, + { + "auxiliary_loss_clip": 0.01259873, + "auxiliary_loss_mlp": 0.01037463, + "balance_loss_clip": 1.06353652, + "balance_loss_mlp": 1.02647805, + "epoch": 0.14056393915709733, + "flos": 18332972501760.0, + "grad_norm": 2.0191449385210496, + "language_loss": 0.75564218, + "learning_rate": 3.8732428159848575e-06, + "loss": 0.77861553, + "num_input_tokens_seen": 24877875, + "step": 1169, + "time_per_iteration": 2.7015297412872314 + }, + { + "auxiliary_loss_clip": 0.01245825, + "auxiliary_loss_mlp": 0.01043792, + "balance_loss_clip": 1.06972909, + "balance_loss_mlp": 1.03252113, + "epoch": 0.14068418204773642, + "flos": 26687517770880.0, + "grad_norm": 1.961337551652606, + "language_loss": 0.7844587, + "learning_rate": 3.872969766509304e-06, + "loss": 0.80735493, + "num_input_tokens_seen": 24898430, + "step": 1170, + "time_per_iteration": 2.7204904556274414 + }, + { + "auxiliary_loss_clip": 0.01140841, + "auxiliary_loss_mlp": 0.01002312, + "balance_loss_clip": 1.021384, + "balance_loss_mlp": 0.99972552, + "epoch": 0.14080442493837553, + "flos": 65259314501760.0, + "grad_norm": 0.760222348036148, + "language_loss": 0.55644792, + "learning_rate": 3.872696432905726e-06, + "loss": 0.57787943, + "num_input_tokens_seen": 24959250, + "step": 1171, + "time_per_iteration": 3.2732512950897217 + }, + { + "auxiliary_loss_clip": 0.01246977, + "auxiliary_loss_mlp": 0.01035909, + "balance_loss_clip": 1.06211567, + "balance_loss_mlp": 1.0249182, + "epoch": 0.1409246678290146, + "flos": 25776715582080.0, + "grad_norm": 2.1937404930947406, + "language_loss": 0.71349508, + "learning_rate": 3.872422815215589e-06, + "loss": 0.73632395, + "num_input_tokens_seen": 24978330, + "step": 1172, + "time_per_iteration": 2.6810762882232666 + }, + { + "auxiliary_loss_clip": 0.0123374, + "auxiliary_loss_mlp": 0.01043955, + "balance_loss_clip": 1.06013727, + "balance_loss_mlp": 1.03183115, + "epoch": 0.1410449107196537, + "flos": 21868521217920.0, + "grad_norm": 2.133964888787409, + "language_loss": 0.74704349, + "learning_rate": 3.8721489134803994e-06, + "loss": 0.76982039, + "num_input_tokens_seen": 24997120, + "step": 1173, + "time_per_iteration": 2.6842119693756104 + }, + { + "auxiliary_loss_clip": 0.01242916, + "auxiliary_loss_mlp": 0.01037765, + "balance_loss_clip": 1.06459451, + "balance_loss_mlp": 1.02651715, + "epoch": 0.1411651536102928, + "flos": 16684133564160.0, + "grad_norm": 2.2718156398620177, + "language_loss": 0.72605681, + "learning_rate": 3.871874727741707e-06, + "loss": 0.74886364, + "num_input_tokens_seen": 25014350, + "step": 1174, + "time_per_iteration": 2.599893093109131 + }, + { + "auxiliary_loss_clip": 0.01239667, + "auxiliary_loss_mlp": 0.01041853, + "balance_loss_clip": 1.06642115, + "balance_loss_mlp": 1.03185689, + "epoch": 0.1412853965009319, + "flos": 20992264934400.0, + "grad_norm": 1.784091406530716, + "language_loss": 0.96777046, + "learning_rate": 3.871600258041108e-06, + "loss": 0.99058568, + "num_input_tokens_seen": 25033875, + "step": 1175, + "time_per_iteration": 2.7669386863708496 + }, + { + "auxiliary_loss_clip": 0.01246412, + "auxiliary_loss_mlp": 0.01034511, + "balance_loss_clip": 1.06126249, + "balance_loss_mlp": 1.02365088, + "epoch": 0.14140563939157097, + "flos": 20335279224960.0, + "grad_norm": 2.20163017425374, + "language_loss": 0.85699081, + "learning_rate": 3.871325504420238e-06, + "loss": 0.87980008, + "num_input_tokens_seen": 25052865, + "step": 1176, + "time_per_iteration": 2.734261989593506 + }, + { + "auxiliary_loss_clip": 0.01235559, + "auxiliary_loss_mlp": 0.01037242, + "balance_loss_clip": 1.06640708, + "balance_loss_mlp": 1.02710342, + "epoch": 0.14152588228221005, + "flos": 21068826773760.0, + "grad_norm": 1.921485800982764, + "language_loss": 0.81732064, + "learning_rate": 3.871050466920776e-06, + "loss": 0.84004867, + "num_input_tokens_seen": 25072770, + "step": 1177, + "time_per_iteration": 2.6459884643554688 + }, + { + "auxiliary_loss_clip": 0.01246117, + "auxiliary_loss_mlp": 0.01040571, + "balance_loss_clip": 1.05870199, + "balance_loss_mlp": 1.03023541, + "epoch": 0.14164612517284916, + "flos": 18223157646720.0, + "grad_norm": 2.0487181788432918, + "language_loss": 0.79925174, + "learning_rate": 3.870775145584447e-06, + "loss": 0.82211858, + "num_input_tokens_seen": 25090550, + "step": 1178, + "time_per_iteration": 2.6496522426605225 + }, + { + "auxiliary_loss_clip": 0.01255086, + "auxiliary_loss_mlp": 0.01043607, + "balance_loss_clip": 1.06261969, + "balance_loss_mlp": 1.03134632, + "epoch": 0.14176636806348825, + "flos": 22744454279040.0, + "grad_norm": 2.8188704415982926, + "language_loss": 0.65152514, + "learning_rate": 3.8704995404530145e-06, + "loss": 0.67451215, + "num_input_tokens_seen": 25106175, + "step": 1179, + "time_per_iteration": 2.668302297592163 + }, + { + "auxiliary_loss_clip": 0.0123501, + "auxiliary_loss_mlp": 0.01038988, + "balance_loss_clip": 1.06863773, + "balance_loss_mlp": 1.02869439, + "epoch": 0.14188661095412733, + "flos": 22091095843200.0, + "grad_norm": 1.683283482151086, + "language_loss": 0.85205686, + "learning_rate": 3.87022365156829e-06, + "loss": 0.87479687, + "num_input_tokens_seen": 25126890, + "step": 1180, + "time_per_iteration": 2.6537461280822754 + }, + { + "auxiliary_loss_clip": 0.01279813, + "auxiliary_loss_mlp": 0.01034399, + "balance_loss_clip": 1.05429387, + "balance_loss_mlp": 1.02321744, + "epoch": 0.14200685384476644, + "flos": 24352390604160.0, + "grad_norm": 1.8194176155664359, + "language_loss": 0.81139374, + "learning_rate": 3.869947478972123e-06, + "loss": 0.83453584, + "num_input_tokens_seen": 25147915, + "step": 1181, + "time_per_iteration": 2.874037742614746 + }, + { + "auxiliary_loss_clip": 0.01233647, + "auxiliary_loss_mlp": 0.01047042, + "balance_loss_clip": 1.06157649, + "balance_loss_mlp": 1.03622389, + "epoch": 0.14212709673540552, + "flos": 24022048199040.0, + "grad_norm": 1.8369731871490134, + "language_loss": 0.82088506, + "learning_rate": 3.869671022706412e-06, + "loss": 0.843692, + "num_input_tokens_seen": 25166645, + "step": 1182, + "time_per_iteration": 2.674643039703369 + }, + { + "auxiliary_loss_clip": 0.01241565, + "auxiliary_loss_mlp": 0.01046299, + "balance_loss_clip": 1.05333972, + "balance_loss_mlp": 1.03489053, + "epoch": 0.1422473396260446, + "flos": 26431797870720.0, + "grad_norm": 1.8261144349811818, + "language_loss": 0.64964598, + "learning_rate": 3.869394282813092e-06, + "loss": 0.67252469, + "num_input_tokens_seen": 25185845, + "step": 1183, + "time_per_iteration": 2.848761558532715 + }, + { + "auxiliary_loss_clip": 0.0126059, + "auxiliary_loss_mlp": 0.01036465, + "balance_loss_clip": 1.05668521, + "balance_loss_mlp": 1.02544451, + "epoch": 0.1423675825166837, + "flos": 17055306754560.0, + "grad_norm": 3.0619742726389227, + "language_loss": 0.89026368, + "learning_rate": 3.869117259334147e-06, + "loss": 0.91323417, + "num_input_tokens_seen": 25203770, + "step": 1184, + "time_per_iteration": 2.730396270751953 + }, + { + "auxiliary_loss_clip": 0.01236702, + "auxiliary_loss_mlp": 0.01043653, + "balance_loss_clip": 1.0633955, + "balance_loss_mlp": 1.03296626, + "epoch": 0.1424878254073228, + "flos": 17929480049280.0, + "grad_norm": 1.9103920406833326, + "language_loss": 0.81914812, + "learning_rate": 3.868839952311599e-06, + "loss": 0.84195173, + "num_input_tokens_seen": 25221725, + "step": 1185, + "time_per_iteration": 3.5649020671844482 + }, + { + "auxiliary_loss_clip": 0.0124493, + "auxiliary_loss_mlp": 0.0104047, + "balance_loss_clip": 1.06201446, + "balance_loss_mlp": 1.02925825, + "epoch": 0.14260806829796188, + "flos": 20303606407680.0, + "grad_norm": 2.556325440317037, + "language_loss": 0.80755043, + "learning_rate": 3.868562361787516e-06, + "loss": 0.8304044, + "num_input_tokens_seen": 25240855, + "step": 1186, + "time_per_iteration": 2.761486530303955 + }, + { + "auxiliary_loss_clip": 0.0127009, + "auxiliary_loss_mlp": 0.01042499, + "balance_loss_clip": 1.0503397, + "balance_loss_mlp": 1.03156757, + "epoch": 0.14272831118860096, + "flos": 23185724860800.0, + "grad_norm": 1.9673244193050154, + "language_loss": 0.68826175, + "learning_rate": 3.868284487804009e-06, + "loss": 0.71138763, + "num_input_tokens_seen": 25260085, + "step": 1187, + "time_per_iteration": 3.0546798706054688 + }, + { + "auxiliary_loss_clip": 0.01253852, + "auxiliary_loss_mlp": 0.01046607, + "balance_loss_clip": 1.06188285, + "balance_loss_mlp": 1.03584194, + "epoch": 0.14284855407924008, + "flos": 27232210586880.0, + "grad_norm": 1.5762973472822934, + "language_loss": 0.77924216, + "learning_rate": 3.86800633040323e-06, + "loss": 0.80224675, + "num_input_tokens_seen": 25280675, + "step": 1188, + "time_per_iteration": 3.668750286102295 + }, + { + "auxiliary_loss_clip": 0.01251927, + "auxiliary_loss_mlp": 0.01381888, + "balance_loss_clip": 1.0674659, + "balance_loss_mlp": 1.00020337, + "epoch": 0.14296879696987916, + "flos": 28184202696960.0, + "grad_norm": 2.0814673651181588, + "language_loss": 0.78144401, + "learning_rate": 3.867727889627376e-06, + "loss": 0.80778217, + "num_input_tokens_seen": 25300290, + "step": 1189, + "time_per_iteration": 3.6818480491638184 + }, + { + "auxiliary_loss_clip": 0.01245832, + "auxiliary_loss_mlp": 0.01043477, + "balance_loss_clip": 1.06126142, + "balance_loss_mlp": 1.03198481, + "epoch": 0.14308903986051824, + "flos": 19390290266880.0, + "grad_norm": 2.418611834791541, + "language_loss": 0.7831167, + "learning_rate": 3.867449165518687e-06, + "loss": 0.80600977, + "num_input_tokens_seen": 25316760, + "step": 1190, + "time_per_iteration": 3.6051743030548096 + }, + { + "auxiliary_loss_clip": 0.01239258, + "auxiliary_loss_mlp": 0.01381796, + "balance_loss_clip": 1.06611848, + "balance_loss_mlp": 1.00003529, + "epoch": 0.14320928275115732, + "flos": 17457506317440.0, + "grad_norm": 1.7539871350980603, + "language_loss": 0.71254206, + "learning_rate": 3.867170158119444e-06, + "loss": 0.7387526, + "num_input_tokens_seen": 25335760, + "step": 1191, + "time_per_iteration": 2.6002626419067383 + }, + { + "auxiliary_loss_clip": 0.01237681, + "auxiliary_loss_mlp": 0.01041547, + "balance_loss_clip": 1.06635904, + "balance_loss_mlp": 1.03024578, + "epoch": 0.14332952564179643, + "flos": 21466070259840.0, + "grad_norm": 2.072035459530546, + "language_loss": 0.7573843, + "learning_rate": 3.866890867471972e-06, + "loss": 0.78017652, + "num_input_tokens_seen": 25354230, + "step": 1192, + "time_per_iteration": 2.6269564628601074 + }, + { + "auxiliary_loss_clip": 0.01243036, + "auxiliary_loss_mlp": 0.01037652, + "balance_loss_clip": 1.05672097, + "balance_loss_mlp": 1.02602363, + "epoch": 0.14344976853243552, + "flos": 16396992241920.0, + "grad_norm": 2.1781097561067764, + "language_loss": 0.89451784, + "learning_rate": 3.86661129361864e-06, + "loss": 0.91732466, + "num_input_tokens_seen": 25368720, + "step": 1193, + "time_per_iteration": 2.5870048999786377 + }, + { + "auxiliary_loss_clip": 0.01250989, + "auxiliary_loss_mlp": 0.0104282, + "balance_loss_clip": 1.0645678, + "balance_loss_mlp": 1.03226423, + "epoch": 0.1435700114230746, + "flos": 18916736336640.0, + "grad_norm": 2.061734484716706, + "language_loss": 0.86335891, + "learning_rate": 3.866331436601859e-06, + "loss": 0.88629699, + "num_input_tokens_seen": 25386715, + "step": 1194, + "time_per_iteration": 2.708169937133789 + }, + { + "auxiliary_loss_clip": 0.01237338, + "auxiliary_loss_mlp": 0.01036308, + "balance_loss_clip": 1.06896329, + "balance_loss_mlp": 1.02543616, + "epoch": 0.1436902543137137, + "flos": 19755394058880.0, + "grad_norm": 2.679888783911945, + "language_loss": 0.73771405, + "learning_rate": 3.866051296464083e-06, + "loss": 0.76045048, + "num_input_tokens_seen": 25405550, + "step": 1195, + "time_per_iteration": 2.552137613296509 + }, + { + "auxiliary_loss_clip": 0.01237096, + "auxiliary_loss_mlp": 0.01382158, + "balance_loss_clip": 1.06565523, + "balance_loss_mlp": 1.00017214, + "epoch": 0.1438104972043528, + "flos": 14684807669760.0, + "grad_norm": 2.91876827999361, + "language_loss": 0.85159576, + "learning_rate": 3.86577087324781e-06, + "loss": 0.87778831, + "num_input_tokens_seen": 25422040, + "step": 1196, + "time_per_iteration": 2.635610342025757 + }, + { + "auxiliary_loss_clip": 0.01241101, + "auxiliary_loss_mlp": 0.01036818, + "balance_loss_clip": 1.06629908, + "balance_loss_mlp": 1.02655458, + "epoch": 0.14393074009499188, + "flos": 17092330698240.0, + "grad_norm": 1.9945291594088503, + "language_loss": 0.77658296, + "learning_rate": 3.865490166995578e-06, + "loss": 0.79936212, + "num_input_tokens_seen": 25440270, + "step": 1197, + "time_per_iteration": 2.643012523651123 + }, + { + "auxiliary_loss_clip": 0.01242689, + "auxiliary_loss_mlp": 0.01037955, + "balance_loss_clip": 1.06417501, + "balance_loss_mlp": 1.02732706, + "epoch": 0.144050982985631, + "flos": 30476200608000.0, + "grad_norm": 2.504147913157106, + "language_loss": 0.83922517, + "learning_rate": 3.86520917774997e-06, + "loss": 0.86203158, + "num_input_tokens_seen": 25459705, + "step": 1198, + "time_per_iteration": 2.7572996616363525 + }, + { + "auxiliary_loss_clip": 0.01241363, + "auxiliary_loss_mlp": 0.01045404, + "balance_loss_clip": 1.06532168, + "balance_loss_mlp": 1.0354383, + "epoch": 0.14417122587627007, + "flos": 17858484817920.0, + "grad_norm": 2.3341685833008756, + "language_loss": 0.75273478, + "learning_rate": 3.864927905553614e-06, + "loss": 0.7756024, + "num_input_tokens_seen": 25477615, + "step": 1199, + "time_per_iteration": 2.670346260070801 + }, + { + "auxiliary_loss_clip": 0.01254975, + "auxiliary_loss_mlp": 0.01038825, + "balance_loss_clip": 1.05935287, + "balance_loss_mlp": 1.02800059, + "epoch": 0.14429146876690915, + "flos": 21613914639360.0, + "grad_norm": 1.6462787880641276, + "language_loss": 0.88683671, + "learning_rate": 3.8646463504491765e-06, + "loss": 0.90977466, + "num_input_tokens_seen": 25497750, + "step": 1200, + "time_per_iteration": 2.7180826663970947 + }, + { + "auxiliary_loss_clip": 0.01248744, + "auxiliary_loss_mlp": 0.01041554, + "balance_loss_clip": 1.06841016, + "balance_loss_mlp": 1.03061664, + "epoch": 0.14441171165754824, + "flos": 23258120722560.0, + "grad_norm": 1.8373049762598082, + "language_loss": 0.83173621, + "learning_rate": 3.8643645124793705e-06, + "loss": 0.85463911, + "num_input_tokens_seen": 25516650, + "step": 1201, + "time_per_iteration": 2.6376466751098633 + }, + { + "auxiliary_loss_clip": 0.01239863, + "auxiliary_loss_mlp": 0.01034253, + "balance_loss_clip": 1.06328285, + "balance_loss_mlp": 1.02344668, + "epoch": 0.14453195454818735, + "flos": 42854213963520.0, + "grad_norm": 1.7157352235901768, + "language_loss": 0.74830514, + "learning_rate": 3.8640823916869515e-06, + "loss": 0.77104628, + "num_input_tokens_seen": 25540960, + "step": 1202, + "time_per_iteration": 2.86295223236084 + }, + { + "auxiliary_loss_clip": 0.01234361, + "auxiliary_loss_mlp": 0.01038571, + "balance_loss_clip": 1.06597471, + "balance_loss_mlp": 1.02788353, + "epoch": 0.14465219743882643, + "flos": 27235873774080.0, + "grad_norm": 1.6271552288609632, + "language_loss": 0.78293544, + "learning_rate": 3.863799988114714e-06, + "loss": 0.80566478, + "num_input_tokens_seen": 25562990, + "step": 1203, + "time_per_iteration": 2.6588082313537598 + }, + { + "auxiliary_loss_clip": 0.01236365, + "auxiliary_loss_mlp": 0.01044212, + "balance_loss_clip": 1.06596136, + "balance_loss_mlp": 1.03225505, + "epoch": 0.1447724403294655, + "flos": 16690705752960.0, + "grad_norm": 2.505742069428079, + "language_loss": 0.70551175, + "learning_rate": 3.863517301805502e-06, + "loss": 0.7283175, + "num_input_tokens_seen": 25581380, + "step": 1204, + "time_per_iteration": 2.560192346572876 + }, + { + "auxiliary_loss_clip": 0.01259919, + "auxiliary_loss_mlp": 0.01041793, + "balance_loss_clip": 1.06452274, + "balance_loss_mlp": 1.03095722, + "epoch": 0.14489268322010462, + "flos": 20073741321600.0, + "grad_norm": 2.6016958804443413, + "language_loss": 0.97447628, + "learning_rate": 3.863234332802196e-06, + "loss": 0.99749345, + "num_input_tokens_seen": 25593585, + "step": 1205, + "time_per_iteration": 2.692931890487671 + }, + { + "auxiliary_loss_clip": 0.01244993, + "auxiliary_loss_mlp": 0.01041529, + "balance_loss_clip": 1.06235588, + "balance_loss_mlp": 1.03085411, + "epoch": 0.1450129261107437, + "flos": 27125627955840.0, + "grad_norm": 2.4213836830465234, + "language_loss": 0.74047399, + "learning_rate": 3.862951081147723e-06, + "loss": 0.76333928, + "num_input_tokens_seen": 25613750, + "step": 1206, + "time_per_iteration": 2.734363555908203 + }, + { + "auxiliary_loss_clip": 0.01247737, + "auxiliary_loss_mlp": 0.01033477, + "balance_loss_clip": 1.06832099, + "balance_loss_mlp": 1.0234282, + "epoch": 0.1451331690013828, + "flos": 25702344472320.0, + "grad_norm": 2.6259517441416316, + "language_loss": 0.78230947, + "learning_rate": 3.862667546885053e-06, + "loss": 0.8051216, + "num_input_tokens_seen": 25632300, + "step": 1207, + "time_per_iteration": 2.700852394104004 + }, + { + "auxiliary_loss_clip": 0.01252449, + "auxiliary_loss_mlp": 0.01045183, + "balance_loss_clip": 1.05894732, + "balance_loss_mlp": 1.03341126, + "epoch": 0.14525341189202187, + "flos": 25737393168000.0, + "grad_norm": 1.9058631477164942, + "language_loss": 0.73380089, + "learning_rate": 3.8623837300571965e-06, + "loss": 0.75677717, + "num_input_tokens_seen": 25651285, + "step": 1208, + "time_per_iteration": 2.677182197570801 + }, + { + "auxiliary_loss_clip": 0.01237649, + "auxiliary_loss_mlp": 0.01049104, + "balance_loss_clip": 1.06716478, + "balance_loss_mlp": 1.03808928, + "epoch": 0.14537365478266098, + "flos": 23073898844160.0, + "grad_norm": 2.118944683676461, + "language_loss": 0.8381598, + "learning_rate": 3.8620996307072085e-06, + "loss": 0.8610273, + "num_input_tokens_seen": 25671990, + "step": 1209, + "time_per_iteration": 2.5845720767974854 + }, + { + "auxiliary_loss_clip": 0.01259438, + "auxiliary_loss_mlp": 0.010454, + "balance_loss_clip": 1.05521393, + "balance_loss_mlp": 1.03467095, + "epoch": 0.14549389767330007, + "flos": 20595021448320.0, + "grad_norm": 2.0006059672851624, + "language_loss": 0.64464706, + "learning_rate": 3.861815248878188e-06, + "loss": 0.66769546, + "num_input_tokens_seen": 25689475, + "step": 1210, + "time_per_iteration": 3.620697498321533 + }, + { + "auxiliary_loss_clip": 0.01247113, + "auxiliary_loss_mlp": 0.01038679, + "balance_loss_clip": 1.06474376, + "balance_loss_mlp": 1.02669907, + "epoch": 0.14561414056393915, + "flos": 15121804533120.0, + "grad_norm": 2.3149556912493354, + "language_loss": 0.79507303, + "learning_rate": 3.861530584613274e-06, + "loss": 0.81793094, + "num_input_tokens_seen": 25707475, + "step": 1211, + "time_per_iteration": 2.5934557914733887 + }, + { + "auxiliary_loss_clip": 0.01241857, + "auxiliary_loss_mlp": 0.01381787, + "balance_loss_clip": 1.06301272, + "balance_loss_mlp": 1.00014687, + "epoch": 0.14573438345457826, + "flos": 19427493778560.0, + "grad_norm": 2.1146836170550345, + "language_loss": 0.8232559, + "learning_rate": 3.86124563795565e-06, + "loss": 0.84949231, + "num_input_tokens_seen": 25726290, + "step": 1212, + "time_per_iteration": 2.623832941055298 + }, + { + "auxiliary_loss_clip": 0.01234485, + "auxiliary_loss_mlp": 0.01036737, + "balance_loss_clip": 1.06623602, + "balance_loss_mlp": 1.0263356, + "epoch": 0.14585462634521734, + "flos": 24828422572800.0, + "grad_norm": 1.5926610430668726, + "language_loss": 0.70190603, + "learning_rate": 3.860960408948543e-06, + "loss": 0.72461832, + "num_input_tokens_seen": 25748040, + "step": 1213, + "time_per_iteration": 2.6593587398529053 + }, + { + "auxiliary_loss_clip": 0.0123744, + "auxiliary_loss_mlp": 0.01040464, + "balance_loss_clip": 1.0668596, + "balance_loss_mlp": 1.03000343, + "epoch": 0.14597486923585642, + "flos": 15448627405440.0, + "grad_norm": 2.2952039106472286, + "language_loss": 0.89926386, + "learning_rate": 3.860674897635222e-06, + "loss": 0.92204285, + "num_input_tokens_seen": 25764525, + "step": 1214, + "time_per_iteration": 3.530162811279297 + }, + { + "auxiliary_loss_clip": 0.0124014, + "auxiliary_loss_mlp": 0.01035457, + "balance_loss_clip": 1.06502295, + "balance_loss_mlp": 1.02455568, + "epoch": 0.1460951121264955, + "flos": 16655154266880.0, + "grad_norm": 1.7773272422543471, + "language_loss": 0.83431089, + "learning_rate": 3.860389104058998e-06, + "loss": 0.85706687, + "num_input_tokens_seen": 25782755, + "step": 1215, + "time_per_iteration": 3.3574562072753906 + }, + { + "auxiliary_loss_clip": 0.01250557, + "auxiliary_loss_mlp": 0.01036897, + "balance_loss_clip": 1.0622046, + "balance_loss_mlp": 1.02587616, + "epoch": 0.14621535501713462, + "flos": 24863291700480.0, + "grad_norm": 2.6276404318133375, + "language_loss": 0.72727007, + "learning_rate": 3.860103028263227e-06, + "loss": 0.7501446, + "num_input_tokens_seen": 25805860, + "step": 1216, + "time_per_iteration": 3.4384267330169678 + }, + { + "auxiliary_loss_clip": 0.01258219, + "auxiliary_loss_mlp": 0.01037176, + "balance_loss_clip": 1.0534482, + "balance_loss_mlp": 1.02668536, + "epoch": 0.1463355979077737, + "flos": 25228000442880.0, + "grad_norm": 2.3068358448769106, + "language_loss": 0.69956201, + "learning_rate": 3.859816670291304e-06, + "loss": 0.72251594, + "num_input_tokens_seen": 25824955, + "step": 1217, + "time_per_iteration": 2.699249505996704 + }, + { + "auxiliary_loss_clip": 0.012608, + "auxiliary_loss_mlp": 0.01045158, + "balance_loss_clip": 1.0548898, + "balance_loss_mlp": 1.03434563, + "epoch": 0.14645584079841278, + "flos": 22054143726720.0, + "grad_norm": 3.1722174821571194, + "language_loss": 0.89575696, + "learning_rate": 3.859530030186672e-06, + "loss": 0.91881657, + "num_input_tokens_seen": 25841965, + "step": 1218, + "time_per_iteration": 2.6557633876800537 + }, + { + "auxiliary_loss_clip": 0.01253485, + "auxiliary_loss_mlp": 0.01044927, + "balance_loss_clip": 1.06436062, + "balance_loss_mlp": 1.03325677, + "epoch": 0.1465760836890519, + "flos": 23623870959360.0, + "grad_norm": 2.4429193656910866, + "language_loss": 0.82666939, + "learning_rate": 3.859243107992813e-06, + "loss": 0.84965354, + "num_input_tokens_seen": 25860770, + "step": 1219, + "time_per_iteration": 2.6700849533081055 + }, + { + "auxiliary_loss_clip": 0.01254165, + "auxiliary_loss_mlp": 0.01040169, + "balance_loss_clip": 1.05265307, + "balance_loss_mlp": 1.02874887, + "epoch": 0.14669632657969098, + "flos": 37407893356800.0, + "grad_norm": 4.333260909830189, + "language_loss": 0.78177047, + "learning_rate": 3.858955903753252e-06, + "loss": 0.80471379, + "num_input_tokens_seen": 25879410, + "step": 1220, + "time_per_iteration": 2.7940683364868164 + }, + { + "auxiliary_loss_clip": 0.01241068, + "auxiliary_loss_mlp": 0.01046355, + "balance_loss_clip": 1.0626328, + "balance_loss_mlp": 1.03671145, + "epoch": 0.14681656947033006, + "flos": 28365910623360.0, + "grad_norm": 1.4820199292324203, + "language_loss": 0.83303934, + "learning_rate": 3.858668417511559e-06, + "loss": 0.85591358, + "num_input_tokens_seen": 25902160, + "step": 1221, + "time_per_iteration": 2.7506678104400635 + }, + { + "auxiliary_loss_clip": 0.0125558, + "auxiliary_loss_mlp": 0.01040587, + "balance_loss_clip": 1.06632447, + "balance_loss_mlp": 1.02966177, + "epoch": 0.14693681236096917, + "flos": 18479488078080.0, + "grad_norm": 2.0163762703520813, + "language_loss": 0.76221657, + "learning_rate": 3.8583806493113445e-06, + "loss": 0.78517824, + "num_input_tokens_seen": 25920505, + "step": 1222, + "time_per_iteration": 2.6534156799316406 + }, + { + "auxiliary_loss_clip": 0.01236665, + "auxiliary_loss_mlp": 0.01041335, + "balance_loss_clip": 1.06252313, + "balance_loss_mlp": 1.02995062, + "epoch": 0.14705705525160825, + "flos": 20777806782720.0, + "grad_norm": 2.126152118238381, + "language_loss": 0.82370818, + "learning_rate": 3.858092599196263e-06, + "loss": 0.84648818, + "num_input_tokens_seen": 25938460, + "step": 1223, + "time_per_iteration": 2.630174160003662 + }, + { + "auxiliary_loss_clip": 0.01243251, + "auxiliary_loss_mlp": 0.01034795, + "balance_loss_clip": 1.06468463, + "balance_loss_mlp": 1.02369642, + "epoch": 0.14717729814224734, + "flos": 29932944336000.0, + "grad_norm": 5.765154514129177, + "language_loss": 0.82613689, + "learning_rate": 3.857804267210012e-06, + "loss": 0.84891737, + "num_input_tokens_seen": 25957760, + "step": 1224, + "time_per_iteration": 2.665001392364502 + }, + { + "auxiliary_loss_clip": 0.01240158, + "auxiliary_loss_mlp": 0.01031178, + "balance_loss_clip": 1.05635834, + "balance_loss_mlp": 1.02077055, + "epoch": 0.14729754103288642, + "flos": 20047491457920.0, + "grad_norm": 2.261164166730256, + "language_loss": 0.88350868, + "learning_rate": 3.857515653396331e-06, + "loss": 0.90622199, + "num_input_tokens_seen": 25974970, + "step": 1225, + "time_per_iteration": 2.6193864345550537 + }, + { + "auxiliary_loss_clip": 0.01265914, + "auxiliary_loss_mlp": 0.01038764, + "balance_loss_clip": 1.0595789, + "balance_loss_mlp": 1.02736807, + "epoch": 0.14741778392352553, + "flos": 19281516906240.0, + "grad_norm": 2.432559341316473, + "language_loss": 0.87115926, + "learning_rate": 3.857226757799002e-06, + "loss": 0.89420605, + "num_input_tokens_seen": 25992525, + "step": 1226, + "time_per_iteration": 2.7351951599121094 + }, + { + "auxiliary_loss_clip": 0.01250135, + "auxiliary_loss_mlp": 0.01039032, + "balance_loss_clip": 1.06148887, + "balance_loss_mlp": 1.02789199, + "epoch": 0.1475380268141646, + "flos": 25411108999680.0, + "grad_norm": 2.56243921941046, + "language_loss": 0.74075413, + "learning_rate": 3.85693758046185e-06, + "loss": 0.76364577, + "num_input_tokens_seen": 26010815, + "step": 1227, + "time_per_iteration": 2.6531858444213867 + }, + { + "auxiliary_loss_clip": 0.01235244, + "auxiliary_loss_mlp": 0.01045468, + "balance_loss_clip": 1.06914401, + "balance_loss_mlp": 1.0350188, + "epoch": 0.1476582697048037, + "flos": 20847652778880.0, + "grad_norm": 2.073841808908013, + "language_loss": 0.8284837, + "learning_rate": 3.8566481214287435e-06, + "loss": 0.85129082, + "num_input_tokens_seen": 26028935, + "step": 1228, + "time_per_iteration": 2.6322953701019287 + }, + { + "auxiliary_loss_clip": 0.01242965, + "auxiliary_loss_mlp": 0.01032554, + "balance_loss_clip": 1.05369067, + "balance_loss_mlp": 1.02217698, + "epoch": 0.1477785125954428, + "flos": 14028109269120.0, + "grad_norm": 2.3613450216922764, + "language_loss": 0.90731382, + "learning_rate": 3.8563583807435935e-06, + "loss": 0.93006909, + "num_input_tokens_seen": 26045080, + "step": 1229, + "time_per_iteration": 2.7543063163757324 + }, + { + "auxiliary_loss_clip": 0.01244027, + "auxiliary_loss_mlp": 0.01381612, + "balance_loss_clip": 1.06436002, + "balance_loss_mlp": 1.00012648, + "epoch": 0.1478987554860819, + "flos": 20516699842560.0, + "grad_norm": 2.0069182680767303, + "language_loss": 0.77749407, + "learning_rate": 3.856068358450353e-06, + "loss": 0.80375046, + "num_input_tokens_seen": 26065030, + "step": 1230, + "time_per_iteration": 2.6671156883239746 + }, + { + "auxiliary_loss_clip": 0.01245459, + "auxiliary_loss_mlp": 0.01042165, + "balance_loss_clip": 1.06671405, + "balance_loss_mlp": 1.03153181, + "epoch": 0.14801899837672097, + "flos": 17857012360320.0, + "grad_norm": 1.6695676914134092, + "language_loss": 0.85843688, + "learning_rate": 3.8557780545930186e-06, + "loss": 0.88131315, + "num_input_tokens_seen": 26083445, + "step": 1231, + "time_per_iteration": 2.6252524852752686 + }, + { + "auxiliary_loss_clip": 0.01244993, + "auxiliary_loss_mlp": 0.01048371, + "balance_loss_clip": 1.06205869, + "balance_loss_mlp": 1.03723121, + "epoch": 0.14813924126736006, + "flos": 20881408584960.0, + "grad_norm": 1.9354448815825036, + "language_loss": 0.79282755, + "learning_rate": 3.855487469215628e-06, + "loss": 0.81576121, + "num_input_tokens_seen": 26102375, + "step": 1232, + "time_per_iteration": 2.669377088546753 + }, + { + "auxiliary_loss_clip": 0.01254477, + "auxiliary_loss_mlp": 0.01045187, + "balance_loss_clip": 1.06119108, + "balance_loss_mlp": 1.03414774, + "epoch": 0.14825948415799917, + "flos": 37414070496000.0, + "grad_norm": 2.1174721567050443, + "language_loss": 0.72433341, + "learning_rate": 3.855196602362264e-06, + "loss": 0.74733001, + "num_input_tokens_seen": 26125295, + "step": 1233, + "time_per_iteration": 2.8212313652038574 + }, + { + "auxiliary_loss_clip": 0.01242046, + "auxiliary_loss_mlp": 0.01036808, + "balance_loss_clip": 1.06259596, + "balance_loss_mlp": 1.02579916, + "epoch": 0.14837972704863825, + "flos": 22014641744640.0, + "grad_norm": 1.9920631660818648, + "language_loss": 0.94334328, + "learning_rate": 3.854905454077051e-06, + "loss": 0.96613181, + "num_input_tokens_seen": 26142905, + "step": 1234, + "time_per_iteration": 2.6197023391723633 + }, + { + "auxiliary_loss_clip": 0.01272123, + "auxiliary_loss_mlp": 0.01033602, + "balance_loss_clip": 1.05100608, + "balance_loss_mlp": 1.02316499, + "epoch": 0.14849996993927733, + "flos": 20996323171200.0, + "grad_norm": 1.7823218250702482, + "language_loss": 0.88213831, + "learning_rate": 3.854614024404155e-06, + "loss": 0.90519553, + "num_input_tokens_seen": 26161215, + "step": 1235, + "time_per_iteration": 2.8798861503601074 + }, + { + "auxiliary_loss_clip": 0.01228867, + "auxiliary_loss_mlp": 0.01043855, + "balance_loss_clip": 1.05416226, + "balance_loss_mlp": 1.03214872, + "epoch": 0.14862021282991644, + "flos": 20047994248320.0, + "grad_norm": 2.9529345707333183, + "language_loss": 0.89312464, + "learning_rate": 3.8543223133877865e-06, + "loss": 0.91585183, + "num_input_tokens_seen": 26179810, + "step": 1236, + "time_per_iteration": 3.7071595191955566 + }, + { + "auxiliary_loss_clip": 0.01225383, + "auxiliary_loss_mlp": 0.01035666, + "balance_loss_clip": 1.05745733, + "balance_loss_mlp": 1.02413273, + "epoch": 0.14874045572055553, + "flos": 22712027276160.0, + "grad_norm": 1.736881754220249, + "language_loss": 0.88513786, + "learning_rate": 3.854030321072198e-06, + "loss": 0.90774834, + "num_input_tokens_seen": 26199715, + "step": 1237, + "time_per_iteration": 2.742809534072876 + }, + { + "auxiliary_loss_clip": 0.01262269, + "auxiliary_loss_mlp": 0.01036149, + "balance_loss_clip": 1.06085396, + "balance_loss_mlp": 1.02458537, + "epoch": 0.1488606986111946, + "flos": 25411288567680.0, + "grad_norm": 1.9392528566088445, + "language_loss": 0.73516971, + "learning_rate": 3.853738047501682e-06, + "loss": 0.75815386, + "num_input_tokens_seen": 26220275, + "step": 1238, + "time_per_iteration": 2.740222692489624 + }, + { + "auxiliary_loss_clip": 0.01244378, + "auxiliary_loss_mlp": 0.01048055, + "balance_loss_clip": 1.06779504, + "balance_loss_mlp": 1.03652167, + "epoch": 0.1489809415018337, + "flos": 17018749687680.0, + "grad_norm": 1.9390989031223553, + "language_loss": 0.77806151, + "learning_rate": 3.85344549272058e-06, + "loss": 0.80098581, + "num_input_tokens_seen": 26238255, + "step": 1239, + "time_per_iteration": 2.6630818843841553 + }, + { + "auxiliary_loss_clip": 0.01233392, + "auxiliary_loss_mlp": 0.01040407, + "balance_loss_clip": 1.06182265, + "balance_loss_mlp": 1.02955258, + "epoch": 0.1491011843924728, + "flos": 33659394860160.0, + "grad_norm": 1.8701860043065608, + "language_loss": 0.82797039, + "learning_rate": 3.853152656773269e-06, + "loss": 0.85070843, + "num_input_tokens_seen": 26259690, + "step": 1240, + "time_per_iteration": 4.564885854721069 + }, + { + "auxiliary_loss_clip": 0.01243765, + "auxiliary_loss_mlp": 0.01041289, + "balance_loss_clip": 1.0605793, + "balance_loss_mlp": 1.03107834, + "epoch": 0.14922142728311188, + "flos": 21179000764800.0, + "grad_norm": 1.6727673965245895, + "language_loss": 0.84802222, + "learning_rate": 3.852859539704174e-06, + "loss": 0.87087274, + "num_input_tokens_seen": 26278990, + "step": 1241, + "time_per_iteration": 2.680511951446533 + }, + { + "auxiliary_loss_clip": 0.0126082, + "auxiliary_loss_mlp": 0.01042634, + "balance_loss_clip": 1.05744529, + "balance_loss_mlp": 1.03178024, + "epoch": 0.14934167017375097, + "flos": 29860548474240.0, + "grad_norm": 1.890506530214889, + "language_loss": 0.76674008, + "learning_rate": 3.85256614155776e-06, + "loss": 0.7897746, + "num_input_tokens_seen": 26299120, + "step": 1242, + "time_per_iteration": 3.788922071456909 + }, + { + "auxiliary_loss_clip": 0.01240369, + "auxiliary_loss_mlp": 0.01047152, + "balance_loss_clip": 1.06187773, + "balance_loss_mlp": 1.03704894, + "epoch": 0.14946191306439008, + "flos": 17019216564480.0, + "grad_norm": 1.8868194501827713, + "language_loss": 0.74192905, + "learning_rate": 3.852272462378535e-06, + "loss": 0.76480424, + "num_input_tokens_seen": 26316995, + "step": 1243, + "time_per_iteration": 2.6517581939697266 + }, + { + "auxiliary_loss_clip": 0.01249096, + "auxiliary_loss_mlp": 0.010356, + "balance_loss_clip": 1.06231952, + "balance_loss_mlp": 1.02562225, + "epoch": 0.14958215595502916, + "flos": 15669047214720.0, + "grad_norm": 2.1225007428476883, + "language_loss": 0.77624118, + "learning_rate": 3.85197850221105e-06, + "loss": 0.79908818, + "num_input_tokens_seen": 26333295, + "step": 1244, + "time_per_iteration": 2.6403255462646484 + }, + { + "auxiliary_loss_clip": 0.01241302, + "auxiliary_loss_mlp": 0.01039709, + "balance_loss_clip": 1.0683701, + "balance_loss_mlp": 1.02931428, + "epoch": 0.14970239884566824, + "flos": 33108560818560.0, + "grad_norm": 1.7556639743076419, + "language_loss": 0.75368714, + "learning_rate": 3.851684261099899e-06, + "loss": 0.77649724, + "num_input_tokens_seen": 26355035, + "step": 1245, + "time_per_iteration": 2.7428112030029297 + }, + { + "auxiliary_loss_clip": 0.01243468, + "auxiliary_loss_mlp": 0.01042383, + "balance_loss_clip": 1.06049609, + "balance_loss_mlp": 1.03069448, + "epoch": 0.14982264173630733, + "flos": 17821245392640.0, + "grad_norm": 2.6439092953150207, + "language_loss": 0.86632133, + "learning_rate": 3.851389739089718e-06, + "loss": 0.88917983, + "num_input_tokens_seen": 26371655, + "step": 1246, + "time_per_iteration": 2.6314680576324463 + }, + { + "auxiliary_loss_clip": 0.01245237, + "auxiliary_loss_mlp": 0.01038224, + "balance_loss_clip": 1.06856573, + "balance_loss_mlp": 1.02592206, + "epoch": 0.14994288462694644, + "flos": 32409559175040.0, + "grad_norm": 1.8449225573188128, + "language_loss": 0.80428183, + "learning_rate": 3.851094936225186e-06, + "loss": 0.82711643, + "num_input_tokens_seen": 26392540, + "step": 1247, + "time_per_iteration": 2.7557215690612793 + }, + { + "auxiliary_loss_clip": 0.0124276, + "auxiliary_loss_mlp": 0.01036277, + "balance_loss_clip": 1.06229663, + "balance_loss_mlp": 1.02579284, + "epoch": 0.15006312751758552, + "flos": 31794661226880.0, + "grad_norm": 1.4163819361845202, + "language_loss": 0.76575351, + "learning_rate": 3.850799852551024e-06, + "loss": 0.78854382, + "num_input_tokens_seen": 26414960, + "step": 1248, + "time_per_iteration": 2.9116315841674805 + }, + { + "auxiliary_loss_clip": 0.01231289, + "auxiliary_loss_mlp": 0.0103495, + "balance_loss_clip": 1.06150711, + "balance_loss_mlp": 1.02403069, + "epoch": 0.1501833704082246, + "flos": 16618022582400.0, + "grad_norm": 2.209885399057956, + "language_loss": 0.86493337, + "learning_rate": 3.850504488111995e-06, + "loss": 0.88759577, + "num_input_tokens_seen": 26431635, + "step": 1249, + "time_per_iteration": 2.656726121902466 + }, + { + "auxiliary_loss_clip": 0.01243708, + "auxiliary_loss_mlp": 0.01042646, + "balance_loss_clip": 1.06185973, + "balance_loss_mlp": 1.03175616, + "epoch": 0.15030361329886371, + "flos": 23471178243840.0, + "grad_norm": 2.3164309179759535, + "language_loss": 0.82617652, + "learning_rate": 3.850208842952907e-06, + "loss": 0.84904009, + "num_input_tokens_seen": 26450440, + "step": 1250, + "time_per_iteration": 2.70021390914917 + }, + { + "auxiliary_loss_clip": 0.01271101, + "auxiliary_loss_mlp": 0.01039805, + "balance_loss_clip": 1.05811858, + "balance_loss_mlp": 1.02911735, + "epoch": 0.1504238561895028, + "flos": 25629409906560.0, + "grad_norm": 1.7079443190326562, + "language_loss": 0.79134274, + "learning_rate": 3.849912917118608e-06, + "loss": 0.81445187, + "num_input_tokens_seen": 26471480, + "step": 1251, + "time_per_iteration": 2.823718547821045 + }, + { + "auxiliary_loss_clip": 0.01144075, + "auxiliary_loss_mlp": 0.01012376, + "balance_loss_clip": 1.03632998, + "balance_loss_mlp": 1.00938416, + "epoch": 0.15054409908014188, + "flos": 52095146129280.0, + "grad_norm": 0.8918779357704675, + "language_loss": 0.59324408, + "learning_rate": 3.849616710653992e-06, + "loss": 0.61480862, + "num_input_tokens_seen": 26532950, + "step": 1252, + "time_per_iteration": 3.238067865371704 + }, + { + "auxiliary_loss_clip": 0.01240018, + "auxiliary_loss_mlp": 0.01044193, + "balance_loss_clip": 1.06411791, + "balance_loss_mlp": 1.03277314, + "epoch": 0.150664341970781, + "flos": 18880251096960.0, + "grad_norm": 1.8573177953246416, + "language_loss": 0.74908006, + "learning_rate": 3.84932022360399e-06, + "loss": 0.77192217, + "num_input_tokens_seen": 26551615, + "step": 1253, + "time_per_iteration": 2.6747000217437744 + }, + { + "auxiliary_loss_clip": 0.01247995, + "auxiliary_loss_mlp": 0.0103454, + "balance_loss_clip": 1.06635618, + "balance_loss_mlp": 1.02416325, + "epoch": 0.15078458486142007, + "flos": 22163240309760.0, + "grad_norm": 3.327089480002559, + "language_loss": 0.84138334, + "learning_rate": 3.849023456013581e-06, + "loss": 0.8642087, + "num_input_tokens_seen": 26569175, + "step": 1254, + "time_per_iteration": 2.6964638233184814 + }, + { + "auxiliary_loss_clip": 0.01248981, + "auxiliary_loss_mlp": 0.01046029, + "balance_loss_clip": 1.06833947, + "balance_loss_mlp": 1.03490651, + "epoch": 0.15090482775205916, + "flos": 26651894457600.0, + "grad_norm": 2.2486909913212525, + "language_loss": 0.62583989, + "learning_rate": 3.848726407927784e-06, + "loss": 0.64879, + "num_input_tokens_seen": 26589560, + "step": 1255, + "time_per_iteration": 2.6672611236572266 + }, + { + "auxiliary_loss_clip": 0.01254074, + "auxiliary_loss_mlp": 0.01033367, + "balance_loss_clip": 1.06569052, + "balance_loss_mlp": 1.02234066, + "epoch": 0.15102507064269824, + "flos": 21798998444160.0, + "grad_norm": 2.2839056772222754, + "language_loss": 0.86728621, + "learning_rate": 3.84842907939166e-06, + "loss": 0.89016062, + "num_input_tokens_seen": 26608785, + "step": 1256, + "time_per_iteration": 2.6920433044433594 + }, + { + "auxiliary_loss_clip": 0.01249625, + "auxiliary_loss_mlp": 0.01039765, + "balance_loss_clip": 1.06002581, + "balance_loss_mlp": 1.02962601, + "epoch": 0.15114531353333735, + "flos": 22820908377600.0, + "grad_norm": 2.5013383728537812, + "language_loss": 0.70755816, + "learning_rate": 3.8481314704503146e-06, + "loss": 0.73045194, + "num_input_tokens_seen": 26628615, + "step": 1257, + "time_per_iteration": 2.714205265045166 + }, + { + "auxiliary_loss_clip": 0.01242728, + "auxiliary_loss_mlp": 0.010478, + "balance_loss_clip": 1.0693841, + "balance_loss_mlp": 1.0378226, + "epoch": 0.15126555642397643, + "flos": 19682674974720.0, + "grad_norm": 2.2025557727922087, + "language_loss": 0.88490438, + "learning_rate": 3.847833581148895e-06, + "loss": 0.90780973, + "num_input_tokens_seen": 26647525, + "step": 1258, + "time_per_iteration": 2.6513967514038086 + }, + { + "auxiliary_loss_clip": 0.01230039, + "auxiliary_loss_mlp": 0.0103377, + "balance_loss_clip": 1.06299031, + "balance_loss_mlp": 1.02335072, + "epoch": 0.15138579931461552, + "flos": 28726022424960.0, + "grad_norm": 2.5431177379688292, + "language_loss": 0.80917358, + "learning_rate": 3.84753541153259e-06, + "loss": 0.83181167, + "num_input_tokens_seen": 26667095, + "step": 1259, + "time_per_iteration": 2.6237621307373047 + }, + { + "auxiliary_loss_clip": 0.01242691, + "auxiliary_loss_mlp": 0.01036714, + "balance_loss_clip": 1.06545115, + "balance_loss_mlp": 1.02556837, + "epoch": 0.15150604220525463, + "flos": 22127006465280.0, + "grad_norm": 1.5649967163281153, + "language_loss": 0.83267576, + "learning_rate": 3.847236961646633e-06, + "loss": 0.85546982, + "num_input_tokens_seen": 26686075, + "step": 1260, + "time_per_iteration": 2.663902521133423 + }, + { + "auxiliary_loss_clip": 0.01238926, + "auxiliary_loss_mlp": 0.01033961, + "balance_loss_clip": 1.06088388, + "balance_loss_mlp": 1.02308333, + "epoch": 0.1516262850958937, + "flos": 12968708515200.0, + "grad_norm": 2.931342662692437, + "language_loss": 0.78219104, + "learning_rate": 3.846938231536296e-06, + "loss": 0.80491996, + "num_input_tokens_seen": 26701695, + "step": 1261, + "time_per_iteration": 2.624307632446289 + }, + { + "auxiliary_loss_clip": 0.01245577, + "auxiliary_loss_mlp": 0.01036793, + "balance_loss_clip": 1.06811786, + "balance_loss_mlp": 1.02670205, + "epoch": 0.1517465279865328, + "flos": 21797130936960.0, + "grad_norm": 1.8339826185979329, + "language_loss": 0.80937016, + "learning_rate": 3.8466392212468995e-06, + "loss": 0.83219391, + "num_input_tokens_seen": 26721885, + "step": 1262, + "time_per_iteration": 3.541339874267578 + }, + { + "auxiliary_loss_clip": 0.01157118, + "auxiliary_loss_mlp": 0.01015323, + "balance_loss_clip": 1.02542496, + "balance_loss_mlp": 1.01231909, + "epoch": 0.15186677087717187, + "flos": 58174569901440.0, + "grad_norm": 0.8223994291592278, + "language_loss": 0.61946034, + "learning_rate": 3.8463399308238e-06, + "loss": 0.64118469, + "num_input_tokens_seen": 26780990, + "step": 1263, + "time_per_iteration": 3.2163665294647217 + }, + { + "auxiliary_loss_clip": 0.01242394, + "auxiliary_loss_mlp": 0.01039078, + "balance_loss_clip": 1.06633055, + "balance_loss_mlp": 1.02887321, + "epoch": 0.15198701376781099, + "flos": 32669696448000.0, + "grad_norm": 2.0116318259965724, + "language_loss": 0.64165282, + "learning_rate": 3.846040360312402e-06, + "loss": 0.66446757, + "num_input_tokens_seen": 26804250, + "step": 1264, + "time_per_iteration": 2.722717523574829 + }, + { + "auxiliary_loss_clip": 0.01232627, + "auxiliary_loss_mlp": 0.01033595, + "balance_loss_clip": 1.06515861, + "balance_loss_mlp": 1.02282476, + "epoch": 0.15210725665845007, + "flos": 28402575431040.0, + "grad_norm": 2.1830117738239068, + "language_loss": 0.80780685, + "learning_rate": 3.8457405097581485e-06, + "loss": 0.83046913, + "num_input_tokens_seen": 26823240, + "step": 1265, + "time_per_iteration": 2.682340621948242 + }, + { + "auxiliary_loss_clip": 0.01264233, + "auxiliary_loss_mlp": 0.01043043, + "balance_loss_clip": 1.05587709, + "balance_loss_mlp": 1.03290975, + "epoch": 0.15222749954908915, + "flos": 19938179393280.0, + "grad_norm": 1.8208539086926256, + "language_loss": 0.77914488, + "learning_rate": 3.8454403792065275e-06, + "loss": 0.80221766, + "num_input_tokens_seen": 26842060, + "step": 1266, + "time_per_iteration": 3.8099865913391113 + }, + { + "auxiliary_loss_clip": 0.0123955, + "auxiliary_loss_mlp": 0.01040307, + "balance_loss_clip": 1.05628085, + "balance_loss_mlp": 1.03057337, + "epoch": 0.15234774243972826, + "flos": 21324223451520.0, + "grad_norm": 2.044273860323503, + "language_loss": 0.85727251, + "learning_rate": 3.845139968703068e-06, + "loss": 0.88007104, + "num_input_tokens_seen": 26859580, + "step": 1267, + "time_per_iteration": 2.6759722232818604 + }, + { + "auxiliary_loss_clip": 0.01258546, + "auxiliary_loss_mlp": 0.01034954, + "balance_loss_clip": 1.0568068, + "balance_loss_mlp": 1.02449965, + "epoch": 0.15246798533036734, + "flos": 25957812977280.0, + "grad_norm": 6.703336671679978, + "language_loss": 0.82810783, + "learning_rate": 3.844839278293342e-06, + "loss": 0.85104281, + "num_input_tokens_seen": 26880430, + "step": 1268, + "time_per_iteration": 3.650465250015259 + }, + { + "auxiliary_loss_clip": 0.01233647, + "auxiliary_loss_mlp": 0.01033952, + "balance_loss_clip": 1.06690073, + "balance_loss_mlp": 1.02294338, + "epoch": 0.15258822822100643, + "flos": 25811907932160.0, + "grad_norm": 2.3583696973059056, + "language_loss": 0.77068806, + "learning_rate": 3.8445383080229654e-06, + "loss": 0.79336405, + "num_input_tokens_seen": 26896445, + "step": 1269, + "time_per_iteration": 2.632754325866699 + }, + { + "auxiliary_loss_clip": 0.01234512, + "auxiliary_loss_mlp": 0.01041547, + "balance_loss_clip": 1.05740833, + "balance_loss_mlp": 1.03049684, + "epoch": 0.1527084711116455, + "flos": 25265455349760.0, + "grad_norm": 1.9871615530738882, + "language_loss": 0.73492694, + "learning_rate": 3.844237057937593e-06, + "loss": 0.75768757, + "num_input_tokens_seen": 26915450, + "step": 1270, + "time_per_iteration": 2.734858751296997 + }, + { + "auxiliary_loss_clip": 0.01243132, + "auxiliary_loss_mlp": 0.01045754, + "balance_loss_clip": 1.06224275, + "balance_loss_mlp": 1.03556776, + "epoch": 0.15282871400228462, + "flos": 29240227572480.0, + "grad_norm": 2.375667212518203, + "language_loss": 0.7797873, + "learning_rate": 3.843935528082926e-06, + "loss": 0.80267614, + "num_input_tokens_seen": 26936475, + "step": 1271, + "time_per_iteration": 2.6545770168304443 + }, + { + "auxiliary_loss_clip": 0.01243423, + "auxiliary_loss_mlp": 0.0103772, + "balance_loss_clip": 1.06469083, + "balance_loss_mlp": 1.02668095, + "epoch": 0.1529489568929237, + "flos": 20882952869760.0, + "grad_norm": 2.240601561360015, + "language_loss": 0.84784913, + "learning_rate": 3.843633718504704e-06, + "loss": 0.87066048, + "num_input_tokens_seen": 26954920, + "step": 1272, + "time_per_iteration": 2.683396100997925 + }, + { + "auxiliary_loss_clip": 0.01259303, + "auxiliary_loss_mlp": 0.01034406, + "balance_loss_clip": 1.06087697, + "balance_loss_mlp": 1.02395689, + "epoch": 0.1530691997835628, + "flos": 20083833043200.0, + "grad_norm": 2.2885108554171465, + "language_loss": 0.90272886, + "learning_rate": 3.843331629248715e-06, + "loss": 0.92566592, + "num_input_tokens_seen": 26972520, + "step": 1273, + "time_per_iteration": 2.7180676460266113 + }, + { + "auxiliary_loss_clip": 0.01235043, + "auxiliary_loss_mlp": 0.01039948, + "balance_loss_clip": 1.06963682, + "balance_loss_mlp": 1.02884984, + "epoch": 0.1531894426742019, + "flos": 28759814144640.0, + "grad_norm": 1.9920132813235931, + "language_loss": 0.76325589, + "learning_rate": 3.843029260360782e-06, + "loss": 0.7860058, + "num_input_tokens_seen": 26990890, + "step": 1274, + "time_per_iteration": 2.672095537185669 + }, + { + "auxiliary_loss_clip": 0.0123946, + "auxiliary_loss_mlp": 0.01038533, + "balance_loss_clip": 1.06747103, + "balance_loss_mlp": 1.02823901, + "epoch": 0.15330968556484098, + "flos": 22236282616320.0, + "grad_norm": 1.8265221945883858, + "language_loss": 0.78973824, + "learning_rate": 3.8427266118867755e-06, + "loss": 0.81251818, + "num_input_tokens_seen": 27010640, + "step": 1275, + "time_per_iteration": 2.6803102493286133 + }, + { + "auxiliary_loss_clip": 0.012432, + "auxiliary_loss_mlp": 0.01036515, + "balance_loss_clip": 1.06209064, + "balance_loss_mlp": 1.02550006, + "epoch": 0.15342992845548006, + "flos": 27527504296320.0, + "grad_norm": 2.077312329810287, + "language_loss": 0.82464838, + "learning_rate": 3.842423683872608e-06, + "loss": 0.84744561, + "num_input_tokens_seen": 27031215, + "step": 1276, + "time_per_iteration": 2.6996984481811523 + }, + { + "auxiliary_loss_clip": 0.01235864, + "auxiliary_loss_mlp": 0.01040539, + "balance_loss_clip": 1.05996501, + "balance_loss_mlp": 1.02970886, + "epoch": 0.15355017134611917, + "flos": 19609596754560.0, + "grad_norm": 2.4176625966667826, + "language_loss": 0.7776252, + "learning_rate": 3.842120476364232e-06, + "loss": 0.80038917, + "num_input_tokens_seen": 27049665, + "step": 1277, + "time_per_iteration": 2.5902175903320312 + }, + { + "auxiliary_loss_clip": 0.01241478, + "auxiliary_loss_mlp": 0.01033334, + "balance_loss_clip": 1.06264722, + "balance_loss_mlp": 1.02212834, + "epoch": 0.15367041423675826, + "flos": 18478590238080.0, + "grad_norm": 2.1005142195135234, + "language_loss": 0.83890027, + "learning_rate": 3.841816989407644e-06, + "loss": 0.86164838, + "num_input_tokens_seen": 27065155, + "step": 1278, + "time_per_iteration": 2.582883834838867 + }, + { + "auxiliary_loss_clip": 0.01249178, + "auxiliary_loss_mlp": 0.01037308, + "balance_loss_clip": 1.05766428, + "balance_loss_mlp": 1.02711606, + "epoch": 0.15379065712739734, + "flos": 41427662342400.0, + "grad_norm": 2.941670692000881, + "language_loss": 0.7668891, + "learning_rate": 3.841513223048884e-06, + "loss": 0.78975397, + "num_input_tokens_seen": 27085840, + "step": 1279, + "time_per_iteration": 2.890280246734619 + }, + { + "auxiliary_loss_clip": 0.01245048, + "auxiliary_loss_mlp": 0.01036479, + "balance_loss_clip": 1.05443239, + "balance_loss_mlp": 1.02605391, + "epoch": 0.15391090001803642, + "flos": 22054215553920.0, + "grad_norm": 2.847816157907329, + "language_loss": 0.78551781, + "learning_rate": 3.841209177334031e-06, + "loss": 0.80833316, + "num_input_tokens_seen": 27104200, + "step": 1280, + "time_per_iteration": 2.718545436859131 + }, + { + "auxiliary_loss_clip": 0.01235674, + "auxiliary_loss_mlp": 0.01042105, + "balance_loss_clip": 1.06203294, + "balance_loss_mlp": 1.03163242, + "epoch": 0.15403114290867553, + "flos": 15450351258240.0, + "grad_norm": 1.932507799351087, + "language_loss": 0.74802643, + "learning_rate": 3.84090485230921e-06, + "loss": 0.77080417, + "num_input_tokens_seen": 27122440, + "step": 1281, + "time_per_iteration": 2.6306874752044678 + }, + { + "auxiliary_loss_clip": 0.01227478, + "auxiliary_loss_mlp": 0.01042584, + "balance_loss_clip": 1.06360412, + "balance_loss_mlp": 1.03175402, + "epoch": 0.15415138579931462, + "flos": 17929156826880.0, + "grad_norm": 2.7549041062364767, + "language_loss": 0.76342571, + "learning_rate": 3.840600248020588e-06, + "loss": 0.78612638, + "num_input_tokens_seen": 27139380, + "step": 1282, + "time_per_iteration": 2.551873207092285 + }, + { + "auxiliary_loss_clip": 0.01249519, + "auxiliary_loss_mlp": 0.01041696, + "balance_loss_clip": 1.05794573, + "balance_loss_mlp": 1.03066945, + "epoch": 0.1542716286899537, + "flos": 11429325296640.0, + "grad_norm": 2.133568361092329, + "language_loss": 0.79717022, + "learning_rate": 3.840295364514371e-06, + "loss": 0.82008237, + "num_input_tokens_seen": 27156760, + "step": 1283, + "time_per_iteration": 2.738250494003296 + }, + { + "auxiliary_loss_clip": 0.0124777, + "auxiliary_loss_mlp": 0.01036518, + "balance_loss_clip": 1.06167793, + "balance_loss_mlp": 1.02571118, + "epoch": 0.1543918715805928, + "flos": 17420338719360.0, + "grad_norm": 4.780084482009469, + "language_loss": 0.78481114, + "learning_rate": 3.83999020183681e-06, + "loss": 0.80765396, + "num_input_tokens_seen": 27175455, + "step": 1284, + "time_per_iteration": 2.7063333988189697 + }, + { + "auxiliary_loss_clip": 0.01258324, + "auxiliary_loss_mlp": 0.01040039, + "balance_loss_clip": 1.0525341, + "balance_loss_mlp": 1.02833903, + "epoch": 0.1545121144712319, + "flos": 17786376264960.0, + "grad_norm": 2.9800590549791335, + "language_loss": 0.7838223, + "learning_rate": 3.839684760034199e-06, + "loss": 0.80680597, + "num_input_tokens_seen": 27193660, + "step": 1285, + "time_per_iteration": 2.720641851425171 + }, + { + "auxiliary_loss_clip": 0.01247702, + "auxiliary_loss_mlp": 0.01039237, + "balance_loss_clip": 1.06036377, + "balance_loss_mlp": 1.02849054, + "epoch": 0.15463235736187098, + "flos": 28220185146240.0, + "grad_norm": 9.197665743329773, + "language_loss": 0.6579113, + "learning_rate": 3.8393790391528716e-06, + "loss": 0.68078065, + "num_input_tokens_seen": 27214355, + "step": 1286, + "time_per_iteration": 2.743739604949951 + }, + { + "auxiliary_loss_clip": 0.01244885, + "auxiliary_loss_mlp": 0.01043936, + "balance_loss_clip": 1.05993652, + "balance_loss_mlp": 1.03352332, + "epoch": 0.15475260025251006, + "flos": 22856890826880.0, + "grad_norm": 2.78691139559293, + "language_loss": 0.89043766, + "learning_rate": 3.8390730392392075e-06, + "loss": 0.91332585, + "num_input_tokens_seen": 27234335, + "step": 1287, + "time_per_iteration": 2.7207603454589844 + }, + { + "auxiliary_loss_clip": 0.01233133, + "auxiliary_loss_mlp": 0.01034835, + "balance_loss_clip": 1.06529725, + "balance_loss_mlp": 1.02404618, + "epoch": 0.15487284314314917, + "flos": 17602872658560.0, + "grad_norm": 3.0384508443103995, + "language_loss": 0.79241008, + "learning_rate": 3.838766760339626e-06, + "loss": 0.8150897, + "num_input_tokens_seen": 27252860, + "step": 1288, + "time_per_iteration": 2.6072452068328857 + }, + { + "auxiliary_loss_clip": 0.01229911, + "auxiliary_loss_mlp": 0.01033293, + "balance_loss_clip": 1.05530763, + "balance_loss_mlp": 1.02211761, + "epoch": 0.15499308603378825, + "flos": 20082037363200.0, + "grad_norm": 2.488456067672113, + "language_loss": 0.79072392, + "learning_rate": 3.838460202500587e-06, + "loss": 0.81335592, + "num_input_tokens_seen": 27268650, + "step": 1289, + "time_per_iteration": 3.5715413093566895 + }, + { + "auxiliary_loss_clip": 0.01250532, + "auxiliary_loss_mlp": 0.01042881, + "balance_loss_clip": 1.06195867, + "balance_loss_mlp": 1.03062654, + "epoch": 0.15511332892442733, + "flos": 15918051271680.0, + "grad_norm": 4.618034285699303, + "language_loss": 0.74556309, + "learning_rate": 3.838153365768599e-06, + "loss": 0.76849723, + "num_input_tokens_seen": 27285160, + "step": 1290, + "time_per_iteration": 2.704063653945923 + }, + { + "auxiliary_loss_clip": 0.01250555, + "auxiliary_loss_mlp": 0.01042784, + "balance_loss_clip": 1.06259382, + "balance_loss_mlp": 1.03164411, + "epoch": 0.15523357181506645, + "flos": 41282475569280.0, + "grad_norm": 3.6758251724369093, + "language_loss": 0.75527477, + "learning_rate": 3.837846250190206e-06, + "loss": 0.77820814, + "num_input_tokens_seen": 27308025, + "step": 1291, + "time_per_iteration": 2.887840747833252 + }, + { + "auxiliary_loss_clip": 0.01259472, + "auxiliary_loss_mlp": 0.01381215, + "balance_loss_clip": 1.05924678, + "balance_loss_mlp": 1.00019598, + "epoch": 0.15535381470570553, + "flos": 18478769806080.0, + "grad_norm": 2.0818656975916694, + "language_loss": 0.7705484, + "learning_rate": 3.837538855811998e-06, + "loss": 0.79695529, + "num_input_tokens_seen": 27326200, + "step": 1292, + "time_per_iteration": 4.594832181930542 + }, + { + "auxiliary_loss_clip": 0.01255602, + "auxiliary_loss_mlp": 0.01046796, + "balance_loss_clip": 1.06570983, + "balance_loss_mlp": 1.03627014, + "epoch": 0.1554740575963446, + "flos": 13918150759680.0, + "grad_norm": 2.124602771175708, + "language_loss": 0.70775735, + "learning_rate": 3.837231182680606e-06, + "loss": 0.73078138, + "num_input_tokens_seen": 27344165, + "step": 1293, + "time_per_iteration": 2.685253381729126 + }, + { + "auxiliary_loss_clip": 0.01243878, + "auxiliary_loss_mlp": 0.01045845, + "balance_loss_clip": 1.06462061, + "balance_loss_mlp": 1.0349853, + "epoch": 0.1555943004869837, + "flos": 20847078161280.0, + "grad_norm": 1.570919168331054, + "language_loss": 0.75813711, + "learning_rate": 3.836923230842706e-06, + "loss": 0.78103429, + "num_input_tokens_seen": 27363280, + "step": 1294, + "time_per_iteration": 3.5656914710998535 + }, + { + "auxiliary_loss_clip": 0.01259175, + "auxiliary_loss_mlp": 0.0103877, + "balance_loss_clip": 1.05349076, + "balance_loss_mlp": 1.02839243, + "epoch": 0.1557145433776228, + "flos": 22085888371200.0, + "grad_norm": 2.258053327508254, + "language_loss": 0.80405319, + "learning_rate": 3.836615000345011e-06, + "loss": 0.82703257, + "num_input_tokens_seen": 27381460, + "step": 1295, + "time_per_iteration": 2.73681378364563 + }, + { + "auxiliary_loss_clip": 0.01231452, + "auxiliary_loss_mlp": 0.01036181, + "balance_loss_clip": 1.0659045, + "balance_loss_mlp": 1.0270133, + "epoch": 0.1558347862682619, + "flos": 19791987039360.0, + "grad_norm": 2.9249590761401936, + "language_loss": 0.78021955, + "learning_rate": 3.836306491234282e-06, + "loss": 0.8028959, + "num_input_tokens_seen": 27399310, + "step": 1296, + "time_per_iteration": 2.646810293197632 + }, + { + "auxiliary_loss_clip": 0.01244871, + "auxiliary_loss_mlp": 0.01045668, + "balance_loss_clip": 1.06544781, + "balance_loss_mlp": 1.03528523, + "epoch": 0.15595502915890097, + "flos": 17237086508160.0, + "grad_norm": 2.1903980326837584, + "language_loss": 0.7522999, + "learning_rate": 3.835997703557317e-06, + "loss": 0.77520531, + "num_input_tokens_seen": 27416050, + "step": 1297, + "time_per_iteration": 2.734849452972412 + }, + { + "auxiliary_loss_clip": 0.01255403, + "auxiliary_loss_mlp": 0.01043589, + "balance_loss_clip": 1.05187964, + "balance_loss_mlp": 1.03361106, + "epoch": 0.15607527204954008, + "flos": 19719519350400.0, + "grad_norm": 2.0455307832723464, + "language_loss": 0.7995913, + "learning_rate": 3.83568863736096e-06, + "loss": 0.82258123, + "num_input_tokens_seen": 27434920, + "step": 1298, + "time_per_iteration": 2.882305860519409 + }, + { + "auxiliary_loss_clip": 0.01252273, + "auxiliary_loss_mlp": 0.01034225, + "balance_loss_clip": 1.05722904, + "balance_loss_mlp": 1.02349043, + "epoch": 0.15619551494017916, + "flos": 18515650095360.0, + "grad_norm": 2.255885150276323, + "language_loss": 0.89145124, + "learning_rate": 3.8353792926920975e-06, + "loss": 0.91431618, + "num_input_tokens_seen": 27453570, + "step": 1299, + "time_per_iteration": 2.7132880687713623 + }, + { + "auxiliary_loss_clip": 0.0124175, + "auxiliary_loss_mlp": 0.01038438, + "balance_loss_clip": 1.06402004, + "balance_loss_mlp": 1.02778053, + "epoch": 0.15631575783081825, + "flos": 19902125116800.0, + "grad_norm": 2.401534286517635, + "language_loss": 0.8183322, + "learning_rate": 3.835069669597655e-06, + "loss": 0.84113407, + "num_input_tokens_seen": 27471960, + "step": 1300, + "time_per_iteration": 2.7121469974517822 + }, + { + "auxiliary_loss_clip": 0.01242264, + "auxiliary_loss_mlp": 0.01381783, + "balance_loss_clip": 1.06307507, + "balance_loss_mlp": 1.00017118, + "epoch": 0.15643600072145733, + "flos": 20777663128320.0, + "grad_norm": 2.002572806599959, + "language_loss": 0.799178, + "learning_rate": 3.834759768124603e-06, + "loss": 0.82541847, + "num_input_tokens_seen": 27490835, + "step": 1301, + "time_per_iteration": 2.6378159523010254 + }, + { + "auxiliary_loss_clip": 0.0126176, + "auxiliary_loss_mlp": 0.01045331, + "balance_loss_clip": 1.06543446, + "balance_loss_mlp": 1.03502548, + "epoch": 0.15655624361209644, + "flos": 18546389159040.0, + "grad_norm": 2.1147798936606534, + "language_loss": 0.76083696, + "learning_rate": 3.834449588319953e-06, + "loss": 0.78390789, + "num_input_tokens_seen": 27508870, + "step": 1302, + "time_per_iteration": 2.8203229904174805 + }, + { + "auxiliary_loss_clip": 0.01239753, + "auxiliary_loss_mlp": 0.01032751, + "balance_loss_clip": 1.0695405, + "balance_loss_mlp": 1.02311277, + "epoch": 0.15667648650273552, + "flos": 25229544727680.0, + "grad_norm": 2.2484507156775275, + "language_loss": 0.85293055, + "learning_rate": 3.834139130230758e-06, + "loss": 0.87565559, + "num_input_tokens_seen": 27528175, + "step": 1303, + "time_per_iteration": 2.687504529953003 + }, + { + "auxiliary_loss_clip": 0.01252337, + "auxiliary_loss_mlp": 0.01035021, + "balance_loss_clip": 1.06361651, + "balance_loss_mlp": 1.02419734, + "epoch": 0.1567967293933746, + "flos": 24827093769600.0, + "grad_norm": 2.515852221190651, + "language_loss": 0.81208915, + "learning_rate": 3.833828393904117e-06, + "loss": 0.83496273, + "num_input_tokens_seen": 27548455, + "step": 1304, + "time_per_iteration": 3.1224112510681152 + }, + { + "auxiliary_loss_clip": 0.01254274, + "auxiliary_loss_mlp": 0.01034463, + "balance_loss_clip": 1.05584085, + "balance_loss_mlp": 1.02456307, + "epoch": 0.15691697228401372, + "flos": 19164555244800.0, + "grad_norm": 2.2126451904924243, + "language_loss": 0.77308941, + "learning_rate": 3.833517379387165e-06, + "loss": 0.79597682, + "num_input_tokens_seen": 27564910, + "step": 1305, + "time_per_iteration": 2.8378748893737793 + }, + { + "auxiliary_loss_clip": 0.01243601, + "auxiliary_loss_mlp": 0.01035081, + "balance_loss_clip": 1.06547022, + "balance_loss_mlp": 1.02407765, + "epoch": 0.1570372151746528, + "flos": 24790931752320.0, + "grad_norm": 1.9399367618654613, + "language_loss": 0.88825357, + "learning_rate": 3.833206086727085e-06, + "loss": 0.91104043, + "num_input_tokens_seen": 27584260, + "step": 1306, + "time_per_iteration": 2.6569836139678955 + }, + { + "auxiliary_loss_clip": 0.01256687, + "auxiliary_loss_mlp": 0.01038927, + "balance_loss_clip": 1.05780613, + "balance_loss_mlp": 1.02826405, + "epoch": 0.15715745806529188, + "flos": 24863650836480.0, + "grad_norm": 2.098250730034012, + "language_loss": 0.70449394, + "learning_rate": 3.8328945159710994e-06, + "loss": 0.72745007, + "num_input_tokens_seen": 27604440, + "step": 1307, + "time_per_iteration": 2.9651341438293457 + }, + { + "auxiliary_loss_clip": 0.01245365, + "auxiliary_loss_mlp": 0.01381254, + "balance_loss_clip": 1.0673089, + "balance_loss_mlp": 1.00015759, + "epoch": 0.157277700955931, + "flos": 21872148491520.0, + "grad_norm": 1.9984569828009286, + "language_loss": 0.88722122, + "learning_rate": 3.832582667166473e-06, + "loss": 0.91348737, + "num_input_tokens_seen": 27624250, + "step": 1308, + "time_per_iteration": 2.6506590843200684 + }, + { + "auxiliary_loss_clip": 0.0124253, + "auxiliary_loss_mlp": 0.01034279, + "balance_loss_clip": 1.05923486, + "balance_loss_mlp": 1.02327597, + "epoch": 0.15739794384657008, + "flos": 24533344344960.0, + "grad_norm": 1.717455376579045, + "language_loss": 0.81293082, + "learning_rate": 3.8322705403605125e-06, + "loss": 0.83569896, + "num_input_tokens_seen": 27644595, + "step": 1309, + "time_per_iteration": 2.697934150695801 + }, + { + "auxiliary_loss_clip": 0.01239376, + "auxiliary_loss_mlp": 0.01031395, + "balance_loss_clip": 1.06080317, + "balance_loss_mlp": 1.0213877, + "epoch": 0.15751818673720916, + "flos": 17745329998080.0, + "grad_norm": 2.4281077438879284, + "language_loss": 0.80729616, + "learning_rate": 3.831958135600568e-06, + "loss": 0.83000386, + "num_input_tokens_seen": 27662145, + "step": 1310, + "time_per_iteration": 2.659219980239868 + }, + { + "auxiliary_loss_clip": 0.01241652, + "auxiliary_loss_mlp": 0.01035178, + "balance_loss_clip": 1.06567693, + "balance_loss_mlp": 1.02446699, + "epoch": 0.15763842962784824, + "flos": 17858520731520.0, + "grad_norm": 2.1988354306116142, + "language_loss": 0.79250133, + "learning_rate": 3.831645452934032e-06, + "loss": 0.81526965, + "num_input_tokens_seen": 27680575, + "step": 1311, + "time_per_iteration": 2.611131191253662 + }, + { + "auxiliary_loss_clip": 0.01234103, + "auxiliary_loss_mlp": 0.01045664, + "balance_loss_clip": 1.06889129, + "balance_loss_mlp": 1.03467333, + "epoch": 0.15775867251848735, + "flos": 26980908059520.0, + "grad_norm": 1.8299152806330041, + "language_loss": 0.80032039, + "learning_rate": 3.831332492408336e-06, + "loss": 0.82311803, + "num_input_tokens_seen": 27701985, + "step": 1312, + "time_per_iteration": 2.649151086807251 + }, + { + "auxiliary_loss_clip": 0.01239497, + "auxiliary_loss_mlp": 0.01040054, + "balance_loss_clip": 1.05787909, + "balance_loss_mlp": 1.03015339, + "epoch": 0.15787891540912644, + "flos": 19240398812160.0, + "grad_norm": 1.9925419426924253, + "language_loss": 0.69510734, + "learning_rate": 3.831019254070957e-06, + "loss": 0.71790284, + "num_input_tokens_seen": 27719770, + "step": 1313, + "time_per_iteration": 2.701174736022949 + }, + { + "auxiliary_loss_clip": 0.01261682, + "auxiliary_loss_mlp": 0.01035471, + "balance_loss_clip": 1.05581474, + "balance_loss_mlp": 1.0256536, + "epoch": 0.15799915829976552, + "flos": 27271102037760.0, + "grad_norm": 2.5083628614555686, + "language_loss": 0.94843268, + "learning_rate": 3.8307057379694135e-06, + "loss": 0.97140419, + "num_input_tokens_seen": 27739105, + "step": 1314, + "time_per_iteration": 2.7995269298553467 + }, + { + "auxiliary_loss_clip": 0.0123303, + "auxiliary_loss_mlp": 0.01039534, + "balance_loss_clip": 1.06582546, + "balance_loss_mlp": 1.02848315, + "epoch": 0.15811940119040463, + "flos": 20405520270720.0, + "grad_norm": 2.3202215134752437, + "language_loss": 0.8232286, + "learning_rate": 3.830391944151264e-06, + "loss": 0.84595418, + "num_input_tokens_seen": 27754985, + "step": 1315, + "time_per_iteration": 3.5286524295806885 + }, + { + "auxiliary_loss_clip": 0.01246026, + "auxiliary_loss_mlp": 0.01042826, + "balance_loss_clip": 1.05939269, + "balance_loss_mlp": 1.03104246, + "epoch": 0.1582396440810437, + "flos": 32599347661440.0, + "grad_norm": 1.8667690488495923, + "language_loss": 0.67288136, + "learning_rate": 3.830077872664114e-06, + "loss": 0.69576985, + "num_input_tokens_seen": 27776110, + "step": 1316, + "time_per_iteration": 2.7980990409851074 + }, + { + "auxiliary_loss_clip": 0.01268382, + "auxiliary_loss_mlp": 0.01040822, + "balance_loss_clip": 1.05293441, + "balance_loss_mlp": 1.03027773, + "epoch": 0.1583598869716828, + "flos": 33800559310080.0, + "grad_norm": 1.6903740467966388, + "language_loss": 0.73291278, + "learning_rate": 3.829763523555604e-06, + "loss": 0.75600481, + "num_input_tokens_seen": 27796510, + "step": 1317, + "time_per_iteration": 2.852170705795288 + }, + { + "auxiliary_loss_clip": 0.01232282, + "auxiliary_loss_mlp": 0.01038499, + "balance_loss_clip": 1.06514573, + "balance_loss_mlp": 1.02903962, + "epoch": 0.15848012986232188, + "flos": 24681332378880.0, + "grad_norm": 1.9876387310922157, + "language_loss": 0.77739894, + "learning_rate": 3.829448896873423e-06, + "loss": 0.80010676, + "num_input_tokens_seen": 27815610, + "step": 1318, + "time_per_iteration": 4.532312631607056 + }, + { + "auxiliary_loss_clip": 0.01250584, + "auxiliary_loss_mlp": 0.01381354, + "balance_loss_clip": 1.05972755, + "balance_loss_mlp": 1.00020409, + "epoch": 0.158600372752961, + "flos": 22602068766720.0, + "grad_norm": 1.9209524336375543, + "language_loss": 0.79057097, + "learning_rate": 3.829133992665299e-06, + "loss": 0.8168903, + "num_input_tokens_seen": 27834735, + "step": 1319, + "time_per_iteration": 2.7511045932769775 + }, + { + "auxiliary_loss_clip": 0.01225767, + "auxiliary_loss_mlp": 0.01045624, + "balance_loss_clip": 1.06432962, + "balance_loss_mlp": 1.03528309, + "epoch": 0.15872061564360007, + "flos": 27927944092800.0, + "grad_norm": 2.9130578338209174, + "language_loss": 0.89032978, + "learning_rate": 3.828818810979002e-06, + "loss": 0.91304374, + "num_input_tokens_seen": 27853065, + "step": 1320, + "time_per_iteration": 2.68623948097229 + }, + { + "auxiliary_loss_clip": 0.01230321, + "auxiliary_loss_mlp": 0.01040459, + "balance_loss_clip": 1.06754398, + "balance_loss_mlp": 1.0297184, + "epoch": 0.15884085853423915, + "flos": 23696805525120.0, + "grad_norm": 2.0742558556616624, + "language_loss": 0.80290979, + "learning_rate": 3.8285033518623454e-06, + "loss": 0.82561755, + "num_input_tokens_seen": 27873315, + "step": 1321, + "time_per_iteration": 3.4874825477600098 + }, + { + "auxiliary_loss_clip": 0.0124594, + "auxiliary_loss_mlp": 0.0104093, + "balance_loss_clip": 1.06653237, + "balance_loss_mlp": 1.02982533, + "epoch": 0.15896110142487826, + "flos": 23112359331840.0, + "grad_norm": 2.902030956950953, + "language_loss": 0.81180978, + "learning_rate": 3.8281876153631845e-06, + "loss": 0.83467847, + "num_input_tokens_seen": 27890070, + "step": 1322, + "time_per_iteration": 2.6395251750946045 + }, + { + "auxiliary_loss_clip": 0.01255698, + "auxiliary_loss_mlp": 0.01038496, + "balance_loss_clip": 1.05696607, + "balance_loss_mlp": 1.02805305, + "epoch": 0.15908134431551735, + "flos": 14685238632960.0, + "grad_norm": 2.137724305940365, + "language_loss": 0.64736384, + "learning_rate": 3.827871601529416e-06, + "loss": 0.67030579, + "num_input_tokens_seen": 27908590, + "step": 1323, + "time_per_iteration": 2.698474168777466 + }, + { + "auxiliary_loss_clip": 0.01246514, + "auxiliary_loss_mlp": 0.01039579, + "balance_loss_clip": 1.05963993, + "balance_loss_mlp": 1.028898, + "epoch": 0.15920158720615643, + "flos": 20193611984640.0, + "grad_norm": 2.9369129153545916, + "language_loss": 0.80447108, + "learning_rate": 3.827555310408979e-06, + "loss": 0.82733202, + "num_input_tokens_seen": 27927985, + "step": 1324, + "time_per_iteration": 2.6824188232421875 + }, + { + "auxiliary_loss_clip": 0.01249699, + "auxiliary_loss_mlp": 0.0103461, + "balance_loss_clip": 1.06010997, + "balance_loss_mlp": 1.0236609, + "epoch": 0.1593218300967955, + "flos": 24826626892800.0, + "grad_norm": 1.7996911485942293, + "language_loss": 0.82766962, + "learning_rate": 3.827238742049854e-06, + "loss": 0.85051268, + "num_input_tokens_seen": 27948280, + "step": 1325, + "time_per_iteration": 2.7027180194854736 + }, + { + "auxiliary_loss_clip": 0.0123101, + "auxiliary_loss_mlp": 0.01045372, + "balance_loss_clip": 1.06457996, + "balance_loss_mlp": 1.03575134, + "epoch": 0.15944207298743462, + "flos": 28328707111680.0, + "grad_norm": 2.438271296612464, + "language_loss": 0.51718658, + "learning_rate": 3.826921896500066e-06, + "loss": 0.53995043, + "num_input_tokens_seen": 27969565, + "step": 1326, + "time_per_iteration": 2.726233959197998 + }, + { + "auxiliary_loss_clip": 0.01258802, + "auxiliary_loss_mlp": 0.01040338, + "balance_loss_clip": 1.0630579, + "balance_loss_mlp": 1.02881026, + "epoch": 0.1595623158780737, + "flos": 22964838174720.0, + "grad_norm": 1.8592740522871338, + "language_loss": 0.77975768, + "learning_rate": 3.826604773807678e-06, + "loss": 0.80274904, + "num_input_tokens_seen": 27987540, + "step": 1327, + "time_per_iteration": 2.6996402740478516 + }, + { + "auxiliary_loss_clip": 0.0124384, + "auxiliary_loss_mlp": 0.01041996, + "balance_loss_clip": 1.05883217, + "balance_loss_mlp": 1.03079581, + "epoch": 0.1596825587687128, + "flos": 19710540950400.0, + "grad_norm": 2.6424004017205953, + "language_loss": 0.73470277, + "learning_rate": 3.826287374020798e-06, + "loss": 0.75756109, + "num_input_tokens_seen": 28002345, + "step": 1328, + "time_per_iteration": 2.7157955169677734 + }, + { + "auxiliary_loss_clip": 0.01234512, + "auxiliary_loss_mlp": 0.01048581, + "balance_loss_clip": 1.06984138, + "balance_loss_mlp": 1.03838289, + "epoch": 0.1598028016593519, + "flos": 22637727993600.0, + "grad_norm": 3.242070171249524, + "language_loss": 0.82271892, + "learning_rate": 3.825969697187575e-06, + "loss": 0.84554988, + "num_input_tokens_seen": 28021675, + "step": 1329, + "time_per_iteration": 2.630504608154297 + }, + { + "auxiliary_loss_clip": 0.01250742, + "auxiliary_loss_mlp": 0.01039987, + "balance_loss_clip": 1.05869865, + "balance_loss_mlp": 1.03014612, + "epoch": 0.15992304454999098, + "flos": 20482908122880.0, + "grad_norm": 1.8717158031386876, + "language_loss": 0.69515848, + "learning_rate": 3.8256517433562015e-06, + "loss": 0.71806574, + "num_input_tokens_seen": 28039615, + "step": 1330, + "time_per_iteration": 2.717594861984253 + }, + { + "auxiliary_loss_clip": 0.01231824, + "auxiliary_loss_mlp": 0.01040983, + "balance_loss_clip": 1.06804347, + "balance_loss_mlp": 1.03167903, + "epoch": 0.16004328744063007, + "flos": 17676094533120.0, + "grad_norm": 2.2185673171526115, + "language_loss": 0.92118108, + "learning_rate": 3.82533351257491e-06, + "loss": 0.94390917, + "num_input_tokens_seen": 28057565, + "step": 1331, + "time_per_iteration": 2.6362287998199463 + }, + { + "auxiliary_loss_clip": 0.01235532, + "auxiliary_loss_mlp": 0.01046993, + "balance_loss_clip": 1.06432915, + "balance_loss_mlp": 1.03696752, + "epoch": 0.16016353033126918, + "flos": 24098717779200.0, + "grad_norm": 1.8368071094986653, + "language_loss": 0.88904124, + "learning_rate": 3.825015004891975e-06, + "loss": 0.91186655, + "num_input_tokens_seen": 28076305, + "step": 1332, + "time_per_iteration": 2.6822221279144287 + }, + { + "auxiliary_loss_clip": 0.01229968, + "auxiliary_loss_mlp": 0.01038673, + "balance_loss_clip": 1.06139278, + "balance_loss_mlp": 1.02902317, + "epoch": 0.16028377322190826, + "flos": 27634841112960.0, + "grad_norm": 1.8787099040935622, + "language_loss": 0.75457525, + "learning_rate": 3.824696220355716e-06, + "loss": 0.77726173, + "num_input_tokens_seen": 28097895, + "step": 1333, + "time_per_iteration": 2.6813597679138184 + }, + { + "auxiliary_loss_clip": 0.01240467, + "auxiliary_loss_mlp": 0.01038942, + "balance_loss_clip": 1.06040466, + "balance_loss_mlp": 1.02859449, + "epoch": 0.16040401611254734, + "flos": 20961202648320.0, + "grad_norm": 2.030275590254589, + "language_loss": 0.78853166, + "learning_rate": 3.824377159014491e-06, + "loss": 0.81132567, + "num_input_tokens_seen": 28118790, + "step": 1334, + "time_per_iteration": 2.694903612136841 + }, + { + "auxiliary_loss_clip": 0.01234523, + "auxiliary_loss_mlp": 0.01036341, + "balance_loss_clip": 1.06431341, + "balance_loss_mlp": 1.02658415, + "epoch": 0.16052425900318643, + "flos": 21247051080960.0, + "grad_norm": 1.7388937514352387, + "language_loss": 0.85000968, + "learning_rate": 3.824057820916702e-06, + "loss": 0.87271833, + "num_input_tokens_seen": 28135995, + "step": 1335, + "time_per_iteration": 2.664322853088379 + }, + { + "auxiliary_loss_clip": 0.01249406, + "auxiliary_loss_mlp": 0.01035183, + "balance_loss_clip": 1.06350255, + "balance_loss_mlp": 1.0237087, + "epoch": 0.16064450189382554, + "flos": 15524004096000.0, + "grad_norm": 2.334830603462103, + "language_loss": 0.71773982, + "learning_rate": 3.8237382061107904e-06, + "loss": 0.7405858, + "num_input_tokens_seen": 28152715, + "step": 1336, + "time_per_iteration": 2.617060899734497 + }, + { + "auxiliary_loss_clip": 0.01261574, + "auxiliary_loss_mlp": 0.01035252, + "balance_loss_clip": 1.04947639, + "balance_loss_mlp": 1.02469563, + "epoch": 0.16076474478446462, + "flos": 21178497974400.0, + "grad_norm": 1.6258616506099948, + "language_loss": 0.78371674, + "learning_rate": 3.823418314645243e-06, + "loss": 0.80668497, + "num_input_tokens_seen": 28171590, + "step": 1337, + "time_per_iteration": 2.896885395050049 + }, + { + "auxiliary_loss_clip": 0.01239285, + "auxiliary_loss_mlp": 0.01043416, + "balance_loss_clip": 1.05597293, + "balance_loss_mlp": 1.03406394, + "epoch": 0.1608849876751037, + "flos": 18366476912640.0, + "grad_norm": 2.071333158984587, + "language_loss": 0.75194812, + "learning_rate": 3.823098146568588e-06, + "loss": 0.77477509, + "num_input_tokens_seen": 28191295, + "step": 1338, + "time_per_iteration": 3.1231727600097656 + }, + { + "auxiliary_loss_clip": 0.01239252, + "auxiliary_loss_mlp": 0.01033167, + "balance_loss_clip": 1.0641346, + "balance_loss_mlp": 1.02326047, + "epoch": 0.1610052305657428, + "flos": 29497024880640.0, + "grad_norm": 1.8510263287302013, + "language_loss": 0.71433151, + "learning_rate": 3.822777701929394e-06, + "loss": 0.73705572, + "num_input_tokens_seen": 28213120, + "step": 1339, + "time_per_iteration": 2.7181503772735596 + }, + { + "auxiliary_loss_clip": 0.01219502, + "auxiliary_loss_mlp": 0.01036993, + "balance_loss_clip": 1.0579114, + "balance_loss_mlp": 1.0269022, + "epoch": 0.1611254734563819, + "flos": 26797871329920.0, + "grad_norm": 1.8542866081186375, + "language_loss": 0.73421204, + "learning_rate": 3.8224569807762714e-06, + "loss": 0.75677705, + "num_input_tokens_seen": 28232440, + "step": 1340, + "time_per_iteration": 2.679706335067749 + }, + { + "auxiliary_loss_clip": 0.01234693, + "auxiliary_loss_mlp": 0.01038501, + "balance_loss_clip": 1.05056596, + "balance_loss_mlp": 1.02758753, + "epoch": 0.16124571634702098, + "flos": 22419570741120.0, + "grad_norm": 2.6136169322752028, + "language_loss": 0.76426649, + "learning_rate": 3.822135983157873e-06, + "loss": 0.78699845, + "num_input_tokens_seen": 28251715, + "step": 1341, + "time_per_iteration": 3.766507387161255 + }, + { + "auxiliary_loss_clip": 0.01224992, + "auxiliary_loss_mlp": 0.0138106, + "balance_loss_clip": 1.06376815, + "balance_loss_mlp": 1.00038242, + "epoch": 0.16136595923766006, + "flos": 10999116103680.0, + "grad_norm": 2.6787967441216436, + "language_loss": 0.84766936, + "learning_rate": 3.821814709122896e-06, + "loss": 0.87372983, + "num_input_tokens_seen": 28269765, + "step": 1342, + "time_per_iteration": 2.6662747859954834 + }, + { + "auxiliary_loss_clip": 0.01240266, + "auxiliary_loss_mlp": 0.01040532, + "balance_loss_clip": 1.06093383, + "balance_loss_mlp": 1.03141308, + "epoch": 0.16148620212829917, + "flos": 21214983214080.0, + "grad_norm": 2.5880417534468654, + "language_loss": 0.85270482, + "learning_rate": 3.821493158720076e-06, + "loss": 0.87551284, + "num_input_tokens_seen": 28288870, + "step": 1343, + "time_per_iteration": 2.6938891410827637 + }, + { + "auxiliary_loss_clip": 0.01248667, + "auxiliary_loss_mlp": 0.01036708, + "balance_loss_clip": 1.05585599, + "balance_loss_mlp": 1.0259676, + "epoch": 0.16160644501893826, + "flos": 16758468760320.0, + "grad_norm": 4.225246174349515, + "language_loss": 0.72592175, + "learning_rate": 3.821171331998191e-06, + "loss": 0.74877548, + "num_input_tokens_seen": 28305400, + "step": 1344, + "time_per_iteration": 4.404429912567139 + }, + { + "auxiliary_loss_clip": 0.01150883, + "auxiliary_loss_mlp": 0.01005168, + "balance_loss_clip": 1.02487469, + "balance_loss_mlp": 1.00240231, + "epoch": 0.16172668790957734, + "flos": 64444967308800.0, + "grad_norm": 0.7112199272267797, + "language_loss": 0.54469359, + "learning_rate": 3.820849229006064e-06, + "loss": 0.56625408, + "num_input_tokens_seen": 28373150, + "step": 1345, + "time_per_iteration": 3.4475138187408447 + }, + { + "auxiliary_loss_clip": 0.01227855, + "auxiliary_loss_mlp": 0.01030751, + "balance_loss_clip": 1.06465662, + "balance_loss_mlp": 1.02095819, + "epoch": 0.16184693080021645, + "flos": 23257689759360.0, + "grad_norm": 1.8529871390307713, + "language_loss": 0.70515072, + "learning_rate": 3.8205268497925564e-06, + "loss": 0.72773677, + "num_input_tokens_seen": 28393620, + "step": 1346, + "time_per_iteration": 3.5360147953033447 + }, + { + "auxiliary_loss_clip": 0.0122986, + "auxiliary_loss_mlp": 0.01036852, + "balance_loss_clip": 1.06596744, + "balance_loss_mlp": 1.02646923, + "epoch": 0.16196717369085553, + "flos": 17451113696640.0, + "grad_norm": 2.1869039027499575, + "language_loss": 0.78476161, + "learning_rate": 3.8202041944065725e-06, + "loss": 0.80742872, + "num_input_tokens_seen": 28409440, + "step": 1347, + "time_per_iteration": 2.5713515281677246 + }, + { + "auxiliary_loss_clip": 0.01232565, + "auxiliary_loss_mlp": 0.01038046, + "balance_loss_clip": 1.06906855, + "balance_loss_mlp": 1.02793145, + "epoch": 0.16208741658149461, + "flos": 23873377806720.0, + "grad_norm": 1.808218667579124, + "language_loss": 0.73797697, + "learning_rate": 3.819881262897061e-06, + "loss": 0.76068312, + "num_input_tokens_seen": 28427575, + "step": 1348, + "time_per_iteration": 2.6844935417175293 + }, + { + "auxiliary_loss_clip": 0.01258207, + "auxiliary_loss_mlp": 0.0103758, + "balance_loss_clip": 1.06487274, + "balance_loss_mlp": 1.02709556, + "epoch": 0.1622076594721337, + "flos": 25884806584320.0, + "grad_norm": 2.149940655594675, + "language_loss": 0.73531497, + "learning_rate": 3.819558055313008e-06, + "loss": 0.75827289, + "num_input_tokens_seen": 28448260, + "step": 1349, + "time_per_iteration": 2.738335609436035 + }, + { + "auxiliary_loss_clip": 0.01239428, + "auxiliary_loss_mlp": 0.01038724, + "balance_loss_clip": 1.0628953, + "balance_loss_mlp": 1.02868116, + "epoch": 0.1623279023627728, + "flos": 21539759011200.0, + "grad_norm": 1.878517122682717, + "language_loss": 0.7742697, + "learning_rate": 3.819234571703444e-06, + "loss": 0.79705131, + "num_input_tokens_seen": 28467085, + "step": 1350, + "time_per_iteration": 2.670393228530884 + }, + { + "auxiliary_loss_clip": 0.01226823, + "auxiliary_loss_mlp": 0.01041212, + "balance_loss_clip": 1.0595324, + "balance_loss_mlp": 1.03091288, + "epoch": 0.1624481452534119, + "flos": 22085421494400.0, + "grad_norm": 2.616515389066127, + "language_loss": 0.85380989, + "learning_rate": 3.8189108121174435e-06, + "loss": 0.87649029, + "num_input_tokens_seen": 28486850, + "step": 1351, + "time_per_iteration": 2.613978147506714 + }, + { + "auxiliary_loss_clip": 0.01245134, + "auxiliary_loss_mlp": 0.0103423, + "balance_loss_clip": 1.05986249, + "balance_loss_mlp": 1.02441943, + "epoch": 0.16256838814405097, + "flos": 27087490690560.0, + "grad_norm": 1.6419979846561774, + "language_loss": 0.83677566, + "learning_rate": 3.818586776604118e-06, + "loss": 0.85956925, + "num_input_tokens_seen": 28507490, + "step": 1352, + "time_per_iteration": 2.7172017097473145 + }, + { + "auxiliary_loss_clip": 0.01236651, + "auxiliary_loss_mlp": 0.01043495, + "balance_loss_clip": 1.06036425, + "balance_loss_mlp": 1.03317738, + "epoch": 0.16268863103469008, + "flos": 20120354196480.0, + "grad_norm": 1.9019097180596487, + "language_loss": 0.61675525, + "learning_rate": 3.818262465212625e-06, + "loss": 0.63955671, + "num_input_tokens_seen": 28527615, + "step": 1353, + "time_per_iteration": 2.630565881729126 + }, + { + "auxiliary_loss_clip": 0.01223628, + "auxiliary_loss_mlp": 0.01048916, + "balance_loss_clip": 1.06257391, + "balance_loss_mlp": 1.03781152, + "epoch": 0.16280887392532917, + "flos": 18332792933760.0, + "grad_norm": 3.0965176185594054, + "language_loss": 0.77303344, + "learning_rate": 3.817937877992161e-06, + "loss": 0.7957589, + "num_input_tokens_seen": 28544910, + "step": 1354, + "time_per_iteration": 2.6227293014526367 + }, + { + "auxiliary_loss_clip": 0.01243144, + "auxiliary_loss_mlp": 0.01381783, + "balance_loss_clip": 1.05364704, + "balance_loss_mlp": 1.0004077, + "epoch": 0.16292911681596825, + "flos": 11874330892800.0, + "grad_norm": 2.5247649945432804, + "language_loss": 0.85416031, + "learning_rate": 3.817613014991967e-06, + "loss": 0.8804096, + "num_input_tokens_seen": 28561050, + "step": 1355, + "time_per_iteration": 2.6590282917022705 + }, + { + "auxiliary_loss_clip": 0.01234845, + "auxiliary_loss_mlp": 0.01036881, + "balance_loss_clip": 1.05470181, + "balance_loss_mlp": 1.02599716, + "epoch": 0.16304935970660733, + "flos": 26103466627200.0, + "grad_norm": 2.4493154063975533, + "language_loss": 0.7679041, + "learning_rate": 3.817287876261323e-06, + "loss": 0.7906214, + "num_input_tokens_seen": 28581385, + "step": 1356, + "time_per_iteration": 2.7896764278411865 + }, + { + "auxiliary_loss_clip": 0.0123964, + "auxiliary_loss_mlp": 0.0103485, + "balance_loss_clip": 1.06141448, + "balance_loss_mlp": 1.02447295, + "epoch": 0.16316960259724644, + "flos": 29351945848320.0, + "grad_norm": 1.7814138710743166, + "language_loss": 0.79957569, + "learning_rate": 3.816962461849553e-06, + "loss": 0.82232052, + "num_input_tokens_seen": 28603255, + "step": 1357, + "time_per_iteration": 2.7269480228424072 + }, + { + "auxiliary_loss_clip": 0.01234965, + "auxiliary_loss_mlp": 0.01045286, + "balance_loss_clip": 1.06032765, + "balance_loss_mlp": 1.03483105, + "epoch": 0.16328984548788553, + "flos": 20886759711360.0, + "grad_norm": 2.1222289723941876, + "language_loss": 0.84754097, + "learning_rate": 3.8166367718060235e-06, + "loss": 0.87034351, + "num_input_tokens_seen": 28623145, + "step": 1358, + "time_per_iteration": 2.723484516143799 + }, + { + "auxiliary_loss_clip": 0.01226846, + "auxiliary_loss_mlp": 0.01029555, + "balance_loss_clip": 1.05861306, + "balance_loss_mlp": 1.01887345, + "epoch": 0.1634100883785246, + "flos": 18041090584320.0, + "grad_norm": 3.0055589432642726, + "language_loss": 0.76484573, + "learning_rate": 3.816310806180139e-06, + "loss": 0.78740978, + "num_input_tokens_seen": 28641555, + "step": 1359, + "time_per_iteration": 2.642019748687744 + }, + { + "auxiliary_loss_clip": 0.01238649, + "auxiliary_loss_mlp": 0.01039484, + "balance_loss_clip": 1.06063437, + "balance_loss_mlp": 1.02881455, + "epoch": 0.16353033126916372, + "flos": 24572128055040.0, + "grad_norm": 1.8000168496360294, + "language_loss": 0.80956745, + "learning_rate": 3.81598456502135e-06, + "loss": 0.83234876, + "num_input_tokens_seen": 28661575, + "step": 1360, + "time_per_iteration": 2.8092100620269775 + }, + { + "auxiliary_loss_clip": 0.0123886, + "auxiliary_loss_mlp": 0.01042405, + "balance_loss_clip": 1.06235158, + "balance_loss_mlp": 1.03251648, + "epoch": 0.1636505741598028, + "flos": 19892895321600.0, + "grad_norm": 2.270722116282552, + "language_loss": 0.86764508, + "learning_rate": 3.8156580483791455e-06, + "loss": 0.89045775, + "num_input_tokens_seen": 28676765, + "step": 1361, + "time_per_iteration": 2.742502450942993 + }, + { + "auxiliary_loss_clip": 0.01227045, + "auxiliary_loss_mlp": 0.01036616, + "balance_loss_clip": 1.06434941, + "balance_loss_mlp": 1.02670979, + "epoch": 0.16377081705044189, + "flos": 28402611344640.0, + "grad_norm": 2.2534566212477456, + "language_loss": 0.76469409, + "learning_rate": 3.815331256303059e-06, + "loss": 0.78733069, + "num_input_tokens_seen": 28696795, + "step": 1362, + "time_per_iteration": 2.634323835372925 + }, + { + "auxiliary_loss_clip": 0.01249365, + "auxiliary_loss_mlp": 0.01036718, + "balance_loss_clip": 1.06332469, + "balance_loss_mlp": 1.02637672, + "epoch": 0.163891059941081, + "flos": 21908059113600.0, + "grad_norm": 2.4329937830084196, + "language_loss": 0.77041173, + "learning_rate": 3.815004188842665e-06, + "loss": 0.79327261, + "num_input_tokens_seen": 28714835, + "step": 1363, + "time_per_iteration": 2.749081611633301 + }, + { + "auxiliary_loss_clip": 0.01235296, + "auxiliary_loss_mlp": 0.01039325, + "balance_loss_clip": 1.05835092, + "balance_loss_mlp": 1.02882886, + "epoch": 0.16401130283172008, + "flos": 26797619934720.0, + "grad_norm": 1.6010675128995995, + "language_loss": 0.79452777, + "learning_rate": 3.814676846047578e-06, + "loss": 0.81727397, + "num_input_tokens_seen": 28735710, + "step": 1364, + "time_per_iteration": 2.713397741317749 + }, + { + "auxiliary_loss_clip": 0.0122811, + "auxiliary_loss_mlp": 0.01047937, + "balance_loss_clip": 1.05986512, + "balance_loss_mlp": 1.03736949, + "epoch": 0.16413154572235916, + "flos": 32997417160320.0, + "grad_norm": 1.835246106978649, + "language_loss": 0.70138729, + "learning_rate": 3.8143492279674565e-06, + "loss": 0.7241478, + "num_input_tokens_seen": 28758405, + "step": 1365, + "time_per_iteration": 2.764618396759033 + }, + { + "auxiliary_loss_clip": 0.01161662, + "auxiliary_loss_mlp": 0.01002606, + "balance_loss_clip": 1.03486562, + "balance_loss_mlp": 0.99990004, + "epoch": 0.16425178861299825, + "flos": 40113622074240.0, + "grad_norm": 0.8423844091155088, + "language_loss": 0.58360219, + "learning_rate": 3.8140213346519997e-06, + "loss": 0.60524487, + "num_input_tokens_seen": 28809000, + "step": 1366, + "time_per_iteration": 3.036161422729492 + }, + { + "auxiliary_loss_clip": 0.0123549, + "auxiliary_loss_mlp": 0.01038064, + "balance_loss_clip": 1.05709112, + "balance_loss_mlp": 1.02787733, + "epoch": 0.16437203150363736, + "flos": 25447486498560.0, + "grad_norm": 3.19863420452937, + "language_loss": 0.76781416, + "learning_rate": 3.813693166150948e-06, + "loss": 0.79054964, + "num_input_tokens_seen": 28829210, + "step": 1367, + "time_per_iteration": 3.6489064693450928 + }, + { + "auxiliary_loss_clip": 0.01242061, + "auxiliary_loss_mlp": 0.01035358, + "balance_loss_clip": 1.05938649, + "balance_loss_mlp": 1.02520704, + "epoch": 0.16449227439427644, + "flos": 23476888506240.0, + "grad_norm": 4.335531059918504, + "language_loss": 0.85573399, + "learning_rate": 3.813364722514086e-06, + "loss": 0.87850821, + "num_input_tokens_seen": 28847545, + "step": 1368, + "time_per_iteration": 2.7530477046966553 + }, + { + "auxiliary_loss_clip": 0.01232086, + "auxiliary_loss_mlp": 0.0104002, + "balance_loss_clip": 1.06243873, + "balance_loss_mlp": 1.02956581, + "epoch": 0.16461251728491552, + "flos": 13545217802880.0, + "grad_norm": 2.185527124253009, + "language_loss": 0.79985613, + "learning_rate": 3.8130360037912368e-06, + "loss": 0.82257724, + "num_input_tokens_seen": 28863990, + "step": 1369, + "time_per_iteration": 2.6439318656921387 + }, + { + "auxiliary_loss_clip": 0.0123477, + "auxiliary_loss_mlp": 0.0103112, + "balance_loss_clip": 1.06417406, + "balance_loss_mlp": 1.02052855, + "epoch": 0.16473276017555463, + "flos": 23003298662400.0, + "grad_norm": 2.2668122881826984, + "language_loss": 0.81858325, + "learning_rate": 3.812707010032268e-06, + "loss": 0.84124219, + "num_input_tokens_seen": 28883045, + "step": 1370, + "time_per_iteration": 4.481522798538208 + }, + { + "auxiliary_loss_clip": 0.01244422, + "auxiliary_loss_mlp": 0.0104138, + "balance_loss_clip": 1.06920505, + "balance_loss_mlp": 1.03036547, + "epoch": 0.16485300306619372, + "flos": 24790680357120.0, + "grad_norm": 6.7913770351563505, + "language_loss": 0.79030669, + "learning_rate": 3.8123777412870863e-06, + "loss": 0.81316471, + "num_input_tokens_seen": 28902545, + "step": 1371, + "time_per_iteration": 2.7580316066741943 + }, + { + "auxiliary_loss_clip": 0.01248419, + "auxiliary_loss_mlp": 0.01039993, + "balance_loss_clip": 1.06262326, + "balance_loss_mlp": 1.02937758, + "epoch": 0.1649732459568328, + "flos": 21106497162240.0, + "grad_norm": 2.0300474057797286, + "language_loss": 0.78383338, + "learning_rate": 3.812048197605643e-06, + "loss": 0.80671746, + "num_input_tokens_seen": 28921440, + "step": 1372, + "time_per_iteration": 3.6056859493255615 + }, + { + "auxiliary_loss_clip": 0.01235188, + "auxiliary_loss_mlp": 0.01033082, + "balance_loss_clip": 1.06293654, + "balance_loss_mlp": 1.02290165, + "epoch": 0.16509348884747188, + "flos": 20266726118400.0, + "grad_norm": 1.9625923458053633, + "language_loss": 0.81426924, + "learning_rate": 3.8117183790379277e-06, + "loss": 0.83695197, + "num_input_tokens_seen": 28939890, + "step": 1373, + "time_per_iteration": 2.7993104457855225 + }, + { + "auxiliary_loss_clip": 0.01229333, + "auxiliary_loss_mlp": 0.01033922, + "balance_loss_clip": 1.06678224, + "balance_loss_mlp": 1.02352655, + "epoch": 0.165213731738111, + "flos": 11035493602560.0, + "grad_norm": 2.7573944668462502, + "language_loss": 0.94379354, + "learning_rate": 3.811388285633976e-06, + "loss": 0.96642607, + "num_input_tokens_seen": 28955875, + "step": 1374, + "time_per_iteration": 2.6309895515441895 + }, + { + "auxiliary_loss_clip": 0.0126219, + "auxiliary_loss_mlp": 0.0103167, + "balance_loss_clip": 1.05785012, + "balance_loss_mlp": 1.02148378, + "epoch": 0.16533397462875007, + "flos": 29972051268480.0, + "grad_norm": 2.0864949944859155, + "language_loss": 0.61890614, + "learning_rate": 3.811057917443861e-06, + "loss": 0.64184475, + "num_input_tokens_seen": 28975140, + "step": 1375, + "time_per_iteration": 2.786975860595703 + }, + { + "auxiliary_loss_clip": 0.01156114, + "auxiliary_loss_mlp": 0.01003535, + "balance_loss_clip": 1.03640318, + "balance_loss_mlp": 1.0008173, + "epoch": 0.16545421751938916, + "flos": 65556763027200.0, + "grad_norm": 0.8503155873980447, + "language_loss": 0.6829381, + "learning_rate": 3.8107272745177e-06, + "loss": 0.70453459, + "num_input_tokens_seen": 29047470, + "step": 1376, + "time_per_iteration": 3.3799681663513184 + }, + { + "auxiliary_loss_clip": 0.01253635, + "auxiliary_loss_mlp": 0.01036395, + "balance_loss_clip": 1.06206942, + "balance_loss_mlp": 1.02586854, + "epoch": 0.16557446041002827, + "flos": 22492361652480.0, + "grad_norm": 1.8079671186521453, + "language_loss": 0.78669536, + "learning_rate": 3.8103963569056513e-06, + "loss": 0.80959558, + "num_input_tokens_seen": 29066605, + "step": 1377, + "time_per_iteration": 2.7367517948150635 + }, + { + "auxiliary_loss_clip": 0.01232676, + "auxiliary_loss_mlp": 0.01037035, + "balance_loss_clip": 1.05683148, + "balance_loss_mlp": 1.02757597, + "epoch": 0.16569470330066735, + "flos": 24602723464320.0, + "grad_norm": 1.6099263500369454, + "language_loss": 0.88148689, + "learning_rate": 3.8100651646579146e-06, + "loss": 0.90418398, + "num_input_tokens_seen": 29085815, + "step": 1378, + "time_per_iteration": 2.760603904724121 + }, + { + "auxiliary_loss_clip": 0.01234417, + "auxiliary_loss_mlp": 0.01044306, + "balance_loss_clip": 1.057724, + "balance_loss_mlp": 1.03322589, + "epoch": 0.16581494619130643, + "flos": 15006207588480.0, + "grad_norm": 2.2976179192581334, + "language_loss": 0.92721003, + "learning_rate": 3.8097336978247317e-06, + "loss": 0.94999731, + "num_input_tokens_seen": 29102520, + "step": 1379, + "time_per_iteration": 2.6549367904663086 + }, + { + "auxiliary_loss_clip": 0.01225814, + "auxiliary_loss_mlp": 0.01033883, + "balance_loss_clip": 1.05823231, + "balance_loss_mlp": 1.02333283, + "epoch": 0.16593518908194552, + "flos": 17420338719360.0, + "grad_norm": 2.0589162581857536, + "language_loss": 0.88898194, + "learning_rate": 3.8094019564563854e-06, + "loss": 0.91157889, + "num_input_tokens_seen": 29119450, + "step": 1380, + "time_per_iteration": 2.6897621154785156 + }, + { + "auxiliary_loss_clip": 0.01228521, + "auxiliary_loss_mlp": 0.0138165, + "balance_loss_clip": 1.06436396, + "balance_loss_mlp": 1.00049973, + "epoch": 0.16605543197258463, + "flos": 20412631163520.0, + "grad_norm": 2.1915162347716097, + "language_loss": 0.7483381, + "learning_rate": 3.809069940603201e-06, + "loss": 0.77443981, + "num_input_tokens_seen": 29137405, + "step": 1381, + "time_per_iteration": 2.601637363433838 + }, + { + "auxiliary_loss_clip": 0.0123382, + "auxiliary_loss_mlp": 0.01037116, + "balance_loss_clip": 1.06139946, + "balance_loss_mlp": 1.02717364, + "epoch": 0.1661756748632237, + "flos": 14209745368320.0, + "grad_norm": 2.2809962438599465, + "language_loss": 0.7806226, + "learning_rate": 3.8087376503155452e-06, + "loss": 0.80333191, + "num_input_tokens_seen": 29154890, + "step": 1382, + "time_per_iteration": 2.7527830600738525 + }, + { + "auxiliary_loss_clip": 0.01148009, + "auxiliary_loss_mlp": 0.0100415, + "balance_loss_clip": 1.03587055, + "balance_loss_mlp": 1.00152731, + "epoch": 0.1662959177538628, + "flos": 66080877350400.0, + "grad_norm": 0.8923700832790884, + "language_loss": 0.56201243, + "learning_rate": 3.808405085643826e-06, + "loss": 0.583534, + "num_input_tokens_seen": 29219770, + "step": 1383, + "time_per_iteration": 3.2793803215026855 + }, + { + "auxiliary_loss_clip": 0.0123112, + "auxiliary_loss_mlp": 0.01381503, + "balance_loss_clip": 1.06679618, + "balance_loss_mlp": 1.00050449, + "epoch": 0.1664161606445019, + "flos": 20740567357440.0, + "grad_norm": 1.9919724060604111, + "language_loss": 0.89085591, + "learning_rate": 3.8080722466384925e-06, + "loss": 0.91698211, + "num_input_tokens_seen": 29237620, + "step": 1384, + "time_per_iteration": 2.614959955215454 + }, + { + "auxiliary_loss_clip": 0.01230081, + "auxiliary_loss_mlp": 0.01036635, + "balance_loss_clip": 1.06416774, + "balance_loss_mlp": 1.02647877, + "epoch": 0.166536403535141, + "flos": 25260930236160.0, + "grad_norm": 2.0100393308294953, + "language_loss": 0.70971525, + "learning_rate": 3.8077391333500376e-06, + "loss": 0.73238242, + "num_input_tokens_seen": 29256760, + "step": 1385, + "time_per_iteration": 2.7053446769714355 + }, + { + "auxiliary_loss_clip": 0.01248918, + "auxiliary_loss_mlp": 0.01033386, + "balance_loss_clip": 1.0666492, + "balance_loss_mlp": 1.02386713, + "epoch": 0.16665664642578007, + "flos": 25447450584960.0, + "grad_norm": 1.9629971937560147, + "language_loss": 0.76439583, + "learning_rate": 3.8074057458289934e-06, + "loss": 0.78721893, + "num_input_tokens_seen": 29277450, + "step": 1386, + "time_per_iteration": 2.8376307487487793 + }, + { + "auxiliary_loss_clip": 0.01242606, + "auxiliary_loss_mlp": 0.01035193, + "balance_loss_clip": 1.05866909, + "balance_loss_mlp": 1.02416575, + "epoch": 0.16677688931641918, + "flos": 22200767043840.0, + "grad_norm": 2.1353240321339544, + "language_loss": 0.82260287, + "learning_rate": 3.807072084125934e-06, + "loss": 0.8453809, + "num_input_tokens_seen": 29299300, + "step": 1387, + "time_per_iteration": 2.807129144668579 + }, + { + "auxiliary_loss_clip": 0.01244918, + "auxiliary_loss_mlp": 0.01041356, + "balance_loss_clip": 1.06478238, + "balance_loss_mlp": 1.03108597, + "epoch": 0.16689713220705826, + "flos": 16945958776320.0, + "grad_norm": 2.382937306128302, + "language_loss": 0.80790937, + "learning_rate": 3.806738148291477e-06, + "loss": 0.83077216, + "num_input_tokens_seen": 29316125, + "step": 1388, + "time_per_iteration": 2.6990749835968018 + }, + { + "auxiliary_loss_clip": 0.01263869, + "auxiliary_loss_mlp": 0.01035529, + "balance_loss_clip": 1.05294418, + "balance_loss_mlp": 1.0237751, + "epoch": 0.16701737509769735, + "flos": 36244423923840.0, + "grad_norm": 2.124618721142698, + "language_loss": 0.7094537, + "learning_rate": 3.8064039383762793e-06, + "loss": 0.73244774, + "num_input_tokens_seen": 29338490, + "step": 1389, + "time_per_iteration": 2.9215965270996094 + }, + { + "auxiliary_loss_clip": 0.01237227, + "auxiliary_loss_mlp": 0.01046407, + "balance_loss_clip": 1.06586409, + "balance_loss_mlp": 1.03553545, + "epoch": 0.16713761798833643, + "flos": 23258659426560.0, + "grad_norm": 5.312180874710829, + "language_loss": 0.77163756, + "learning_rate": 3.8060694544310396e-06, + "loss": 0.79447389, + "num_input_tokens_seen": 29357000, + "step": 1390, + "time_per_iteration": 2.66805362701416 + }, + { + "auxiliary_loss_clip": 0.01229636, + "auxiliary_loss_mlp": 0.0103602, + "balance_loss_clip": 1.06469512, + "balance_loss_mlp": 1.0252192, + "epoch": 0.16725786087897554, + "flos": 25302515207040.0, + "grad_norm": 1.786557662857729, + "language_loss": 0.78641915, + "learning_rate": 3.8057346965065006e-06, + "loss": 0.80907571, + "num_input_tokens_seen": 29378230, + "step": 1391, + "time_per_iteration": 3.08475923538208 + }, + { + "auxiliary_loss_clip": 0.0124493, + "auxiliary_loss_mlp": 0.01038251, + "balance_loss_clip": 1.06428194, + "balance_loss_mlp": 1.02805233, + "epoch": 0.16737810376961462, + "flos": 31831541516160.0, + "grad_norm": 1.9160148592866257, + "language_loss": 0.84424615, + "learning_rate": 3.805399664653443e-06, + "loss": 0.86707795, + "num_input_tokens_seen": 29400370, + "step": 1392, + "time_per_iteration": 2.830423355102539 + }, + { + "auxiliary_loss_clip": 0.01233842, + "auxiliary_loss_mlp": 0.01040275, + "balance_loss_clip": 1.06743002, + "balance_loss_mlp": 1.02881348, + "epoch": 0.1674983466602537, + "flos": 27961843553280.0, + "grad_norm": 2.5473012301841425, + "language_loss": 0.74117333, + "learning_rate": 3.805064358922692e-06, + "loss": 0.76391447, + "num_input_tokens_seen": 29418660, + "step": 1393, + "time_per_iteration": 3.6478915214538574 + }, + { + "auxiliary_loss_clip": 0.01244692, + "auxiliary_loss_mlp": 0.01038368, + "balance_loss_clip": 1.06623888, + "balance_loss_mlp": 1.0274725, + "epoch": 0.16761858955089282, + "flos": 21762656858880.0, + "grad_norm": 2.0084647447424255, + "language_loss": 0.8132174, + "learning_rate": 3.8047287793651136e-06, + "loss": 0.83604801, + "num_input_tokens_seen": 29440105, + "step": 1394, + "time_per_iteration": 2.6540863513946533 + }, + { + "auxiliary_loss_clip": 0.01257736, + "auxiliary_loss_mlp": 0.01038004, + "balance_loss_clip": 1.06187117, + "balance_loss_mlp": 1.02726889, + "epoch": 0.1677388324415319, + "flos": 23805507058560.0, + "grad_norm": 1.8076306407705511, + "language_loss": 0.8897869, + "learning_rate": 3.8043929260316137e-06, + "loss": 0.91274428, + "num_input_tokens_seen": 29458260, + "step": 1395, + "time_per_iteration": 2.709956169128418 + }, + { + "auxiliary_loss_clip": 0.01248591, + "auxiliary_loss_mlp": 0.01042179, + "balance_loss_clip": 1.06596923, + "balance_loss_mlp": 1.03098559, + "epoch": 0.16785907533217098, + "flos": 20558859431040.0, + "grad_norm": 1.9306742248525577, + "language_loss": 0.83647287, + "learning_rate": 3.8040567989731417e-06, + "loss": 0.8593806, + "num_input_tokens_seen": 29476205, + "step": 1396, + "time_per_iteration": 4.4638636112213135 + }, + { + "auxiliary_loss_clip": 0.01234272, + "auxiliary_loss_mlp": 0.01032242, + "balance_loss_clip": 1.06479931, + "balance_loss_mlp": 1.0224545, + "epoch": 0.16797931822281006, + "flos": 15669657745920.0, + "grad_norm": 1.9661750209658464, + "language_loss": 0.79595578, + "learning_rate": 3.8037203982406876e-06, + "loss": 0.81862086, + "num_input_tokens_seen": 29494370, + "step": 1397, + "time_per_iteration": 2.689168691635132 + }, + { + "auxiliary_loss_clip": 0.01232618, + "auxiliary_loss_mlp": 0.01040274, + "balance_loss_clip": 1.06881547, + "balance_loss_mlp": 1.02946186, + "epoch": 0.16809956111344918, + "flos": 16541101607040.0, + "grad_norm": 2.4199943112266085, + "language_loss": 0.73058003, + "learning_rate": 3.8033837238852835e-06, + "loss": 0.75330889, + "num_input_tokens_seen": 29511070, + "step": 1398, + "time_per_iteration": 3.551035165786743 + }, + { + "auxiliary_loss_clip": 0.01234331, + "auxiliary_loss_mlp": 0.0103553, + "balance_loss_clip": 1.0591557, + "balance_loss_mlp": 1.02557635, + "epoch": 0.16821980400408826, + "flos": 23258084808960.0, + "grad_norm": 1.7682422878908706, + "language_loss": 0.69816023, + "learning_rate": 3.8030467759580017e-06, + "loss": 0.72085881, + "num_input_tokens_seen": 29531990, + "step": 1399, + "time_per_iteration": 2.7235732078552246 + }, + { + "auxiliary_loss_clip": 0.0123898, + "auxiliary_loss_mlp": 0.01039176, + "balance_loss_clip": 1.06250513, + "balance_loss_mlp": 1.02772641, + "epoch": 0.16834004689472734, + "flos": 20774754126720.0, + "grad_norm": 1.9061546892606545, + "language_loss": 0.87029427, + "learning_rate": 3.802709554509958e-06, + "loss": 0.89307582, + "num_input_tokens_seen": 29549790, + "step": 1400, + "time_per_iteration": 2.695544958114624 + }, + { + "auxiliary_loss_clip": 0.0124143, + "auxiliary_loss_mlp": 0.01035702, + "balance_loss_clip": 1.06014967, + "balance_loss_mlp": 1.02573574, + "epoch": 0.16846028978536645, + "flos": 26687302289280.0, + "grad_norm": 1.7371243495421875, + "language_loss": 0.79377055, + "learning_rate": 3.8023720595923083e-06, + "loss": 0.81654185, + "num_input_tokens_seen": 29569045, + "step": 1401, + "time_per_iteration": 2.706847667694092 + }, + { + "auxiliary_loss_clip": 0.01254917, + "auxiliary_loss_mlp": 0.01034863, + "balance_loss_clip": 1.05489874, + "balance_loss_mlp": 1.02450991, + "epoch": 0.16858053267600553, + "flos": 18843298980480.0, + "grad_norm": 2.3680656403797853, + "language_loss": 0.87396115, + "learning_rate": 3.80203429125625e-06, + "loss": 0.89685893, + "num_input_tokens_seen": 29587220, + "step": 1402, + "time_per_iteration": 2.803422689437866 + }, + { + "auxiliary_loss_clip": 0.01258215, + "auxiliary_loss_mlp": 0.0103708, + "balance_loss_clip": 1.05548, + "balance_loss_mlp": 1.02625036, + "epoch": 0.16870077556664462, + "flos": 27744548227200.0, + "grad_norm": 1.711913279725502, + "language_loss": 0.70478988, + "learning_rate": 3.8016962495530225e-06, + "loss": 0.72774285, + "num_input_tokens_seen": 29606410, + "step": 1403, + "time_per_iteration": 2.8210065364837646 + }, + { + "auxiliary_loss_clip": 0.01231584, + "auxiliary_loss_mlp": 0.01038858, + "balance_loss_clip": 1.06604147, + "balance_loss_mlp": 1.02812886, + "epoch": 0.1688210184572837, + "flos": 13730768484480.0, + "grad_norm": 3.36259265365524, + "language_loss": 0.77208304, + "learning_rate": 3.8013579345339063e-06, + "loss": 0.79478747, + "num_input_tokens_seen": 29621275, + "step": 1404, + "time_per_iteration": 2.6208322048187256 + }, + { + "auxiliary_loss_clip": 0.01255084, + "auxiliary_loss_mlp": 0.01047113, + "balance_loss_clip": 1.05958843, + "balance_loss_mlp": 1.03599644, + "epoch": 0.1689412613479228, + "flos": 26468785900800.0, + "grad_norm": 1.8973287950849766, + "language_loss": 0.69435114, + "learning_rate": 3.801019346250224e-06, + "loss": 0.71737307, + "num_input_tokens_seen": 29641420, + "step": 1405, + "time_per_iteration": 2.7025461196899414 + }, + { + "auxiliary_loss_clip": 0.01236245, + "auxiliary_loss_mlp": 0.01045216, + "balance_loss_clip": 1.06290507, + "balance_loss_mlp": 1.03463674, + "epoch": 0.1690615042385619, + "flos": 21138852337920.0, + "grad_norm": 2.369015514610765, + "language_loss": 0.83468091, + "learning_rate": 3.8006804847533395e-06, + "loss": 0.85749555, + "num_input_tokens_seen": 29660935, + "step": 1406, + "time_per_iteration": 2.662839412689209 + }, + { + "auxiliary_loss_clip": 0.01230936, + "auxiliary_loss_mlp": 0.01041364, + "balance_loss_clip": 1.06576669, + "balance_loss_mlp": 1.03089714, + "epoch": 0.16918174712920098, + "flos": 20849340718080.0, + "grad_norm": 1.960234989485044, + "language_loss": 0.85482866, + "learning_rate": 3.8003413500946556e-06, + "loss": 0.87755167, + "num_input_tokens_seen": 29681045, + "step": 1407, + "time_per_iteration": 2.60711669921875 + }, + { + "auxiliary_loss_clip": 0.01249096, + "auxiliary_loss_mlp": 0.01045064, + "balance_loss_clip": 1.06347835, + "balance_loss_mlp": 1.03373349, + "epoch": 0.1693019900198401, + "flos": 16983270028800.0, + "grad_norm": 3.0805021457928317, + "language_loss": 0.83193356, + "learning_rate": 3.8000019423256216e-06, + "loss": 0.85487515, + "num_input_tokens_seen": 29698810, + "step": 1408, + "time_per_iteration": 2.6597180366516113 + }, + { + "auxiliary_loss_clip": 0.01231445, + "auxiliary_loss_mlp": 0.01040037, + "balance_loss_clip": 1.0611788, + "balance_loss_mlp": 1.02986884, + "epoch": 0.16942223291047917, + "flos": 26796901662720.0, + "grad_norm": 1.665948769265328, + "language_loss": 0.882285, + "learning_rate": 3.7996622614977234e-06, + "loss": 0.90499979, + "num_input_tokens_seen": 29720000, + "step": 1409, + "time_per_iteration": 2.7369158267974854 + }, + { + "auxiliary_loss_clip": 0.01245435, + "auxiliary_loss_mlp": 0.01033853, + "balance_loss_clip": 1.06338632, + "balance_loss_mlp": 1.02348733, + "epoch": 0.16954247580111825, + "flos": 18583700411520.0, + "grad_norm": 1.8537191112081177, + "language_loss": 0.79028916, + "learning_rate": 3.799322307662492e-06, + "loss": 0.81308204, + "num_input_tokens_seen": 29737820, + "step": 1410, + "time_per_iteration": 2.658952236175537 + }, + { + "auxiliary_loss_clip": 0.01263113, + "auxiliary_loss_mlp": 0.01045388, + "balance_loss_clip": 1.05821955, + "balance_loss_mlp": 1.03356266, + "epoch": 0.16966271869175734, + "flos": 13983651210240.0, + "grad_norm": 4.199271139606433, + "language_loss": 0.839921, + "learning_rate": 3.798982080871496e-06, + "loss": 0.863006, + "num_input_tokens_seen": 29752960, + "step": 1411, + "time_per_iteration": 2.6967320442199707 + }, + { + "auxiliary_loss_clip": 0.01230293, + "auxiliary_loss_mlp": 0.01038112, + "balance_loss_clip": 1.06404316, + "balance_loss_mlp": 1.02716291, + "epoch": 0.16978296158239645, + "flos": 37487328284160.0, + "grad_norm": 2.0248872061086747, + "language_loss": 0.67724407, + "learning_rate": 3.798641581176349e-06, + "loss": 0.6999281, + "num_input_tokens_seen": 29775240, + "step": 1412, + "time_per_iteration": 2.698676109313965 + }, + { + "auxiliary_loss_clip": 0.01244702, + "auxiliary_loss_mlp": 0.01034617, + "balance_loss_clip": 1.06027508, + "balance_loss_mlp": 1.0243175, + "epoch": 0.16990320447303553, + "flos": 28328958506880.0, + "grad_norm": 1.9630117610510192, + "language_loss": 0.74425399, + "learning_rate": 3.7983008086287044e-06, + "loss": 0.76704723, + "num_input_tokens_seen": 29796560, + "step": 1413, + "time_per_iteration": 2.732853651046753 + }, + { + "auxiliary_loss_clip": 0.01241792, + "auxiliary_loss_mlp": 0.01035377, + "balance_loss_clip": 1.05950797, + "balance_loss_mlp": 1.02405775, + "epoch": 0.1700234473636746, + "flos": 20188189031040.0, + "grad_norm": 2.1723966020831584, + "language_loss": 0.79444063, + "learning_rate": 3.797959763280257e-06, + "loss": 0.81721234, + "num_input_tokens_seen": 29815245, + "step": 1414, + "time_per_iteration": 2.6305739879608154 + }, + { + "auxiliary_loss_clip": 0.01240268, + "auxiliary_loss_mlp": 0.010341, + "balance_loss_clip": 1.06443381, + "balance_loss_mlp": 1.02391958, + "epoch": 0.17014369025431372, + "flos": 24858658846080.0, + "grad_norm": 1.960525287404172, + "language_loss": 0.79309207, + "learning_rate": 3.797618445182743e-06, + "loss": 0.81583571, + "num_input_tokens_seen": 29836640, + "step": 1415, + "time_per_iteration": 2.763334035873413 + }, + { + "auxiliary_loss_clip": 0.01252207, + "auxiliary_loss_mlp": 0.01041313, + "balance_loss_clip": 1.05407047, + "balance_loss_mlp": 1.02976131, + "epoch": 0.1702639331449528, + "flos": 16467233287680.0, + "grad_norm": 2.046179779576239, + "language_loss": 0.85007679, + "learning_rate": 3.79727685438794e-06, + "loss": 0.87301195, + "num_input_tokens_seen": 29850830, + "step": 1416, + "time_per_iteration": 2.6967010498046875 + }, + { + "auxiliary_loss_clip": 0.01142098, + "auxiliary_loss_mlp": 0.01007609, + "balance_loss_clip": 1.03775668, + "balance_loss_mlp": 1.00489068, + "epoch": 0.1703841760355919, + "flos": 52508870979840.0, + "grad_norm": 0.8382142031767728, + "language_loss": 0.61653489, + "learning_rate": 3.796934990947667e-06, + "loss": 0.63803196, + "num_input_tokens_seen": 29912515, + "step": 1417, + "time_per_iteration": 3.2530908584594727 + }, + { + "auxiliary_loss_clip": 0.01141623, + "auxiliary_loss_mlp": 0.01004183, + "balance_loss_clip": 1.03817225, + "balance_loss_mlp": 1.00157261, + "epoch": 0.170504418926231, + "flos": 49370637576960.0, + "grad_norm": 0.8776078217973854, + "language_loss": 0.62499779, + "learning_rate": 3.7965928549137854e-06, + "loss": 0.64645588, + "num_input_tokens_seen": 29969330, + "step": 1418, + "time_per_iteration": 3.175645589828491 + }, + { + "auxiliary_loss_clip": 0.01255577, + "auxiliary_loss_mlp": 0.01044459, + "balance_loss_clip": 1.05485284, + "balance_loss_mlp": 1.03262746, + "epoch": 0.17062466181687008, + "flos": 25849219184640.0, + "grad_norm": 1.9908122145038711, + "language_loss": 0.77927965, + "learning_rate": 3.7962504463381953e-06, + "loss": 0.80228007, + "num_input_tokens_seen": 29990820, + "step": 1419, + "time_per_iteration": 3.6623916625976562 + }, + { + "auxiliary_loss_clip": 0.01233381, + "auxiliary_loss_mlp": 0.01382384, + "balance_loss_clip": 1.05871391, + "balance_loss_mlp": 1.00061464, + "epoch": 0.17074490470750917, + "flos": 20960412549120.0, + "grad_norm": 1.7106947797021865, + "language_loss": 0.78854895, + "learning_rate": 3.7959077652728412e-06, + "loss": 0.81470656, + "num_input_tokens_seen": 30009275, + "step": 1420, + "time_per_iteration": 2.696183443069458 + }, + { + "auxiliary_loss_clip": 0.01246821, + "auxiliary_loss_mlp": 0.01034856, + "balance_loss_clip": 1.06046867, + "balance_loss_mlp": 1.0242877, + "epoch": 0.17086514759814825, + "flos": 20959766104320.0, + "grad_norm": 1.9516783835437739, + "language_loss": 0.77487648, + "learning_rate": 3.795564811769707e-06, + "loss": 0.79769325, + "num_input_tokens_seen": 30027630, + "step": 1421, + "time_per_iteration": 2.685779094696045 + }, + { + "auxiliary_loss_clip": 0.01249928, + "auxiliary_loss_mlp": 0.01036043, + "balance_loss_clip": 1.0663327, + "balance_loss_mlp": 1.02598739, + "epoch": 0.17098539048878736, + "flos": 28474073452800.0, + "grad_norm": 2.2262360073842826, + "language_loss": 0.78106207, + "learning_rate": 3.795221585880818e-06, + "loss": 0.8039217, + "num_input_tokens_seen": 30048310, + "step": 1422, + "time_per_iteration": 3.5796585083007812 + }, + { + "auxiliary_loss_clip": 0.01254789, + "auxiliary_loss_mlp": 0.01043791, + "balance_loss_clip": 1.06293726, + "balance_loss_mlp": 1.03314543, + "epoch": 0.17110563337942644, + "flos": 16290014561280.0, + "grad_norm": 1.783939674655806, + "language_loss": 0.90964532, + "learning_rate": 3.794878087658242e-06, + "loss": 0.93263108, + "num_input_tokens_seen": 30066080, + "step": 1423, + "time_per_iteration": 3.6446568965911865 + }, + { + "auxiliary_loss_clip": 0.01234918, + "auxiliary_loss_mlp": 0.0103887, + "balance_loss_clip": 1.05887604, + "balance_loss_mlp": 1.0286057, + "epoch": 0.17122587627006552, + "flos": 29674207693440.0, + "grad_norm": 2.6003069327416646, + "language_loss": 0.78705496, + "learning_rate": 3.7945343171540873e-06, + "loss": 0.80979288, + "num_input_tokens_seen": 30086955, + "step": 1424, + "time_per_iteration": 3.6843390464782715 + }, + { + "auxiliary_loss_clip": 0.01230584, + "auxiliary_loss_mlp": 0.01041587, + "balance_loss_clip": 1.06466246, + "balance_loss_mlp": 1.03016663, + "epoch": 0.17134611916070464, + "flos": 25338389915520.0, + "grad_norm": 1.8532314131208842, + "language_loss": 0.78850007, + "learning_rate": 3.7941902744205033e-06, + "loss": 0.81122172, + "num_input_tokens_seen": 30107990, + "step": 1425, + "time_per_iteration": 2.6942074298858643 + }, + { + "auxiliary_loss_clip": 0.01250041, + "auxiliary_loss_mlp": 0.01039797, + "balance_loss_clip": 1.06255865, + "balance_loss_mlp": 1.02874589, + "epoch": 0.17146636205134372, + "flos": 13953845900160.0, + "grad_norm": 2.0437522541425053, + "language_loss": 0.83385539, + "learning_rate": 3.7938459595096817e-06, + "loss": 0.85675377, + "num_input_tokens_seen": 30126535, + "step": 1426, + "time_per_iteration": 2.682443857192993 + }, + { + "auxiliary_loss_clip": 0.01241362, + "auxiliary_loss_mlp": 0.01039795, + "balance_loss_clip": 1.06182003, + "balance_loss_mlp": 1.0274266, + "epoch": 0.1715866049419828, + "flos": 23915214172800.0, + "grad_norm": 3.4308398773941025, + "language_loss": 0.86312151, + "learning_rate": 3.7935013724738545e-06, + "loss": 0.88593304, + "num_input_tokens_seen": 30147035, + "step": 1427, + "time_per_iteration": 2.739349603652954 + }, + { + "auxiliary_loss_clip": 0.0123023, + "auxiliary_loss_mlp": 0.01032988, + "balance_loss_clip": 1.05992723, + "balance_loss_mlp": 1.02212167, + "epoch": 0.17170684783262188, + "flos": 22709369669760.0, + "grad_norm": 2.049872190671265, + "language_loss": 0.78043836, + "learning_rate": 3.7931565133652945e-06, + "loss": 0.80307055, + "num_input_tokens_seen": 30167110, + "step": 1428, + "time_per_iteration": 2.9186816215515137 + }, + { + "auxiliary_loss_clip": 0.01228845, + "auxiliary_loss_mlp": 0.01043398, + "balance_loss_clip": 1.06306171, + "balance_loss_mlp": 1.03266299, + "epoch": 0.171827090723261, + "flos": 26613290315520.0, + "grad_norm": 2.709022099032585, + "language_loss": 0.67641389, + "learning_rate": 3.792811382236317e-06, + "loss": 0.69913638, + "num_input_tokens_seen": 30185620, + "step": 1429, + "time_per_iteration": 2.6466944217681885 + }, + { + "auxiliary_loss_clip": 0.01239086, + "auxiliary_loss_mlp": 0.01035278, + "balance_loss_clip": 1.0607276, + "balance_loss_mlp": 1.02510357, + "epoch": 0.17194733361390008, + "flos": 28148507556480.0, + "grad_norm": 3.186707926590084, + "language_loss": 0.78318977, + "learning_rate": 3.792465979139279e-06, + "loss": 0.80593342, + "num_input_tokens_seen": 30208225, + "step": 1430, + "time_per_iteration": 2.7551400661468506 + }, + { + "auxiliary_loss_clip": 0.01157542, + "auxiliary_loss_mlp": 0.01011358, + "balance_loss_clip": 1.03325844, + "balance_loss_mlp": 1.00889015, + "epoch": 0.17206757650453916, + "flos": 65530689753600.0, + "grad_norm": 0.9201220996590606, + "language_loss": 0.6565457, + "learning_rate": 3.792120304126576e-06, + "loss": 0.6782347, + "num_input_tokens_seen": 30271600, + "step": 1431, + "time_per_iteration": 3.3488457202911377 + }, + { + "auxiliary_loss_clip": 0.01278826, + "auxiliary_loss_mlp": 0.0103603, + "balance_loss_clip": 1.05300605, + "balance_loss_mlp": 1.02598107, + "epoch": 0.17218781939517827, + "flos": 22273486128000.0, + "grad_norm": 3.666717153646395, + "language_loss": 0.83386087, + "learning_rate": 3.791774357250649e-06, + "loss": 0.85700947, + "num_input_tokens_seen": 30290430, + "step": 1432, + "time_per_iteration": 3.4250335693359375 + }, + { + "auxiliary_loss_clip": 0.01242035, + "auxiliary_loss_mlp": 0.01046351, + "balance_loss_clip": 1.0607177, + "balance_loss_mlp": 1.03556836, + "epoch": 0.17230806228581735, + "flos": 14137313592960.0, + "grad_norm": 3.4992848602113646, + "language_loss": 0.79189992, + "learning_rate": 3.7914281385639757e-06, + "loss": 0.81478375, + "num_input_tokens_seen": 30308305, + "step": 1433, + "time_per_iteration": 2.771554470062256 + }, + { + "auxiliary_loss_clip": 0.01233649, + "auxiliary_loss_mlp": 0.01029859, + "balance_loss_clip": 1.05787182, + "balance_loss_mlp": 1.01940429, + "epoch": 0.17242830517645644, + "flos": 20704836303360.0, + "grad_norm": 1.7600172800320564, + "language_loss": 0.79359865, + "learning_rate": 3.7910816481190784e-06, + "loss": 0.81623375, + "num_input_tokens_seen": 30328120, + "step": 1434, + "time_per_iteration": 2.690605640411377 + }, + { + "auxiliary_loss_clip": 0.01231369, + "auxiliary_loss_mlp": 0.01037454, + "balance_loss_clip": 1.05775416, + "balance_loss_mlp": 1.02618885, + "epoch": 0.17254854806709552, + "flos": 30774582887040.0, + "grad_norm": 1.9746673937993962, + "language_loss": 0.74777138, + "learning_rate": 3.7907348859685193e-06, + "loss": 0.77045965, + "num_input_tokens_seen": 30349825, + "step": 1435, + "time_per_iteration": 2.8297369480133057 + }, + { + "auxiliary_loss_clip": 0.01229586, + "auxiliary_loss_mlp": 0.01033752, + "balance_loss_clip": 1.06191778, + "balance_loss_mlp": 1.02327943, + "epoch": 0.17266879095773463, + "flos": 26614726859520.0, + "grad_norm": 1.9730575793759146, + "language_loss": 0.80480373, + "learning_rate": 3.790387852164902e-06, + "loss": 0.82743704, + "num_input_tokens_seen": 30370555, + "step": 1436, + "time_per_iteration": 2.709707021713257 + }, + { + "auxiliary_loss_clip": 0.01238964, + "auxiliary_loss_mlp": 0.01038794, + "balance_loss_clip": 1.06318879, + "balance_loss_mlp": 1.02832723, + "epoch": 0.1727890338483737, + "flos": 20266295155200.0, + "grad_norm": 1.7417055734577653, + "language_loss": 0.76784599, + "learning_rate": 3.7900405467608707e-06, + "loss": 0.79062355, + "num_input_tokens_seen": 30390100, + "step": 1437, + "time_per_iteration": 2.678910493850708 + }, + { + "auxiliary_loss_clip": 0.01231802, + "auxiliary_loss_mlp": 0.01042378, + "balance_loss_clip": 1.04779625, + "balance_loss_mlp": 1.03139257, + "epoch": 0.1729092767390128, + "flos": 18179812909440.0, + "grad_norm": 3.048525168473818, + "language_loss": 0.78676689, + "learning_rate": 3.7896929698091114e-06, + "loss": 0.80950868, + "num_input_tokens_seen": 30402915, + "step": 1438, + "time_per_iteration": 2.706895112991333 + }, + { + "auxiliary_loss_clip": 0.01229862, + "auxiliary_loss_mlp": 0.01040844, + "balance_loss_clip": 1.06573212, + "balance_loss_mlp": 1.03047919, + "epoch": 0.1730295196296519, + "flos": 26759518583040.0, + "grad_norm": 4.869688436756144, + "language_loss": 0.68082082, + "learning_rate": 3.7893451213623518e-06, + "loss": 0.70352793, + "num_input_tokens_seen": 30420145, + "step": 1439, + "time_per_iteration": 2.641766309738159 + }, + { + "auxiliary_loss_clip": 0.01235166, + "auxiliary_loss_mlp": 0.01381821, + "balance_loss_clip": 1.06346488, + "balance_loss_mlp": 1.00041056, + "epoch": 0.173149762520291, + "flos": 23842531002240.0, + "grad_norm": 2.2180481744650864, + "language_loss": 0.8226639, + "learning_rate": 3.7889970014733606e-06, + "loss": 0.84883374, + "num_input_tokens_seen": 30439250, + "step": 1440, + "time_per_iteration": 2.902454137802124 + }, + { + "auxiliary_loss_clip": 0.01235407, + "auxiliary_loss_mlp": 0.01040597, + "balance_loss_clip": 1.05220497, + "balance_loss_mlp": 1.0303092, + "epoch": 0.17327000541093007, + "flos": 23368186972800.0, + "grad_norm": 1.6833992379878302, + "language_loss": 0.77827168, + "learning_rate": 3.7886486101949463e-06, + "loss": 0.80103171, + "num_input_tokens_seen": 30460430, + "step": 1441, + "time_per_iteration": 2.759709119796753 + }, + { + "auxiliary_loss_clip": 0.01239637, + "auxiliary_loss_mlp": 0.01038479, + "balance_loss_clip": 1.05239487, + "balance_loss_mlp": 1.02853703, + "epoch": 0.17339024830156918, + "flos": 18221290139520.0, + "grad_norm": 1.8961654170270696, + "language_loss": 0.87948525, + "learning_rate": 3.7882999475799594e-06, + "loss": 0.90226644, + "num_input_tokens_seen": 30478465, + "step": 1442, + "time_per_iteration": 2.6934053897857666 + }, + { + "auxiliary_loss_clip": 0.01234445, + "auxiliary_loss_mlp": 0.01040023, + "balance_loss_clip": 1.05404198, + "balance_loss_mlp": 1.02954423, + "epoch": 0.17351049119220827, + "flos": 23332024955520.0, + "grad_norm": 1.9570269594609566, + "language_loss": 0.81329334, + "learning_rate": 3.787951013681293e-06, + "loss": 0.83603799, + "num_input_tokens_seen": 30496510, + "step": 1443, + "time_per_iteration": 2.783843755722046 + }, + { + "auxiliary_loss_clip": 0.01232214, + "auxiliary_loss_mlp": 0.01041981, + "balance_loss_clip": 1.06137085, + "balance_loss_mlp": 1.03012538, + "epoch": 0.17363073408284735, + "flos": 23803495896960.0, + "grad_norm": 5.928252727618891, + "language_loss": 0.77654743, + "learning_rate": 3.787601808551879e-06, + "loss": 0.79928941, + "num_input_tokens_seen": 30516325, + "step": 1444, + "time_per_iteration": 3.6301369667053223 + }, + { + "auxiliary_loss_clip": 0.0125209, + "auxiliary_loss_mlp": 0.01044573, + "balance_loss_clip": 1.05966532, + "balance_loss_mlp": 1.03410053, + "epoch": 0.17375097697348643, + "flos": 18515290959360.0, + "grad_norm": 2.2138462830633387, + "language_loss": 0.83763099, + "learning_rate": 3.7872523322446926e-06, + "loss": 0.86059761, + "num_input_tokens_seen": 30535210, + "step": 1445, + "time_per_iteration": 2.74275279045105 + }, + { + "auxiliary_loss_clip": 0.01262062, + "auxiliary_loss_mlp": 0.01036891, + "balance_loss_clip": 1.05287266, + "balance_loss_mlp": 1.02715135, + "epoch": 0.17387121986412554, + "flos": 38877897456000.0, + "grad_norm": 2.453818869507688, + "language_loss": 0.60232723, + "learning_rate": 3.7869025848127478e-06, + "loss": 0.6253168, + "num_input_tokens_seen": 30559405, + "step": 1446, + "time_per_iteration": 2.991410732269287 + }, + { + "auxiliary_loss_clip": 0.01236272, + "auxiliary_loss_mlp": 0.01037216, + "balance_loss_clip": 1.06112409, + "balance_loss_mlp": 1.02672553, + "epoch": 0.17399146275476463, + "flos": 20375714960640.0, + "grad_norm": 2.841977203884939, + "language_loss": 0.80950606, + "learning_rate": 3.786552566309102e-06, + "loss": 0.83224094, + "num_input_tokens_seen": 30577615, + "step": 1447, + "time_per_iteration": 2.6662395000457764 + }, + { + "auxiliary_loss_clip": 0.0123983, + "auxiliary_loss_mlp": 0.01381125, + "balance_loss_clip": 1.06269073, + "balance_loss_mlp": 1.00027156, + "epoch": 0.1741117056454037, + "flos": 19164339763200.0, + "grad_norm": 2.575452913866764, + "language_loss": 0.85465741, + "learning_rate": 3.7862022767868517e-06, + "loss": 0.88086694, + "num_input_tokens_seen": 30595205, + "step": 1448, + "time_per_iteration": 3.8646156787872314 + }, + { + "auxiliary_loss_clip": 0.01247938, + "auxiliary_loss_mlp": 0.01042283, + "balance_loss_clip": 1.06221545, + "balance_loss_mlp": 1.03206706, + "epoch": 0.17423194853604282, + "flos": 25374300537600.0, + "grad_norm": 1.9711615695778673, + "language_loss": 0.84047961, + "learning_rate": 3.7858517162991367e-06, + "loss": 0.8633818, + "num_input_tokens_seen": 30615280, + "step": 1449, + "time_per_iteration": 3.66196346282959 + }, + { + "auxiliary_loss_clip": 0.01249853, + "auxiliary_loss_mlp": 0.01035872, + "balance_loss_clip": 1.05708003, + "balance_loss_mlp": 1.02576911, + "epoch": 0.1743521914266819, + "flos": 25191874339200.0, + "grad_norm": 2.463197704570653, + "language_loss": 0.61025518, + "learning_rate": 3.7855008848991363e-06, + "loss": 0.63311243, + "num_input_tokens_seen": 30633485, + "step": 1450, + "time_per_iteration": 3.737910270690918 + }, + { + "auxiliary_loss_clip": 0.01237736, + "auxiliary_loss_mlp": 0.01038323, + "balance_loss_clip": 1.06076288, + "balance_loss_mlp": 1.02836323, + "epoch": 0.17447243431732098, + "flos": 25666577504640.0, + "grad_norm": 2.031236842377336, + "language_loss": 0.77769381, + "learning_rate": 3.7851497826400714e-06, + "loss": 0.80045438, + "num_input_tokens_seen": 30653625, + "step": 1451, + "time_per_iteration": 2.661381721496582 + }, + { + "auxiliary_loss_clip": 0.01228429, + "auxiliary_loss_mlp": 0.01035326, + "balance_loss_clip": 1.0656966, + "balance_loss_mlp": 1.024019, + "epoch": 0.17459267720796007, + "flos": 36281950657920.0, + "grad_norm": 1.831841746838739, + "language_loss": 0.76170552, + "learning_rate": 3.7847984095752034e-06, + "loss": 0.78434306, + "num_input_tokens_seen": 30677080, + "step": 1452, + "time_per_iteration": 2.727843999862671 + }, + { + "auxiliary_loss_clip": 0.01225726, + "auxiliary_loss_mlp": 0.01037763, + "balance_loss_clip": 1.06297433, + "balance_loss_mlp": 1.02796388, + "epoch": 0.17471292009859918, + "flos": 20011113959040.0, + "grad_norm": 12.819032573544042, + "language_loss": 0.8013804, + "learning_rate": 3.784446765757836e-06, + "loss": 0.82401526, + "num_input_tokens_seen": 30695725, + "step": 1453, + "time_per_iteration": 2.575303554534912 + }, + { + "auxiliary_loss_clip": 0.01225033, + "auxiliary_loss_mlp": 0.01038866, + "balance_loss_clip": 1.05342674, + "balance_loss_mlp": 1.02769566, + "epoch": 0.17483316298923826, + "flos": 27819242559360.0, + "grad_norm": 2.3762743924244374, + "language_loss": 0.78051454, + "learning_rate": 3.7840948512413133e-06, + "loss": 0.80315351, + "num_input_tokens_seen": 30713310, + "step": 1454, + "time_per_iteration": 2.7374441623687744 + }, + { + "auxiliary_loss_clip": 0.01245532, + "auxiliary_loss_mlp": 0.01042513, + "balance_loss_clip": 1.0598681, + "balance_loss_mlp": 1.03138459, + "epoch": 0.17495340587987734, + "flos": 44017934791680.0, + "grad_norm": 1.8557488272531872, + "language_loss": 0.7863971, + "learning_rate": 3.7837426660790196e-06, + "loss": 0.80927753, + "num_input_tokens_seen": 30734725, + "step": 1455, + "time_per_iteration": 2.9281342029571533 + }, + { + "auxiliary_loss_clip": 0.01225086, + "auxiliary_loss_mlp": 0.01033282, + "balance_loss_clip": 1.06299329, + "balance_loss_mlp": 1.02377546, + "epoch": 0.17507364877051645, + "flos": 20885825957760.0, + "grad_norm": 1.9290274275841843, + "language_loss": 0.82005686, + "learning_rate": 3.783390210324382e-06, + "loss": 0.84264052, + "num_input_tokens_seen": 30754450, + "step": 1456, + "time_per_iteration": 2.5959391593933105 + }, + { + "auxiliary_loss_clip": 0.01250152, + "auxiliary_loss_mlp": 0.01039116, + "balance_loss_clip": 1.06090021, + "balance_loss_mlp": 1.02919829, + "epoch": 0.17519389166115554, + "flos": 24717602136960.0, + "grad_norm": 2.6843336555266224, + "language_loss": 0.72450191, + "learning_rate": 3.7830374840308676e-06, + "loss": 0.74739462, + "num_input_tokens_seen": 30774605, + "step": 1457, + "time_per_iteration": 2.7080941200256348 + }, + { + "auxiliary_loss_clip": 0.01236268, + "auxiliary_loss_mlp": 0.01033675, + "balance_loss_clip": 1.06399298, + "balance_loss_mlp": 1.02348316, + "epoch": 0.17531413455179462, + "flos": 23798144770560.0, + "grad_norm": 2.6959253547740225, + "language_loss": 0.82603312, + "learning_rate": 3.7826844872519842e-06, + "loss": 0.84873259, + "num_input_tokens_seen": 30792460, + "step": 1458, + "time_per_iteration": 2.666579484939575 + }, + { + "auxiliary_loss_clip": 0.0123945, + "auxiliary_loss_mlp": 0.01034069, + "balance_loss_clip": 1.06179595, + "balance_loss_mlp": 1.02420998, + "epoch": 0.1754343774424337, + "flos": 24572379450240.0, + "grad_norm": 1.9995516694176123, + "language_loss": 0.72693837, + "learning_rate": 3.782331220041282e-06, + "loss": 0.74967355, + "num_input_tokens_seen": 30812525, + "step": 1459, + "time_per_iteration": 2.759674549102783 + }, + { + "auxiliary_loss_clip": 0.0125882, + "auxiliary_loss_mlp": 0.01040679, + "balance_loss_clip": 1.05934465, + "balance_loss_mlp": 1.02944386, + "epoch": 0.17555462033307281, + "flos": 18114599767680.0, + "grad_norm": 2.7235515629820046, + "language_loss": 0.83210081, + "learning_rate": 3.7819776824523504e-06, + "loss": 0.8550958, + "num_input_tokens_seen": 30830390, + "step": 1460, + "time_per_iteration": 2.7024941444396973 + }, + { + "auxiliary_loss_clip": 0.01248693, + "auxiliary_loss_mlp": 0.01043578, + "balance_loss_clip": 1.06108284, + "balance_loss_mlp": 1.03203869, + "epoch": 0.1756748632237119, + "flos": 28366018364160.0, + "grad_norm": 2.116957275908154, + "language_loss": 0.83752042, + "learning_rate": 3.7816238745388213e-06, + "loss": 0.86044312, + "num_input_tokens_seen": 30849935, + "step": 1461, + "time_per_iteration": 2.7612240314483643 + }, + { + "auxiliary_loss_clip": 0.01244477, + "auxiliary_loss_mlp": 0.01036808, + "balance_loss_clip": 1.05797505, + "balance_loss_mlp": 1.0274446, + "epoch": 0.17579510611435098, + "flos": 25732939881600.0, + "grad_norm": 2.197570483492395, + "language_loss": 0.87005216, + "learning_rate": 3.781269796354367e-06, + "loss": 0.89286506, + "num_input_tokens_seen": 30869555, + "step": 1462, + "time_per_iteration": 2.7088396549224854 + }, + { + "auxiliary_loss_clip": 0.01244344, + "auxiliary_loss_mlp": 0.01045966, + "balance_loss_clip": 1.0626086, + "balance_loss_mlp": 1.03496933, + "epoch": 0.1759153490049901, + "flos": 18588081870720.0, + "grad_norm": 1.7061489689467335, + "language_loss": 0.86039716, + "learning_rate": 3.7809154479527006e-06, + "loss": 0.8833003, + "num_input_tokens_seen": 30888760, + "step": 1463, + "time_per_iteration": 2.689037561416626 + }, + { + "auxiliary_loss_clip": 0.01241798, + "auxiliary_loss_mlp": 0.01038563, + "balance_loss_clip": 1.05884242, + "balance_loss_mlp": 1.02864516, + "epoch": 0.17603559189562917, + "flos": 18619323724800.0, + "grad_norm": 2.1223764468907063, + "language_loss": 0.84623957, + "learning_rate": 3.780560829387577e-06, + "loss": 0.86904317, + "num_input_tokens_seen": 30907260, + "step": 1464, + "time_per_iteration": 2.70805287361145 + }, + { + "auxiliary_loss_clip": 0.01137839, + "auxiliary_loss_mlp": 0.01000246, + "balance_loss_clip": 1.03495336, + "balance_loss_mlp": 0.99780226, + "epoch": 0.17615583478626826, + "flos": 60530775373440.0, + "grad_norm": 0.8534514732617525, + "language_loss": 0.57922482, + "learning_rate": 3.7802059407127915e-06, + "loss": 0.60060561, + "num_input_tokens_seen": 30965810, + "step": 1465, + "time_per_iteration": 3.191629409790039 + }, + { + "auxiliary_loss_clip": 0.01231672, + "auxiliary_loss_mlp": 0.01032685, + "balance_loss_clip": 1.05606556, + "balance_loss_mlp": 1.02258229, + "epoch": 0.17627607767690734, + "flos": 23616221362560.0, + "grad_norm": 1.992210659847845, + "language_loss": 0.86162174, + "learning_rate": 3.7798507819821797e-06, + "loss": 0.8842653, + "num_input_tokens_seen": 30982935, + "step": 1466, + "time_per_iteration": 2.7208423614501953 + }, + { + "auxiliary_loss_clip": 0.01240042, + "auxiliary_loss_mlp": 0.01039275, + "balance_loss_clip": 1.05882096, + "balance_loss_mlp": 1.02893972, + "epoch": 0.17639632056754645, + "flos": 17639070589440.0, + "grad_norm": 2.6240976315397453, + "language_loss": 0.79168695, + "learning_rate": 3.7794953532496197e-06, + "loss": 0.81448013, + "num_input_tokens_seen": 30998840, + "step": 1467, + "time_per_iteration": 2.6789045333862305 + }, + { + "auxiliary_loss_clip": 0.01164761, + "auxiliary_loss_mlp": 0.01376529, + "balance_loss_clip": 1.03806806, + "balance_loss_mlp": 0.99991208, + "epoch": 0.17651656345818553, + "flos": 57932604910080.0, + "grad_norm": 0.8985418838191094, + "language_loss": 0.57962537, + "learning_rate": 3.7791396545690295e-06, + "loss": 0.60503823, + "num_input_tokens_seen": 31060075, + "step": 1468, + "time_per_iteration": 3.3316023349761963 + }, + { + "auxiliary_loss_clip": 0.0123207, + "auxiliary_loss_mlp": 0.01031573, + "balance_loss_clip": 1.06482029, + "balance_loss_mlp": 1.02130294, + "epoch": 0.17663680634882462, + "flos": 22929502170240.0, + "grad_norm": 2.3325254093391212, + "language_loss": 0.80501485, + "learning_rate": 3.7787836859943685e-06, + "loss": 0.82765126, + "num_input_tokens_seen": 31078800, + "step": 1469, + "time_per_iteration": 2.6074013710021973 + }, + { + "auxiliary_loss_clip": 0.01233748, + "auxiliary_loss_mlp": 0.01035386, + "balance_loss_clip": 1.06242406, + "balance_loss_mlp": 1.02489531, + "epoch": 0.17675704923946373, + "flos": 22637979388800.0, + "grad_norm": 2.9261634420932308, + "language_loss": 0.78907043, + "learning_rate": 3.7784274475796363e-06, + "loss": 0.81176186, + "num_input_tokens_seen": 31097430, + "step": 1470, + "time_per_iteration": 3.636002779006958 + }, + { + "auxiliary_loss_clip": 0.01250106, + "auxiliary_loss_mlp": 0.01035594, + "balance_loss_clip": 1.05748379, + "balance_loss_mlp": 1.02435851, + "epoch": 0.1768772921301028, + "flos": 27126525795840.0, + "grad_norm": 2.165307719761903, + "language_loss": 0.76114398, + "learning_rate": 3.7780709393788745e-06, + "loss": 0.78400099, + "num_input_tokens_seen": 31117905, + "step": 1471, + "time_per_iteration": 2.7614262104034424 + }, + { + "auxiliary_loss_clip": 0.01223179, + "auxiliary_loss_mlp": 0.01039576, + "balance_loss_clip": 1.0638181, + "balance_loss_mlp": 1.02964544, + "epoch": 0.1769975350207419, + "flos": 19172133014400.0, + "grad_norm": 1.9692303817606533, + "language_loss": 0.75063527, + "learning_rate": 3.777714161446165e-06, + "loss": 0.77326286, + "num_input_tokens_seen": 31137610, + "step": 1472, + "time_per_iteration": 2.6271891593933105 + }, + { + "auxiliary_loss_clip": 0.01236768, + "auxiliary_loss_mlp": 0.01035613, + "balance_loss_clip": 1.0651679, + "balance_loss_mlp": 1.02608252, + "epoch": 0.177117777911381, + "flos": 36134932291200.0, + "grad_norm": 2.191403048733398, + "language_loss": 0.69423407, + "learning_rate": 3.7773571138356304e-06, + "loss": 0.71695793, + "num_input_tokens_seen": 31157780, + "step": 1473, + "time_per_iteration": 2.828432083129883 + }, + { + "auxiliary_loss_clip": 0.01241115, + "auxiliary_loss_mlp": 0.01030151, + "balance_loss_clip": 1.0566237, + "balance_loss_mlp": 1.02076304, + "epoch": 0.17723802080202009, + "flos": 22090593052800.0, + "grad_norm": 2.2012732382213165, + "language_loss": 0.89031678, + "learning_rate": 3.776999796601435e-06, + "loss": 0.91302943, + "num_input_tokens_seen": 31176540, + "step": 1474, + "time_per_iteration": 3.627471446990967 + }, + { + "auxiliary_loss_clip": 0.01237104, + "auxiliary_loss_mlp": 0.01035931, + "balance_loss_clip": 1.06329966, + "balance_loss_mlp": 1.02610207, + "epoch": 0.17735826369265917, + "flos": 30222671437440.0, + "grad_norm": 2.3671675766239897, + "language_loss": 0.72403514, + "learning_rate": 3.776642209797783e-06, + "loss": 0.74676549, + "num_input_tokens_seen": 31198370, + "step": 1475, + "time_per_iteration": 3.812432289123535 + }, + { + "auxiliary_loss_clip": 0.01224868, + "auxiliary_loss_mlp": 0.01033366, + "balance_loss_clip": 1.06018615, + "balance_loss_mlp": 1.02263677, + "epoch": 0.17747850658329825, + "flos": 21397588980480.0, + "grad_norm": 2.2927708981876855, + "language_loss": 0.77884722, + "learning_rate": 3.7762843534789205e-06, + "loss": 0.80142957, + "num_input_tokens_seen": 31217120, + "step": 1476, + "time_per_iteration": 3.6553423404693604 + }, + { + "auxiliary_loss_clip": 0.01246638, + "auxiliary_loss_mlp": 0.01044333, + "balance_loss_clip": 1.06031346, + "balance_loss_mlp": 1.03419375, + "epoch": 0.17759874947393736, + "flos": 16983341856000.0, + "grad_norm": 2.8312642243508774, + "language_loss": 0.88431382, + "learning_rate": 3.7759262276991343e-06, + "loss": 0.90722358, + "num_input_tokens_seen": 31234730, + "step": 1477, + "time_per_iteration": 2.646496295928955 + }, + { + "auxiliary_loss_clip": 0.01245421, + "auxiliary_loss_mlp": 0.01034712, + "balance_loss_clip": 1.06115663, + "balance_loss_mlp": 1.0252645, + "epoch": 0.17771899236457644, + "flos": 11546107390080.0, + "grad_norm": 2.140470177257644, + "language_loss": 0.80524153, + "learning_rate": 3.7755678325127506e-06, + "loss": 0.82804286, + "num_input_tokens_seen": 31252410, + "step": 1478, + "time_per_iteration": 2.694952964782715 + }, + { + "auxiliary_loss_clip": 0.01251558, + "auxiliary_loss_mlp": 0.01034027, + "balance_loss_clip": 1.05879116, + "balance_loss_mlp": 1.02393007, + "epoch": 0.17783923525521553, + "flos": 18807747494400.0, + "grad_norm": 2.2197318014132916, + "language_loss": 0.75858736, + "learning_rate": 3.7752091679741393e-06, + "loss": 0.78144324, + "num_input_tokens_seen": 31270200, + "step": 1479, + "time_per_iteration": 2.7566070556640625 + }, + { + "auxiliary_loss_clip": 0.01232889, + "auxiliary_loss_mlp": 0.0103845, + "balance_loss_clip": 1.06346774, + "balance_loss_mlp": 1.02837682, + "epoch": 0.17795947814585464, + "flos": 30408365773440.0, + "grad_norm": 3.4515074522680607, + "language_loss": 0.77549154, + "learning_rate": 3.774850234137708e-06, + "loss": 0.79820496, + "num_input_tokens_seen": 31287495, + "step": 1480, + "time_per_iteration": 2.7120323181152344 + }, + { + "auxiliary_loss_clip": 0.01225951, + "auxiliary_loss_mlp": 0.01032503, + "balance_loss_clip": 1.05919659, + "balance_loss_mlp": 1.02178001, + "epoch": 0.17807972103649372, + "flos": 24389055411840.0, + "grad_norm": 2.472787032985449, + "language_loss": 0.82589281, + "learning_rate": 3.7744910310579076e-06, + "loss": 0.84847742, + "num_input_tokens_seen": 31306420, + "step": 1481, + "time_per_iteration": 2.6722609996795654 + }, + { + "auxiliary_loss_clip": 0.01223327, + "auxiliary_loss_mlp": 0.01036194, + "balance_loss_clip": 1.0642333, + "balance_loss_mlp": 1.02608538, + "epoch": 0.1781999639271328, + "flos": 20301559332480.0, + "grad_norm": 3.2679535295233273, + "language_loss": 0.85375106, + "learning_rate": 3.774131558789229e-06, + "loss": 0.87634635, + "num_input_tokens_seen": 31325750, + "step": 1482, + "time_per_iteration": 2.6041219234466553 + }, + { + "auxiliary_loss_clip": 0.01227459, + "auxiliary_loss_mlp": 0.01380726, + "balance_loss_clip": 1.06658053, + "balance_loss_mlp": 1.00028503, + "epoch": 0.1783202068177719, + "flos": 15924479806080.0, + "grad_norm": 2.393663028552901, + "language_loss": 0.69662279, + "learning_rate": 3.773771817386203e-06, + "loss": 0.72270465, + "num_input_tokens_seen": 31343080, + "step": 1483, + "time_per_iteration": 2.5688366889953613 + }, + { + "auxiliary_loss_clip": 0.01238464, + "auxiliary_loss_mlp": 0.01038883, + "balance_loss_clip": 1.05958581, + "balance_loss_mlp": 1.0284822, + "epoch": 0.178440449708411, + "flos": 20631758083200.0, + "grad_norm": 1.5919834832506046, + "language_loss": 0.79390466, + "learning_rate": 3.773411806903403e-06, + "loss": 0.81667817, + "num_input_tokens_seen": 31362160, + "step": 1484, + "time_per_iteration": 2.6785285472869873 + }, + { + "auxiliary_loss_clip": 0.01262373, + "auxiliary_loss_mlp": 0.01043432, + "balance_loss_clip": 1.05304229, + "balance_loss_mlp": 1.03319788, + "epoch": 0.17856069259905008, + "flos": 21686059105920.0, + "grad_norm": 3.791993313592001, + "language_loss": 0.94689059, + "learning_rate": 3.7730515273954415e-06, + "loss": 0.96994859, + "num_input_tokens_seen": 31380770, + "step": 1485, + "time_per_iteration": 2.8039982318878174 + }, + { + "auxiliary_loss_clip": 0.01225478, + "auxiliary_loss_mlp": 0.01042, + "balance_loss_clip": 1.06580424, + "balance_loss_mlp": 1.0317421, + "epoch": 0.17868093548968916, + "flos": 26572962320640.0, + "grad_norm": 1.8660508874390873, + "language_loss": 0.84976971, + "learning_rate": 3.772690978916973e-06, + "loss": 0.87244451, + "num_input_tokens_seen": 31400525, + "step": 1486, + "time_per_iteration": 2.624284267425537 + }, + { + "auxiliary_loss_clip": 0.01232721, + "auxiliary_loss_mlp": 0.0103052, + "balance_loss_clip": 1.06453204, + "balance_loss_mlp": 1.02012503, + "epoch": 0.17880117838032827, + "flos": 18581006891520.0, + "grad_norm": 2.2573200281004637, + "language_loss": 0.86699641, + "learning_rate": 3.772330161522693e-06, + "loss": 0.88962883, + "num_input_tokens_seen": 31418435, + "step": 1487, + "time_per_iteration": 2.6687088012695312 + }, + { + "auxiliary_loss_clip": 0.01238165, + "auxiliary_loss_mlp": 0.0104513, + "balance_loss_clip": 1.06200135, + "balance_loss_mlp": 1.03458548, + "epoch": 0.17892142127096736, + "flos": 26541217676160.0, + "grad_norm": 1.9357913097867148, + "language_loss": 0.7964232, + "learning_rate": 3.7719690752673365e-06, + "loss": 0.81925607, + "num_input_tokens_seen": 31439230, + "step": 1488, + "time_per_iteration": 2.7243340015411377 + }, + { + "auxiliary_loss_clip": 0.01248657, + "auxiliary_loss_mlp": 0.01038329, + "balance_loss_clip": 1.0600276, + "balance_loss_mlp": 1.02866066, + "epoch": 0.17904166416160644, + "flos": 23872623621120.0, + "grad_norm": 2.2092260895481766, + "language_loss": 0.78144217, + "learning_rate": 3.7716077202056796e-06, + "loss": 0.80431199, + "num_input_tokens_seen": 31457705, + "step": 1489, + "time_per_iteration": 2.7625374794006348 + }, + { + "auxiliary_loss_clip": 0.0122236, + "auxiliary_loss_mlp": 0.01036665, + "balance_loss_clip": 1.05839944, + "balance_loss_mlp": 1.02635312, + "epoch": 0.17916190705224552, + "flos": 19134426712320.0, + "grad_norm": 2.0538040679913716, + "language_loss": 0.93618846, + "learning_rate": 3.7712460963925404e-06, + "loss": 0.95877874, + "num_input_tokens_seen": 31473645, + "step": 1490, + "time_per_iteration": 2.6756041049957275 + }, + { + "auxiliary_loss_clip": 0.01229588, + "auxiliary_loss_mlp": 0.01035248, + "balance_loss_clip": 1.05899882, + "balance_loss_mlp": 1.02612865, + "epoch": 0.17928214994288463, + "flos": 25152120961920.0, + "grad_norm": 2.0876673825073326, + "language_loss": 0.75328279, + "learning_rate": 3.7708842038827775e-06, + "loss": 0.77593118, + "num_input_tokens_seen": 31492605, + "step": 1491, + "time_per_iteration": 2.780306577682495 + }, + { + "auxiliary_loss_clip": 0.01232451, + "auxiliary_loss_mlp": 0.01030903, + "balance_loss_clip": 1.06214976, + "balance_loss_mlp": 1.02150345, + "epoch": 0.17940239283352372, + "flos": 22384629786240.0, + "grad_norm": 2.0715617997414593, + "language_loss": 0.85949409, + "learning_rate": 3.770522042731288e-06, + "loss": 0.88212764, + "num_input_tokens_seen": 31514500, + "step": 1492, + "time_per_iteration": 2.6666274070739746 + }, + { + "auxiliary_loss_clip": 0.01241661, + "auxiliary_loss_mlp": 0.01044305, + "balance_loss_clip": 1.05664277, + "balance_loss_mlp": 1.03414869, + "epoch": 0.1795226357241628, + "flos": 23178685795200.0, + "grad_norm": 1.931477903697411, + "language_loss": 0.87555933, + "learning_rate": 3.7701596129930122e-06, + "loss": 0.89841902, + "num_input_tokens_seen": 31533225, + "step": 1493, + "time_per_iteration": 2.73946475982666 + }, + { + "auxiliary_loss_clip": 0.01254939, + "auxiliary_loss_mlp": 0.01031372, + "balance_loss_clip": 1.05887818, + "balance_loss_mlp": 1.02118564, + "epoch": 0.1796428786148019, + "flos": 22090413484800.0, + "grad_norm": 2.070695108733145, + "language_loss": 0.73413634, + "learning_rate": 3.7697969147229315e-06, + "loss": 0.75699949, + "num_input_tokens_seen": 31551385, + "step": 1494, + "time_per_iteration": 2.714667558670044 + }, + { + "auxiliary_loss_clip": 0.01231042, + "auxiliary_loss_mlp": 0.01035739, + "balance_loss_clip": 1.06327009, + "balance_loss_mlp": 1.02630949, + "epoch": 0.179763121505441, + "flos": 21324618501120.0, + "grad_norm": 1.8938081018501278, + "language_loss": 0.85327303, + "learning_rate": 3.7694339479760647e-06, + "loss": 0.87594092, + "num_input_tokens_seen": 31570415, + "step": 1495, + "time_per_iteration": 2.621367931365967 + }, + { + "auxiliary_loss_clip": 0.01143386, + "auxiliary_loss_mlp": 0.01010508, + "balance_loss_clip": 1.03297031, + "balance_loss_mlp": 1.00795674, + "epoch": 0.17988336439608008, + "flos": 68161864815360.0, + "grad_norm": 0.7813530105864278, + "language_loss": 0.57324457, + "learning_rate": 3.769070712807476e-06, + "loss": 0.59478348, + "num_input_tokens_seen": 31632445, + "step": 1496, + "time_per_iteration": 4.250871896743774 + }, + { + "auxiliary_loss_clip": 0.01250618, + "auxiliary_loss_mlp": 0.01037358, + "balance_loss_clip": 1.054389, + "balance_loss_mlp": 1.02794051, + "epoch": 0.18000360728671919, + "flos": 21945047143680.0, + "grad_norm": 1.7800556697467853, + "language_loss": 0.78841126, + "learning_rate": 3.768707209272266e-06, + "loss": 0.81129098, + "num_input_tokens_seen": 31652575, + "step": 1497, + "time_per_iteration": 2.804029941558838 + }, + { + "auxiliary_loss_clip": 0.01229124, + "auxiliary_loss_mlp": 0.0103324, + "balance_loss_clip": 1.05657208, + "balance_loss_mlp": 1.02341735, + "epoch": 0.18012385017735827, + "flos": 18986330937600.0, + "grad_norm": 2.2102747389403112, + "language_loss": 0.76897615, + "learning_rate": 3.768343437425579e-06, + "loss": 0.79159987, + "num_input_tokens_seen": 31671145, + "step": 1498, + "time_per_iteration": 2.612928628921509 + }, + { + "auxiliary_loss_clip": 0.01261423, + "auxiliary_loss_mlp": 0.01033825, + "balance_loss_clip": 1.05042219, + "balance_loss_mlp": 1.0235554, + "epoch": 0.18024409306799735, + "flos": 19748103598080.0, + "grad_norm": 2.529548022212386, + "language_loss": 0.8568933, + "learning_rate": 3.7679793973225987e-06, + "loss": 0.87984586, + "num_input_tokens_seen": 31686955, + "step": 1499, + "time_per_iteration": 2.7863430976867676 + }, + { + "auxiliary_loss_clip": 0.01158371, + "auxiliary_loss_mlp": 0.01004811, + "balance_loss_clip": 1.02929366, + "balance_loss_mlp": 1.00234365, + "epoch": 0.18036433595863643, + "flos": 67227183060480.0, + "grad_norm": 0.8594523138091991, + "language_loss": 0.61610973, + "learning_rate": 3.767615089018549e-06, + "loss": 0.63774157, + "num_input_tokens_seen": 31749300, + "step": 1500, + "time_per_iteration": 4.204959869384766 + }, + { + "auxiliary_loss_clip": 0.01230411, + "auxiliary_loss_mlp": 0.01033834, + "balance_loss_clip": 1.05836189, + "balance_loss_mlp": 1.02368951, + "epoch": 0.18048457884927555, + "flos": 18181464935040.0, + "grad_norm": 2.0557559338696905, + "language_loss": 0.86270642, + "learning_rate": 3.7672505125686966e-06, + "loss": 0.8853488, + "num_input_tokens_seen": 31765665, + "step": 1501, + "time_per_iteration": 3.5932066440582275 + }, + { + "auxiliary_loss_clip": 0.01252345, + "auxiliary_loss_mlp": 0.01038733, + "balance_loss_clip": 1.05642271, + "balance_loss_mlp": 1.02787852, + "epoch": 0.18060482173991463, + "flos": 15813767111040.0, + "grad_norm": 3.5484925543880674, + "language_loss": 0.84664047, + "learning_rate": 3.7668856680283455e-06, + "loss": 0.8695513, + "num_input_tokens_seen": 31782690, + "step": 1502, + "time_per_iteration": 3.6366989612579346 + }, + { + "auxiliary_loss_clip": 0.01246276, + "auxiliary_loss_mlp": 0.01033574, + "balance_loss_clip": 1.06327772, + "balance_loss_mlp": 1.02327442, + "epoch": 0.1807250646305537, + "flos": 18587399512320.0, + "grad_norm": 1.8683033403481686, + "language_loss": 0.81951642, + "learning_rate": 3.7665205554528437e-06, + "loss": 0.84231496, + "num_input_tokens_seen": 31802045, + "step": 1503, + "time_per_iteration": 2.648536205291748 + }, + { + "auxiliary_loss_clip": 0.01240159, + "auxiliary_loss_mlp": 0.01035456, + "balance_loss_clip": 1.06181335, + "balance_loss_mlp": 1.02532959, + "epoch": 0.18084530752119282, + "flos": 23149131880320.0, + "grad_norm": 1.7219746379687564, + "language_loss": 0.74004316, + "learning_rate": 3.7661551748975782e-06, + "loss": 0.76279926, + "num_input_tokens_seen": 31820220, + "step": 1504, + "time_per_iteration": 2.693272113800049 + }, + { + "auxiliary_loss_clip": 0.01144586, + "auxiliary_loss_mlp": 0.01006422, + "balance_loss_clip": 1.03364825, + "balance_loss_mlp": 1.00410891, + "epoch": 0.1809655504118319, + "flos": 59803153568640.0, + "grad_norm": 0.8110294312390796, + "language_loss": 0.60417396, + "learning_rate": 3.7657895264179772e-06, + "loss": 0.62568408, + "num_input_tokens_seen": 31876195, + "step": 1505, + "time_per_iteration": 3.22245454788208 + }, + { + "auxiliary_loss_clip": 0.01228881, + "auxiliary_loss_mlp": 0.01037564, + "balance_loss_clip": 1.05904055, + "balance_loss_mlp": 1.02709746, + "epoch": 0.181085793302471, + "flos": 44201941188480.0, + "grad_norm": 1.7105363837896668, + "language_loss": 0.74479294, + "learning_rate": 3.765423610069509e-06, + "loss": 0.76745737, + "num_input_tokens_seen": 31901585, + "step": 1506, + "time_per_iteration": 2.8635640144348145 + }, + { + "auxiliary_loss_clip": 0.01241914, + "auxiliary_loss_mlp": 0.01034098, + "balance_loss_clip": 1.06341493, + "balance_loss_mlp": 1.02369714, + "epoch": 0.18120603619311007, + "flos": 34898384638080.0, + "grad_norm": 2.184996946815326, + "language_loss": 0.72603703, + "learning_rate": 3.765057425907683e-06, + "loss": 0.74879718, + "num_input_tokens_seen": 31923045, + "step": 1507, + "time_per_iteration": 2.826634645462036 + }, + { + "auxiliary_loss_clip": 0.0123651, + "auxiliary_loss_mlp": 0.01037617, + "balance_loss_clip": 1.06206703, + "balance_loss_mlp": 1.02716279, + "epoch": 0.18132627908374918, + "flos": 21506757390720.0, + "grad_norm": 4.495039806325034, + "language_loss": 0.78397369, + "learning_rate": 3.764690973988048e-06, + "loss": 0.80671501, + "num_input_tokens_seen": 31943385, + "step": 1508, + "time_per_iteration": 2.6281206607818604 + }, + { + "auxiliary_loss_clip": 0.01253198, + "auxiliary_loss_mlp": 0.01036857, + "balance_loss_clip": 1.06025374, + "balance_loss_mlp": 1.02687311, + "epoch": 0.18144652197438826, + "flos": 29057693633280.0, + "grad_norm": 1.878098371971028, + "language_loss": 0.73983294, + "learning_rate": 3.7643242543661967e-06, + "loss": 0.76273358, + "num_input_tokens_seen": 31966045, + "step": 1509, + "time_per_iteration": 2.799891471862793 + }, + { + "auxiliary_loss_clip": 0.01134076, + "auxiliary_loss_mlp": 0.01006522, + "balance_loss_clip": 1.02928543, + "balance_loss_mlp": 1.00428057, + "epoch": 0.18156676486502735, + "flos": 68675064382080.0, + "grad_norm": 0.8216727313815354, + "language_loss": 0.60531425, + "learning_rate": 3.7639572670977573e-06, + "loss": 0.62672019, + "num_input_tokens_seen": 32021540, + "step": 1510, + "time_per_iteration": 3.1273887157440186 + }, + { + "auxiliary_loss_clip": 0.01253283, + "auxiliary_loss_mlp": 0.010402, + "balance_loss_clip": 1.05989075, + "balance_loss_mlp": 1.02983451, + "epoch": 0.18168700775566646, + "flos": 26471515334400.0, + "grad_norm": 2.1383693271704827, + "language_loss": 0.76429474, + "learning_rate": 3.7635900122384042e-06, + "loss": 0.7872296, + "num_input_tokens_seen": 32044535, + "step": 1511, + "time_per_iteration": 2.793210983276367 + }, + { + "auxiliary_loss_clip": 0.01242225, + "auxiliary_loss_mlp": 0.01038103, + "balance_loss_clip": 1.06089973, + "balance_loss_mlp": 1.02715397, + "epoch": 0.18180725064630554, + "flos": 15005668884480.0, + "grad_norm": 2.663470290051152, + "language_loss": 0.86782563, + "learning_rate": 3.7632224898438477e-06, + "loss": 0.89062881, + "num_input_tokens_seen": 32061010, + "step": 1512, + "time_per_iteration": 2.756561279296875 + }, + { + "auxiliary_loss_clip": 0.01250762, + "auxiliary_loss_mlp": 0.01034955, + "balance_loss_clip": 1.05835152, + "balance_loss_mlp": 1.02498925, + "epoch": 0.18192749353694462, + "flos": 19682387665920.0, + "grad_norm": 1.4874570686036255, + "language_loss": 0.7926712, + "learning_rate": 3.762854699969842e-06, + "loss": 0.81552833, + "num_input_tokens_seen": 32081520, + "step": 1513, + "time_per_iteration": 2.741952896118164 + }, + { + "auxiliary_loss_clip": 0.01232868, + "auxiliary_loss_mlp": 0.01037777, + "balance_loss_clip": 1.06555974, + "balance_loss_mlp": 1.02753735, + "epoch": 0.1820477364275837, + "flos": 20702717400960.0, + "grad_norm": 2.486189631065677, + "language_loss": 0.73164165, + "learning_rate": 3.762486642672179e-06, + "loss": 0.75434804, + "num_input_tokens_seen": 32098460, + "step": 1514, + "time_per_iteration": 2.6635806560516357 + }, + { + "auxiliary_loss_clip": 0.01241804, + "auxiliary_loss_mlp": 0.01038591, + "balance_loss_clip": 1.06139052, + "balance_loss_mlp": 1.02873278, + "epoch": 0.18216797931822282, + "flos": 17128708197120.0, + "grad_norm": 1.7647960132778298, + "language_loss": 0.87188941, + "learning_rate": 3.7621183180066946e-06, + "loss": 0.89469337, + "num_input_tokens_seen": 32116420, + "step": 1515, + "time_per_iteration": 2.670232057571411 + }, + { + "auxiliary_loss_clip": 0.01236031, + "auxiliary_loss_mlp": 0.01033503, + "balance_loss_clip": 1.05893922, + "balance_loss_mlp": 1.02390695, + "epoch": 0.1822882222088619, + "flos": 29242561956480.0, + "grad_norm": 1.5608568732301005, + "language_loss": 0.73698443, + "learning_rate": 3.7617497260292625e-06, + "loss": 0.75967979, + "num_input_tokens_seen": 32138475, + "step": 1516, + "time_per_iteration": 2.748472213745117 + }, + { + "auxiliary_loss_clip": 0.01237335, + "auxiliary_loss_mlp": 0.01039435, + "balance_loss_clip": 1.06247926, + "balance_loss_mlp": 1.02808642, + "epoch": 0.18240846509950098, + "flos": 17702739446400.0, + "grad_norm": 2.7199006811599067, + "language_loss": 0.78652644, + "learning_rate": 3.7613808667957967e-06, + "loss": 0.8092941, + "num_input_tokens_seen": 32151165, + "step": 1517, + "time_per_iteration": 2.6308646202087402 + }, + { + "auxiliary_loss_clip": 0.0124034, + "auxiliary_loss_mlp": 0.01039784, + "balance_loss_clip": 1.05958724, + "balance_loss_mlp": 1.02929366, + "epoch": 0.1825287079901401, + "flos": 14790025584000.0, + "grad_norm": 2.186410531030179, + "language_loss": 0.91472995, + "learning_rate": 3.7610117403622547e-06, + "loss": 0.93753117, + "num_input_tokens_seen": 32167725, + "step": 1518, + "time_per_iteration": 2.7311999797821045 + }, + { + "auxiliary_loss_clip": 0.01239775, + "auxiliary_loss_mlp": 0.01036555, + "balance_loss_clip": 1.05603004, + "balance_loss_mlp": 1.02644038, + "epoch": 0.18264895088077918, + "flos": 21946232292480.0, + "grad_norm": 1.6081785999283438, + "language_loss": 0.90107095, + "learning_rate": 3.7606423467846313e-06, + "loss": 0.9238342, + "num_input_tokens_seen": 32187330, + "step": 1519, + "time_per_iteration": 2.7110788822174072 + }, + { + "auxiliary_loss_clip": 0.01255075, + "auxiliary_loss_mlp": 0.01050625, + "balance_loss_clip": 1.06190383, + "balance_loss_mlp": 1.03983641, + "epoch": 0.18276919377141826, + "flos": 20886759711360.0, + "grad_norm": 1.844780207231808, + "language_loss": 0.79282796, + "learning_rate": 3.760272686118964e-06, + "loss": 0.81588501, + "num_input_tokens_seen": 32205550, + "step": 1520, + "time_per_iteration": 2.7002413272857666 + }, + { + "auxiliary_loss_clip": 0.01245566, + "auxiliary_loss_mlp": 0.01041432, + "balance_loss_clip": 1.06457198, + "balance_loss_mlp": 1.03066754, + "epoch": 0.18288943666205737, + "flos": 21469877101440.0, + "grad_norm": 2.1563574540158297, + "language_loss": 0.92312133, + "learning_rate": 3.7599027584213297e-06, + "loss": 0.94599128, + "num_input_tokens_seen": 32224430, + "step": 1521, + "time_per_iteration": 2.6884891986846924 + }, + { + "auxiliary_loss_clip": 0.01238202, + "auxiliary_loss_mlp": 0.01040764, + "balance_loss_clip": 1.06398654, + "balance_loss_mlp": 1.02906978, + "epoch": 0.18300967955269645, + "flos": 21539363961600.0, + "grad_norm": 1.8851679687519547, + "language_loss": 0.77901936, + "learning_rate": 3.7595325637478465e-06, + "loss": 0.80180907, + "num_input_tokens_seen": 32242455, + "step": 1522, + "time_per_iteration": 3.61053729057312 + }, + { + "auxiliary_loss_clip": 0.01231968, + "auxiliary_loss_mlp": 0.010431, + "balance_loss_clip": 1.06227541, + "balance_loss_mlp": 1.03344393, + "epoch": 0.18312992244333554, + "flos": 28876237102080.0, + "grad_norm": 1.6389246975745138, + "language_loss": 0.81539536, + "learning_rate": 3.7591621021546723e-06, + "loss": 0.83814609, + "num_input_tokens_seen": 32264450, + "step": 1523, + "time_per_iteration": 2.7203593254089355 + }, + { + "auxiliary_loss_clip": 0.01226014, + "auxiliary_loss_mlp": 0.01029744, + "balance_loss_clip": 1.06036115, + "balance_loss_mlp": 1.01968277, + "epoch": 0.18325016533397462, + "flos": 20120102801280.0, + "grad_norm": 1.7519316457967986, + "language_loss": 0.81547242, + "learning_rate": 3.7587913736980062e-06, + "loss": 0.83803004, + "num_input_tokens_seen": 32284090, + "step": 1524, + "time_per_iteration": 2.71160626411438 + }, + { + "auxiliary_loss_clip": 0.0124075, + "auxiliary_loss_mlp": 0.01040732, + "balance_loss_clip": 1.04999983, + "balance_loss_mlp": 1.03069472, + "epoch": 0.18337040822461373, + "flos": 23329187781120.0, + "grad_norm": 1.7221539423693542, + "language_loss": 0.84377813, + "learning_rate": 3.7584203784340865e-06, + "loss": 0.866593, + "num_input_tokens_seen": 32303260, + "step": 1525, + "time_per_iteration": 3.7030320167541504 + }, + { + "auxiliary_loss_clip": 0.01237969, + "auxiliary_loss_mlp": 0.01045246, + "balance_loss_clip": 1.0598712, + "balance_loss_mlp": 1.03442752, + "epoch": 0.1834906511152528, + "flos": 25009555881600.0, + "grad_norm": 2.0258546758659786, + "language_loss": 0.85779864, + "learning_rate": 3.7580491164191938e-06, + "loss": 0.88063073, + "num_input_tokens_seen": 32321570, + "step": 1526, + "time_per_iteration": 3.6088404655456543 + }, + { + "auxiliary_loss_clip": 0.0113365, + "auxiliary_loss_mlp": 0.01003393, + "balance_loss_clip": 1.03243804, + "balance_loss_mlp": 1.00112844, + "epoch": 0.1836108940058919, + "flos": 67251493589760.0, + "grad_norm": 0.7477688853056712, + "language_loss": 0.61250985, + "learning_rate": 3.757677587709648e-06, + "loss": 0.63388026, + "num_input_tokens_seen": 32384835, + "step": 1527, + "time_per_iteration": 4.25897741317749 + }, + { + "auxiliary_loss_clip": 0.01244267, + "auxiliary_loss_mlp": 0.01035171, + "balance_loss_clip": 1.05957186, + "balance_loss_mlp": 1.02453804, + "epoch": 0.183731136896531, + "flos": 25738721971200.0, + "grad_norm": 1.8368790343975971, + "language_loss": 0.75625664, + "learning_rate": 3.7573057923618095e-06, + "loss": 0.77905107, + "num_input_tokens_seen": 32404930, + "step": 1528, + "time_per_iteration": 2.7907354831695557 + }, + { + "auxiliary_loss_clip": 0.01252943, + "auxiliary_loss_mlp": 0.01035449, + "balance_loss_clip": 1.05436826, + "balance_loss_mlp": 1.02534008, + "epoch": 0.1838513797871701, + "flos": 20449403712000.0, + "grad_norm": 1.8393352202286672, + "language_loss": 0.74283552, + "learning_rate": 3.7569337304320793e-06, + "loss": 0.76571941, + "num_input_tokens_seen": 32424515, + "step": 1529, + "time_per_iteration": 2.751526355743408 + }, + { + "auxiliary_loss_clip": 0.01139133, + "auxiliary_loss_mlp": 0.01003651, + "balance_loss_clip": 1.02915883, + "balance_loss_mlp": 1.00112391, + "epoch": 0.18397162267780917, + "flos": 68565141786240.0, + "grad_norm": 0.8365888075574488, + "language_loss": 0.64496493, + "learning_rate": 3.756561401976899e-06, + "loss": 0.6663928, + "num_input_tokens_seen": 32484220, + "step": 1530, + "time_per_iteration": 3.119494676589966 + }, + { + "auxiliary_loss_clip": 0.01228498, + "auxiliary_loss_mlp": 0.01040137, + "balance_loss_clip": 1.06588149, + "balance_loss_mlp": 1.03049314, + "epoch": 0.18409186556844825, + "flos": 31941104976000.0, + "grad_norm": 2.071543279555242, + "language_loss": 0.82418978, + "learning_rate": 3.7561888070527514e-06, + "loss": 0.84687614, + "num_input_tokens_seen": 32506260, + "step": 1531, + "time_per_iteration": 2.695582628250122 + }, + { + "auxiliary_loss_clip": 0.0123098, + "auxiliary_loss_mlp": 0.0138086, + "balance_loss_clip": 1.05947685, + "balance_loss_mlp": 1.00014174, + "epoch": 0.18421210845908736, + "flos": 20120533764480.0, + "grad_norm": 1.9361731770330597, + "language_loss": 0.80045092, + "learning_rate": 3.7558159457161577e-06, + "loss": 0.82656932, + "num_input_tokens_seen": 32524225, + "step": 1532, + "time_per_iteration": 2.630397319793701 + }, + { + "auxiliary_loss_clip": 0.01241315, + "auxiliary_loss_mlp": 0.01380828, + "balance_loss_clip": 1.06154537, + "balance_loss_mlp": 1.00015843, + "epoch": 0.18433235134972645, + "flos": 23110491824640.0, + "grad_norm": 2.1855779974457175, + "language_loss": 0.77940798, + "learning_rate": 3.755442818023681e-06, + "loss": 0.80562937, + "num_input_tokens_seen": 32543850, + "step": 1533, + "time_per_iteration": 2.660860061645508 + }, + { + "auxiliary_loss_clip": 0.01247037, + "auxiliary_loss_mlp": 0.01030458, + "balance_loss_clip": 1.05942464, + "balance_loss_mlp": 1.02071309, + "epoch": 0.18445259424036553, + "flos": 18291351617280.0, + "grad_norm": 2.052812045911955, + "language_loss": 0.76256186, + "learning_rate": 3.7550694240319246e-06, + "loss": 0.78533685, + "num_input_tokens_seen": 32561725, + "step": 1534, + "time_per_iteration": 2.7437891960144043 + }, + { + "auxiliary_loss_clip": 0.01232772, + "auxiliary_loss_mlp": 0.01037347, + "balance_loss_clip": 1.05947578, + "balance_loss_mlp": 1.02651644, + "epoch": 0.18457283713100464, + "flos": 21324079797120.0, + "grad_norm": 2.6278356495932114, + "language_loss": 0.76190567, + "learning_rate": 3.7546957637975326e-06, + "loss": 0.78460681, + "num_input_tokens_seen": 32579135, + "step": 1535, + "time_per_iteration": 2.697728395462036 + }, + { + "auxiliary_loss_clip": 0.01248212, + "auxiliary_loss_mlp": 0.01044903, + "balance_loss_clip": 1.0490725, + "balance_loss_mlp": 1.03533065, + "epoch": 0.18469308002164372, + "flos": 20375679047040.0, + "grad_norm": 1.5882651502007188, + "language_loss": 0.7388593, + "learning_rate": 3.7543218373771873e-06, + "loss": 0.76179045, + "num_input_tokens_seen": 32598460, + "step": 1536, + "time_per_iteration": 2.7829174995422363 + }, + { + "auxiliary_loss_clip": 0.01249737, + "auxiliary_loss_mlp": 0.01380711, + "balance_loss_clip": 1.05202425, + "balance_loss_mlp": 1.0001334, + "epoch": 0.1848133229122828, + "flos": 26435892021120.0, + "grad_norm": 1.4341576244952257, + "language_loss": 0.78264755, + "learning_rate": 3.753947644827615e-06, + "loss": 0.80895203, + "num_input_tokens_seen": 32621920, + "step": 1537, + "time_per_iteration": 2.9628684520721436 + }, + { + "auxiliary_loss_clip": 0.01139898, + "auxiliary_loss_mlp": 0.01002548, + "balance_loss_clip": 1.02763534, + "balance_loss_mlp": 1.00015152, + "epoch": 0.1849335658029219, + "flos": 70547447612160.0, + "grad_norm": 0.9300129753849445, + "language_loss": 0.57200509, + "learning_rate": 3.753573186205579e-06, + "loss": 0.59342957, + "num_input_tokens_seen": 32690040, + "step": 1538, + "time_per_iteration": 3.410865545272827 + }, + { + "auxiliary_loss_clip": 0.01223592, + "auxiliary_loss_mlp": 0.0138055, + "balance_loss_clip": 1.05619311, + "balance_loss_mlp": 1.0001024, + "epoch": 0.185053808693561, + "flos": 17384140788480.0, + "grad_norm": 2.8381367570412275, + "language_loss": 0.77805454, + "learning_rate": 3.753198461567885e-06, + "loss": 0.80409592, + "num_input_tokens_seen": 32707285, + "step": 1539, + "time_per_iteration": 2.6557788848876953 + }, + { + "auxiliary_loss_clip": 0.01247065, + "auxiliary_loss_mlp": 0.01034305, + "balance_loss_clip": 1.0645988, + "balance_loss_mlp": 1.02494717, + "epoch": 0.18517405158420008, + "flos": 28986159697920.0, + "grad_norm": 1.6471058892978654, + "language_loss": 0.91583252, + "learning_rate": 3.7528234709713783e-06, + "loss": 0.93864626, + "num_input_tokens_seen": 32730030, + "step": 1540, + "time_per_iteration": 2.8525776863098145 + }, + { + "auxiliary_loss_clip": 0.01235094, + "auxiliary_loss_mlp": 0.01039572, + "balance_loss_clip": 1.0639255, + "balance_loss_mlp": 1.02881956, + "epoch": 0.18529429447483917, + "flos": 26794962328320.0, + "grad_norm": 1.804430051354442, + "language_loss": 0.83995759, + "learning_rate": 3.7524482144729447e-06, + "loss": 0.86270428, + "num_input_tokens_seen": 32749485, + "step": 1541, + "time_per_iteration": 2.704251766204834 + }, + { + "auxiliary_loss_clip": 0.01239124, + "auxiliary_loss_mlp": 0.01037439, + "balance_loss_clip": 1.05579662, + "balance_loss_mlp": 1.02700174, + "epoch": 0.18541453736547828, + "flos": 13581595301760.0, + "grad_norm": 2.169183654163221, + "language_loss": 0.83876014, + "learning_rate": 3.7520726921295106e-06, + "loss": 0.86152577, + "num_input_tokens_seen": 32766205, + "step": 1542, + "time_per_iteration": 2.7559444904327393 + }, + { + "auxiliary_loss_clip": 0.01222524, + "auxiliary_loss_mlp": 0.01037834, + "balance_loss_clip": 1.05814743, + "balance_loss_mlp": 1.02727175, + "epoch": 0.18553478025611736, + "flos": 24025424077440.0, + "grad_norm": 2.67884623373119, + "language_loss": 0.72442627, + "learning_rate": 3.751696903998042e-06, + "loss": 0.74702984, + "num_input_tokens_seen": 32784840, + "step": 1543, + "time_per_iteration": 2.667715311050415 + }, + { + "auxiliary_loss_clip": 0.01230993, + "auxiliary_loss_mlp": 0.01039218, + "balance_loss_clip": 1.06555605, + "balance_loss_mlp": 1.02978849, + "epoch": 0.18565502314675644, + "flos": 25885165720320.0, + "grad_norm": 1.599994283681934, + "language_loss": 0.69700706, + "learning_rate": 3.7513208501355456e-06, + "loss": 0.71970916, + "num_input_tokens_seen": 32805945, + "step": 1544, + "time_per_iteration": 2.656189203262329 + }, + { + "auxiliary_loss_clip": 0.01235965, + "auxiliary_loss_mlp": 0.01031715, + "balance_loss_clip": 1.06036258, + "balance_loss_mlp": 1.02180886, + "epoch": 0.18577526603739553, + "flos": 19610063631360.0, + "grad_norm": 2.570427013973831, + "language_loss": 0.83575654, + "learning_rate": 3.750944530599069e-06, + "loss": 0.85843337, + "num_input_tokens_seen": 32825515, + "step": 1545, + "time_per_iteration": 2.7282843589782715 + }, + { + "auxiliary_loss_clip": 0.01239353, + "auxiliary_loss_mlp": 0.01042702, + "balance_loss_clip": 1.06434727, + "balance_loss_mlp": 1.03214574, + "epoch": 0.18589550892803464, + "flos": 18474891137280.0, + "grad_norm": 1.8835830650940744, + "language_loss": 0.80853295, + "learning_rate": 3.7505679454456992e-06, + "loss": 0.83135349, + "num_input_tokens_seen": 32842125, + "step": 1546, + "time_per_iteration": 2.6076982021331787 + }, + { + "auxiliary_loss_clip": 0.01265205, + "auxiliary_loss_mlp": 0.01034046, + "balance_loss_clip": 1.05343235, + "balance_loss_mlp": 1.0241276, + "epoch": 0.18601575181867372, + "flos": 23549966726400.0, + "grad_norm": 2.1889852775260006, + "language_loss": 0.7016685, + "learning_rate": 3.750191094732564e-06, + "loss": 0.72466099, + "num_input_tokens_seen": 32862990, + "step": 1547, + "time_per_iteration": 2.8435769081115723 + }, + { + "auxiliary_loss_clip": 0.01260703, + "auxiliary_loss_mlp": 0.01381031, + "balance_loss_clip": 1.05179763, + "balance_loss_mlp": 1.00011659, + "epoch": 0.1861359947093128, + "flos": 26360192108160.0, + "grad_norm": 1.779606464410969, + "language_loss": 0.75432038, + "learning_rate": 3.7498139785168313e-06, + "loss": 0.78073776, + "num_input_tokens_seen": 32883595, + "step": 1548, + "time_per_iteration": 3.891428232192993 + }, + { + "auxiliary_loss_clip": 0.01228884, + "auxiliary_loss_mlp": 0.0103338, + "balance_loss_clip": 1.06164718, + "balance_loss_mlp": 1.02370656, + "epoch": 0.1862562375999519, + "flos": 23331198942720.0, + "grad_norm": 9.988025329831252, + "language_loss": 0.77624393, + "learning_rate": 3.749436596855709e-06, + "loss": 0.79886657, + "num_input_tokens_seen": 32902895, + "step": 1549, + "time_per_iteration": 2.7718117237091064 + }, + { + "auxiliary_loss_clip": 0.0122528, + "auxiliary_loss_mlp": 0.0103616, + "balance_loss_clip": 1.06116521, + "balance_loss_mlp": 1.02580118, + "epoch": 0.186376480490591, + "flos": 16648222942080.0, + "grad_norm": 1.8399956172015528, + "language_loss": 0.90346992, + "learning_rate": 3.749058949806446e-06, + "loss": 0.92608428, + "num_input_tokens_seen": 32919620, + "step": 1550, + "time_per_iteration": 2.6452488899230957 + }, + { + "auxiliary_loss_clip": 0.01234415, + "auxiliary_loss_mlp": 0.01032926, + "balance_loss_clip": 1.06361008, + "balance_loss_mlp": 1.0221014, + "epoch": 0.18649672338123008, + "flos": 21468656039040.0, + "grad_norm": 1.8985218294566872, + "language_loss": 0.84394276, + "learning_rate": 3.748681037426331e-06, + "loss": 0.86661625, + "num_input_tokens_seen": 32938830, + "step": 1551, + "time_per_iteration": 3.517639636993408 + }, + { + "auxiliary_loss_clip": 0.01225788, + "auxiliary_loss_mlp": 0.01040519, + "balance_loss_clip": 1.06547618, + "balance_loss_mlp": 1.03055894, + "epoch": 0.1866169662718692, + "flos": 12312728386560.0, + "grad_norm": 2.164330873669751, + "language_loss": 0.91904366, + "learning_rate": 3.7483028597726936e-06, + "loss": 0.94170666, + "num_input_tokens_seen": 32955600, + "step": 1552, + "time_per_iteration": 3.677345037460327 + }, + { + "auxiliary_loss_clip": 0.01247598, + "auxiliary_loss_mlp": 0.01040954, + "balance_loss_clip": 1.06055999, + "balance_loss_mlp": 1.02927125, + "epoch": 0.18673720916250827, + "flos": 23581280407680.0, + "grad_norm": 1.6875751814368474, + "language_loss": 0.62581658, + "learning_rate": 3.7479244169029017e-06, + "loss": 0.64870203, + "num_input_tokens_seen": 32975390, + "step": 1553, + "time_per_iteration": 2.733825206756592 + }, + { + "auxiliary_loss_clip": 0.01235395, + "auxiliary_loss_mlp": 0.01031206, + "balance_loss_clip": 1.06111991, + "balance_loss_mlp": 1.02131772, + "epoch": 0.18685745205314735, + "flos": 19718370115200.0, + "grad_norm": 2.2189480610920045, + "language_loss": 0.7368722, + "learning_rate": 3.7475457088743658e-06, + "loss": 0.75953817, + "num_input_tokens_seen": 32992640, + "step": 1554, + "time_per_iteration": 3.484128952026367 + }, + { + "auxiliary_loss_clip": 0.01229353, + "auxiliary_loss_mlp": 0.01038422, + "balance_loss_clip": 1.06018567, + "balance_loss_mlp": 1.02871895, + "epoch": 0.18697769494378644, + "flos": 34204123589760.0, + "grad_norm": 1.8864647715748608, + "language_loss": 0.74334657, + "learning_rate": 3.7471667357445348e-06, + "loss": 0.76602429, + "num_input_tokens_seen": 33012470, + "step": 1555, + "time_per_iteration": 2.791921615600586 + }, + { + "auxiliary_loss_clip": 0.01262785, + "auxiliary_loss_mlp": 0.01038612, + "balance_loss_clip": 1.05501246, + "balance_loss_mlp": 1.02812123, + "epoch": 0.18709793783442555, + "flos": 34241327101440.0, + "grad_norm": 3.589060153219514, + "language_loss": 0.72253346, + "learning_rate": 3.7467874975709e-06, + "loss": 0.74554741, + "num_input_tokens_seen": 33033275, + "step": 1556, + "time_per_iteration": 2.8365585803985596 + }, + { + "auxiliary_loss_clip": 0.01236427, + "auxiliary_loss_mlp": 0.01031229, + "balance_loss_clip": 1.06488788, + "balance_loss_mlp": 1.02132833, + "epoch": 0.18721818072506463, + "flos": 40734550529280.0, + "grad_norm": 2.162001133641721, + "language_loss": 0.77940285, + "learning_rate": 3.7464079944109904e-06, + "loss": 0.80207944, + "num_input_tokens_seen": 33055135, + "step": 1557, + "time_per_iteration": 2.8184053897857666 + }, + { + "auxiliary_loss_clip": 0.01247146, + "auxiliary_loss_mlp": 0.01035858, + "balance_loss_clip": 1.0573293, + "balance_loss_mlp": 1.02644634, + "epoch": 0.18733842361570371, + "flos": 22157386392960.0, + "grad_norm": 1.875635058296114, + "language_loss": 0.77732003, + "learning_rate": 3.746028226322376e-06, + "loss": 0.8001501, + "num_input_tokens_seen": 33071015, + "step": 1558, + "time_per_iteration": 2.6540541648864746 + }, + { + "auxiliary_loss_clip": 0.01236065, + "auxiliary_loss_mlp": 0.01040212, + "balance_loss_clip": 1.0594877, + "balance_loss_mlp": 1.03062749, + "epoch": 0.18745866650634282, + "flos": 18914940656640.0, + "grad_norm": 1.760424286851501, + "language_loss": 0.75557977, + "learning_rate": 3.745648193362669e-06, + "loss": 0.77834255, + "num_input_tokens_seen": 33090370, + "step": 1559, + "time_per_iteration": 2.647026538848877 + }, + { + "auxiliary_loss_clip": 0.01239355, + "auxiliary_loss_mlp": 0.01030733, + "balance_loss_clip": 1.06156111, + "balance_loss_mlp": 1.02139258, + "epoch": 0.1875789093969819, + "flos": 19314626267520.0, + "grad_norm": 1.8796652844728379, + "language_loss": 0.72394156, + "learning_rate": 3.745267895589518e-06, + "loss": 0.74664247, + "num_input_tokens_seen": 33108910, + "step": 1560, + "time_per_iteration": 2.715069532394409 + }, + { + "auxiliary_loss_clip": 0.01238648, + "auxiliary_loss_mlp": 0.01034553, + "balance_loss_clip": 1.06026554, + "balance_loss_mlp": 1.0246172, + "epoch": 0.187699152287621, + "flos": 17018965169280.0, + "grad_norm": 4.115178007064644, + "language_loss": 0.82359517, + "learning_rate": 3.7448873330606154e-06, + "loss": 0.84632713, + "num_input_tokens_seen": 33126680, + "step": 1561, + "time_per_iteration": 2.657895803451538 + }, + { + "auxiliary_loss_clip": 0.01237906, + "auxiliary_loss_mlp": 0.01043229, + "balance_loss_clip": 1.0599575, + "balance_loss_mlp": 1.03189802, + "epoch": 0.18781939517826007, + "flos": 22346384780160.0, + "grad_norm": 1.9581753930847612, + "language_loss": 0.87640905, + "learning_rate": 3.7445065058336914e-06, + "loss": 0.89922035, + "num_input_tokens_seen": 33145550, + "step": 1562, + "time_per_iteration": 2.7121682167053223 + }, + { + "auxiliary_loss_clip": 0.01234725, + "auxiliary_loss_mlp": 0.0103476, + "balance_loss_clip": 1.04904163, + "balance_loss_mlp": 1.02488351, + "epoch": 0.18793963806889918, + "flos": 14611478054400.0, + "grad_norm": 2.406185649785801, + "language_loss": 0.86539507, + "learning_rate": 3.7441254139665176e-06, + "loss": 0.8880899, + "num_input_tokens_seen": 33161735, + "step": 1563, + "time_per_iteration": 2.701606512069702 + }, + { + "auxiliary_loss_clip": 0.01227733, + "auxiliary_loss_mlp": 0.01032195, + "balance_loss_clip": 1.06880045, + "balance_loss_mlp": 1.02313554, + "epoch": 0.18805988095953827, + "flos": 17457075354240.0, + "grad_norm": 5.402973731480193, + "language_loss": 0.82498729, + "learning_rate": 3.743744057516905e-06, + "loss": 0.84758657, + "num_input_tokens_seen": 33179795, + "step": 1564, + "time_per_iteration": 2.602174997329712 + }, + { + "auxiliary_loss_clip": 0.01251342, + "auxiliary_loss_mlp": 0.01038124, + "balance_loss_clip": 1.0560534, + "balance_loss_mlp": 1.02742469, + "epoch": 0.18818012385017735, + "flos": 15043877976960.0, + "grad_norm": 2.8739805934949763, + "language_loss": 0.8785255, + "learning_rate": 3.743362436542706e-06, + "loss": 0.90142018, + "num_input_tokens_seen": 33194485, + "step": 1565, + "time_per_iteration": 2.7005839347839355 + }, + { + "auxiliary_loss_clip": 0.01221426, + "auxiliary_loss_mlp": 0.01035919, + "balance_loss_clip": 1.06322527, + "balance_loss_mlp": 1.02644825, + "epoch": 0.18830036674081646, + "flos": 47551975136640.0, + "grad_norm": 1.771854317003793, + "language_loss": 0.77034795, + "learning_rate": 3.7429805511018115e-06, + "loss": 0.79292136, + "num_input_tokens_seen": 33216145, + "step": 1566, + "time_per_iteration": 2.858099937438965 + }, + { + "auxiliary_loss_clip": 0.01239207, + "auxiliary_loss_mlp": 0.01381133, + "balance_loss_clip": 1.05703449, + "balance_loss_mlp": 1.00011778, + "epoch": 0.18842060963145554, + "flos": 30044626698240.0, + "grad_norm": 2.021225080552375, + "language_loss": 0.78475136, + "learning_rate": 3.7425984012521524e-06, + "loss": 0.81095481, + "num_input_tokens_seen": 33236345, + "step": 1567, + "time_per_iteration": 2.806978702545166 + }, + { + "auxiliary_loss_clip": 0.01147247, + "auxiliary_loss_mlp": 0.01376575, + "balance_loss_clip": 1.02974749, + "balance_loss_mlp": 0.99999046, + "epoch": 0.18854085252209463, + "flos": 70318372625280.0, + "grad_norm": 1.013759990508112, + "language_loss": 0.60420382, + "learning_rate": 3.7422159870517025e-06, + "loss": 0.62944204, + "num_input_tokens_seen": 33301600, + "step": 1568, + "time_per_iteration": 3.3003761768341064 + }, + { + "auxiliary_loss_clip": 0.01231458, + "auxiliary_loss_mlp": 0.01027646, + "balance_loss_clip": 1.05788183, + "balance_loss_mlp": 1.01878285, + "epoch": 0.1886610954127337, + "flos": 21289318410240.0, + "grad_norm": 1.4614899819729195, + "language_loss": 0.78842622, + "learning_rate": 3.7418333085584717e-06, + "loss": 0.81101727, + "num_input_tokens_seen": 33322785, + "step": 1569, + "time_per_iteration": 2.7576382160186768 + }, + { + "auxiliary_loss_clip": 0.01248367, + "auxiliary_loss_mlp": 0.01048616, + "balance_loss_clip": 1.06187582, + "balance_loss_mlp": 1.03745818, + "epoch": 0.18878133830337282, + "flos": 17266819991040.0, + "grad_norm": 2.2099259388873778, + "language_loss": 0.90830868, + "learning_rate": 3.7414503658305128e-06, + "loss": 0.93127847, + "num_input_tokens_seen": 33340020, + "step": 1570, + "time_per_iteration": 2.7090225219726562 + }, + { + "auxiliary_loss_clip": 0.01255757, + "auxiliary_loss_mlp": 0.01037171, + "balance_loss_clip": 1.05336237, + "balance_loss_mlp": 1.02736652, + "epoch": 0.1889015811940119, + "flos": 25775207210880.0, + "grad_norm": 2.1557938045811738, + "language_loss": 0.77622455, + "learning_rate": 3.7410671589259185e-06, + "loss": 0.79915386, + "num_input_tokens_seen": 33358620, + "step": 1571, + "time_per_iteration": 2.797093629837036 + }, + { + "auxiliary_loss_clip": 0.01223552, + "auxiliary_loss_mlp": 0.01043806, + "balance_loss_clip": 1.06505609, + "balance_loss_mlp": 1.03406644, + "epoch": 0.18902182408465099, + "flos": 21032197879680.0, + "grad_norm": 1.7521949903689162, + "language_loss": 0.79800916, + "learning_rate": 3.7406836879028205e-06, + "loss": 0.82068276, + "num_input_tokens_seen": 33378845, + "step": 1572, + "time_per_iteration": 2.6049511432647705 + }, + { + "auxiliary_loss_clip": 0.01228467, + "auxiliary_loss_mlp": 0.01039512, + "balance_loss_clip": 1.06330633, + "balance_loss_mlp": 1.02998745, + "epoch": 0.1891420669752901, + "flos": 22272121411200.0, + "grad_norm": 1.9386455620745162, + "language_loss": 0.7641921, + "learning_rate": 3.7402999528193907e-06, + "loss": 0.78687191, + "num_input_tokens_seen": 33398345, + "step": 1573, + "time_per_iteration": 2.702700138092041 + }, + { + "auxiliary_loss_clip": 0.01228192, + "auxiliary_loss_mlp": 0.01380816, + "balance_loss_clip": 1.05811238, + "balance_loss_mlp": 1.00012875, + "epoch": 0.18926230986592918, + "flos": 22017802141440.0, + "grad_norm": 3.5939036105657682, + "language_loss": 0.85342526, + "learning_rate": 3.739915953733842e-06, + "loss": 0.87951529, + "num_input_tokens_seen": 33416390, + "step": 1574, + "time_per_iteration": 3.6618852615356445 + }, + { + "auxiliary_loss_clip": 0.01219076, + "auxiliary_loss_mlp": 0.01039265, + "balance_loss_clip": 1.06312203, + "balance_loss_mlp": 1.02884054, + "epoch": 0.18938255275656826, + "flos": 24462672336000.0, + "grad_norm": 1.605006974160165, + "language_loss": 0.82201594, + "learning_rate": 3.7395316907044264e-06, + "loss": 0.84459937, + "num_input_tokens_seen": 33437175, + "step": 1575, + "time_per_iteration": 2.66750168800354 + }, + { + "auxiliary_loss_clip": 0.01230902, + "auxiliary_loss_mlp": 0.01035901, + "balance_loss_clip": 1.06254625, + "balance_loss_mlp": 1.02546406, + "epoch": 0.18950279564720737, + "flos": 24427049022720.0, + "grad_norm": 1.5711006513198251, + "language_loss": 0.7950114, + "learning_rate": 3.7391471637894364e-06, + "loss": 0.81767941, + "num_input_tokens_seen": 33459440, + "step": 1576, + "time_per_iteration": 2.7070839405059814 + }, + { + "auxiliary_loss_clip": 0.01243141, + "auxiliary_loss_mlp": 0.01037243, + "balance_loss_clip": 1.05465126, + "balance_loss_mlp": 1.02781963, + "epoch": 0.18962303853784646, + "flos": 19756291898880.0, + "grad_norm": 1.7548075972352148, + "language_loss": 0.84609675, + "learning_rate": 3.738762373047205e-06, + "loss": 0.8689006, + "num_input_tokens_seen": 33479360, + "step": 1577, + "time_per_iteration": 3.6335361003875732 + }, + { + "auxiliary_loss_clip": 0.0124829, + "auxiliary_loss_mlp": 0.01040628, + "balance_loss_clip": 1.05992913, + "balance_loss_mlp": 1.03036404, + "epoch": 0.18974328142848554, + "flos": 21032054225280.0, + "grad_norm": 2.037571460502913, + "language_loss": 0.83591807, + "learning_rate": 3.738377318536103e-06, + "loss": 0.85880733, + "num_input_tokens_seen": 33499245, + "step": 1578, + "time_per_iteration": 3.6407015323638916 + }, + { + "auxiliary_loss_clip": 0.01217879, + "auxiliary_loss_mlp": 0.01037502, + "balance_loss_clip": 1.06320333, + "balance_loss_mlp": 1.02809048, + "epoch": 0.18986352431912462, + "flos": 12966122736000.0, + "grad_norm": 2.7462925665077065, + "language_loss": 0.71193534, + "learning_rate": 3.7379920003145447e-06, + "loss": 0.7344892, + "num_input_tokens_seen": 33513520, + "step": 1579, + "time_per_iteration": 3.4839580059051514 + }, + { + "auxiliary_loss_clip": 0.01228538, + "auxiliary_loss_mlp": 0.01035551, + "balance_loss_clip": 1.06062925, + "balance_loss_mlp": 1.02499485, + "epoch": 0.18998376720976373, + "flos": 23767908497280.0, + "grad_norm": 2.0312279773010564, + "language_loss": 0.83592618, + "learning_rate": 3.7376064184409817e-06, + "loss": 0.85856712, + "num_input_tokens_seen": 33533100, + "step": 1580, + "time_per_iteration": 2.6684727668762207 + }, + { + "auxiliary_loss_clip": 0.01232772, + "auxiliary_loss_mlp": 0.01041025, + "balance_loss_clip": 1.05968714, + "balance_loss_mlp": 1.03121972, + "epoch": 0.19010401010040281, + "flos": 22966023323520.0, + "grad_norm": 1.3969122591162533, + "language_loss": 0.87109905, + "learning_rate": 3.7372205729739063e-06, + "loss": 0.89383698, + "num_input_tokens_seen": 33554915, + "step": 1581, + "time_per_iteration": 2.7382593154907227 + }, + { + "auxiliary_loss_clip": 0.01232304, + "auxiliary_loss_mlp": 0.01034163, + "balance_loss_clip": 1.06250048, + "balance_loss_mlp": 1.02402413, + "epoch": 0.1902242529910419, + "flos": 19135647774720.0, + "grad_norm": 3.520973189757873, + "language_loss": 0.71851569, + "learning_rate": 3.7368344639718514e-06, + "loss": 0.74118036, + "num_input_tokens_seen": 33572850, + "step": 1582, + "time_per_iteration": 2.6034059524536133 + }, + { + "auxiliary_loss_clip": 0.01231837, + "auxiliary_loss_mlp": 0.01041204, + "balance_loss_clip": 1.06306458, + "balance_loss_mlp": 1.03166127, + "epoch": 0.190344495881681, + "flos": 25483935824640.0, + "grad_norm": 1.499473804627554, + "language_loss": 0.80430043, + "learning_rate": 3.7364480914933895e-06, + "loss": 0.82703078, + "num_input_tokens_seen": 33593090, + "step": 1583, + "time_per_iteration": 2.688896656036377 + }, + { + "auxiliary_loss_clip": 0.01247772, + "auxiliary_loss_mlp": 0.013806, + "balance_loss_clip": 1.05602908, + "balance_loss_mlp": 1.00007796, + "epoch": 0.1904647387723201, + "flos": 26792843425920.0, + "grad_norm": 1.8598651092509813, + "language_loss": 0.81130224, + "learning_rate": 3.7360614555971325e-06, + "loss": 0.83758593, + "num_input_tokens_seen": 33612745, + "step": 1584, + "time_per_iteration": 2.762960910797119 + }, + { + "auxiliary_loss_clip": 0.01231152, + "auxiliary_loss_mlp": 0.01380437, + "balance_loss_clip": 1.06372464, + "balance_loss_mlp": 1.00004363, + "epoch": 0.19058498166295917, + "flos": 23987753688960.0, + "grad_norm": 1.8715761315066528, + "language_loss": 0.84873903, + "learning_rate": 3.735674556341733e-06, + "loss": 0.87485492, + "num_input_tokens_seen": 33632360, + "step": 1585, + "time_per_iteration": 2.6568944454193115 + }, + { + "auxiliary_loss_clip": 0.01238715, + "auxiliary_loss_mlp": 0.01040305, + "balance_loss_clip": 1.06426454, + "balance_loss_mlp": 1.03040504, + "epoch": 0.19070522455359826, + "flos": 28293299280000.0, + "grad_norm": 1.8839643220024835, + "language_loss": 0.82790053, + "learning_rate": 3.7352873937858835e-06, + "loss": 0.85069072, + "num_input_tokens_seen": 33653895, + "step": 1586, + "time_per_iteration": 2.716492176055908 + }, + { + "auxiliary_loss_clip": 0.01235911, + "auxiliary_loss_mlp": 0.01380731, + "balance_loss_clip": 1.05634093, + "balance_loss_mlp": 1.00003386, + "epoch": 0.19082546744423737, + "flos": 25660220797440.0, + "grad_norm": 2.112192687902865, + "language_loss": 0.721398, + "learning_rate": 3.734899967988316e-06, + "loss": 0.74756444, + "num_input_tokens_seen": 33672075, + "step": 1587, + "time_per_iteration": 2.745380163192749 + }, + { + "auxiliary_loss_clip": 0.01234269, + "auxiliary_loss_mlp": 0.01034412, + "balance_loss_clip": 1.05606604, + "balance_loss_mlp": 1.02523899, + "epoch": 0.19094571033487645, + "flos": 19719483436800.0, + "grad_norm": 1.7818295065303604, + "language_loss": 0.83953118, + "learning_rate": 3.7345122790078026e-06, + "loss": 0.86221802, + "num_input_tokens_seen": 33689640, + "step": 1588, + "time_per_iteration": 2.7147083282470703 + }, + { + "auxiliary_loss_clip": 0.01228103, + "auxiliary_loss_mlp": 0.01037162, + "balance_loss_clip": 1.06136715, + "balance_loss_mlp": 1.02640939, + "epoch": 0.19106595322551553, + "flos": 21616320850560.0, + "grad_norm": 4.654835117496724, + "language_loss": 0.92550647, + "learning_rate": 3.7341243269031556e-06, + "loss": 0.9481591, + "num_input_tokens_seen": 33708630, + "step": 1589, + "time_per_iteration": 2.6361019611358643 + }, + { + "auxiliary_loss_clip": 0.01228883, + "auxiliary_loss_mlp": 0.01038979, + "balance_loss_clip": 1.06028259, + "balance_loss_mlp": 1.02976441, + "epoch": 0.19118619611615464, + "flos": 29896890059520.0, + "grad_norm": 1.690826990560581, + "language_loss": 0.77170444, + "learning_rate": 3.7337361117332275e-06, + "loss": 0.79438311, + "num_input_tokens_seen": 33730370, + "step": 1590, + "time_per_iteration": 2.7732415199279785 + }, + { + "auxiliary_loss_clip": 0.01245514, + "auxiliary_loss_mlp": 0.01035577, + "balance_loss_clip": 1.05507696, + "balance_loss_mlp": 1.02596283, + "epoch": 0.19130643900679373, + "flos": 17273428093440.0, + "grad_norm": 1.889326696195407, + "language_loss": 0.7709322, + "learning_rate": 3.7333476335569087e-06, + "loss": 0.79374313, + "num_input_tokens_seen": 33748370, + "step": 1591, + "time_per_iteration": 2.6905453205108643 + }, + { + "auxiliary_loss_clip": 0.01235055, + "auxiliary_loss_mlp": 0.01033172, + "balance_loss_clip": 1.05860984, + "balance_loss_mlp": 1.02330732, + "epoch": 0.1914266818974328, + "flos": 24826339584000.0, + "grad_norm": 2.2867094879553123, + "language_loss": 0.67181462, + "learning_rate": 3.7329588924331325e-06, + "loss": 0.69449687, + "num_input_tokens_seen": 33769575, + "step": 1592, + "time_per_iteration": 2.720346689224243 + }, + { + "auxiliary_loss_clip": 0.0123589, + "auxiliary_loss_mlp": 0.01030934, + "balance_loss_clip": 1.05510199, + "balance_loss_mlp": 1.02100348, + "epoch": 0.1915469247880719, + "flos": 18952467390720.0, + "grad_norm": 1.713287706824984, + "language_loss": 0.82616866, + "learning_rate": 3.732569888420871e-06, + "loss": 0.84883684, + "num_input_tokens_seen": 33789110, + "step": 1593, + "time_per_iteration": 2.7002265453338623 + }, + { + "auxiliary_loss_clip": 0.01221552, + "auxiliary_loss_mlp": 0.01032825, + "balance_loss_clip": 1.06126499, + "balance_loss_mlp": 1.02267456, + "epoch": 0.191667167678711, + "flos": 21032952065280.0, + "grad_norm": 2.2780469777594345, + "language_loss": 0.82194573, + "learning_rate": 3.732180621579134e-06, + "loss": 0.84448957, + "num_input_tokens_seen": 33808325, + "step": 1594, + "time_per_iteration": 2.6328988075256348 + }, + { + "auxiliary_loss_clip": 0.01256609, + "auxiliary_loss_mlp": 0.01036612, + "balance_loss_clip": 1.0626992, + "balance_loss_mlp": 1.02720594, + "epoch": 0.1917874105693501, + "flos": 34237663914240.0, + "grad_norm": 2.4703866651286654, + "language_loss": 0.80960268, + "learning_rate": 3.7317910919669745e-06, + "loss": 0.83253485, + "num_input_tokens_seen": 33829520, + "step": 1595, + "time_per_iteration": 2.8041324615478516 + }, + { + "auxiliary_loss_clip": 0.01227706, + "auxiliary_loss_mlp": 0.010353, + "balance_loss_clip": 1.06272256, + "balance_loss_mlp": 1.02522075, + "epoch": 0.19190765345998917, + "flos": 23550613171200.0, + "grad_norm": 2.809022588433715, + "language_loss": 0.75890398, + "learning_rate": 3.7314012996434826e-06, + "loss": 0.78153402, + "num_input_tokens_seen": 33848250, + "step": 1596, + "time_per_iteration": 2.7029097080230713 + }, + { + "auxiliary_loss_clip": 0.01236921, + "auxiliary_loss_mlp": 0.01033499, + "balance_loss_clip": 1.06143284, + "balance_loss_mlp": 1.0238018, + "epoch": 0.19202789635062828, + "flos": 19861330245120.0, + "grad_norm": 1.945911881685662, + "language_loss": 0.80905622, + "learning_rate": 3.7310112446677907e-06, + "loss": 0.83176047, + "num_input_tokens_seen": 33866160, + "step": 1597, + "time_per_iteration": 2.6590588092803955 + }, + { + "auxiliary_loss_clip": 0.01224723, + "auxiliary_loss_mlp": 0.0103361, + "balance_loss_clip": 1.06554532, + "balance_loss_mlp": 1.02407312, + "epoch": 0.19214813924126736, + "flos": 20922957642240.0, + "grad_norm": 1.9238078313647016, + "language_loss": 0.68694627, + "learning_rate": 3.7306209270990695e-06, + "loss": 0.70952964, + "num_input_tokens_seen": 33884165, + "step": 1598, + "time_per_iteration": 2.655240774154663 + }, + { + "auxiliary_loss_clip": 0.01238279, + "auxiliary_loss_mlp": 0.01037684, + "balance_loss_clip": 1.06076837, + "balance_loss_mlp": 1.02825463, + "epoch": 0.19226838213190645, + "flos": 26359725231360.0, + "grad_norm": 1.7664146670507825, + "language_loss": 0.86550415, + "learning_rate": 3.7302303469965292e-06, + "loss": 0.88826382, + "num_input_tokens_seen": 33903705, + "step": 1599, + "time_per_iteration": 2.7002055644989014 + }, + { + "auxiliary_loss_clip": 0.0122856, + "auxiliary_loss_mlp": 0.01035732, + "balance_loss_clip": 1.06305909, + "balance_loss_mlp": 1.02660048, + "epoch": 0.19238862502254553, + "flos": 20850525866880.0, + "grad_norm": 1.8607422335907147, + "language_loss": 0.70468175, + "learning_rate": 3.7298395044194206e-06, + "loss": 0.72732472, + "num_input_tokens_seen": 33922515, + "step": 1600, + "time_per_iteration": 3.603997230529785 + }, + { + "auxiliary_loss_clip": 0.01227779, + "auxiliary_loss_mlp": 0.01038325, + "balance_loss_clip": 1.06955671, + "balance_loss_mlp": 1.02881193, + "epoch": 0.19250886791318464, + "flos": 21726063878400.0, + "grad_norm": 1.8374679537990837, + "language_loss": 0.94349623, + "learning_rate": 3.7294483994270356e-06, + "loss": 0.96615726, + "num_input_tokens_seen": 33940840, + "step": 1601, + "time_per_iteration": 2.6222376823425293 + }, + { + "auxiliary_loss_clip": 0.01238129, + "auxiliary_loss_mlp": 0.01040066, + "balance_loss_clip": 1.0517658, + "balance_loss_mlp": 1.03042769, + "epoch": 0.19262911080382372, + "flos": 23367827836800.0, + "grad_norm": 2.2093181370881254, + "language_loss": 0.78073281, + "learning_rate": 3.7290570320787033e-06, + "loss": 0.80351478, + "num_input_tokens_seen": 33960420, + "step": 1602, + "time_per_iteration": 2.756331443786621 + }, + { + "auxiliary_loss_clip": 0.01227841, + "auxiliary_loss_mlp": 0.01033818, + "balance_loss_clip": 1.06274509, + "balance_loss_mlp": 1.02456152, + "epoch": 0.1927493536944628, + "flos": 21943502858880.0, + "grad_norm": 2.126508197418027, + "language_loss": 0.71572888, + "learning_rate": 3.728665402433793e-06, + "loss": 0.7383455, + "num_input_tokens_seen": 33978990, + "step": 1603, + "time_per_iteration": 3.612764835357666 + }, + { + "auxiliary_loss_clip": 0.01233949, + "auxiliary_loss_mlp": 0.01033376, + "balance_loss_clip": 1.05935705, + "balance_loss_mlp": 1.02402425, + "epoch": 0.19286959658510192, + "flos": 16545590807040.0, + "grad_norm": 2.526169084300282, + "language_loss": 0.8620317, + "learning_rate": 3.7282735105517164e-06, + "loss": 0.88470495, + "num_input_tokens_seen": 33997115, + "step": 1604, + "time_per_iteration": 2.7041923999786377 + }, + { + "auxiliary_loss_clip": 0.01258478, + "auxiliary_loss_mlp": 0.01040125, + "balance_loss_clip": 1.0554775, + "balance_loss_mlp": 1.0301528, + "epoch": 0.192989839475741, + "flos": 21616967295360.0, + "grad_norm": 2.057997999255952, + "language_loss": 0.67105454, + "learning_rate": 3.727881356491922e-06, + "loss": 0.69404066, + "num_input_tokens_seen": 34015525, + "step": 1605, + "time_per_iteration": 3.615204334259033 + }, + { + "auxiliary_loss_clip": 0.01221891, + "auxiliary_loss_mlp": 0.0104103, + "balance_loss_clip": 1.06686211, + "balance_loss_mlp": 1.03169572, + "epoch": 0.19311008236638008, + "flos": 19281516906240.0, + "grad_norm": 1.8346809654641734, + "language_loss": 0.76000285, + "learning_rate": 3.7274889403139002e-06, + "loss": 0.78263205, + "num_input_tokens_seen": 34033150, + "step": 1606, + "time_per_iteration": 2.6053965091705322 + }, + { + "auxiliary_loss_clip": 0.01247293, + "auxiliary_loss_mlp": 0.01028845, + "balance_loss_clip": 1.05771637, + "balance_loss_mlp": 1.01986814, + "epoch": 0.1932303252570192, + "flos": 28652369587200.0, + "grad_norm": 2.250523641082838, + "language_loss": 0.7843051, + "learning_rate": 3.727096262077179e-06, + "loss": 0.8070665, + "num_input_tokens_seen": 34052145, + "step": 1607, + "time_per_iteration": 2.7635905742645264 + }, + { + "auxiliary_loss_clip": 0.01229219, + "auxiliary_loss_mlp": 0.01037119, + "balance_loss_clip": 1.0620507, + "balance_loss_mlp": 1.02757645, + "epoch": 0.19335056814765827, + "flos": 18368990864640.0, + "grad_norm": 1.6845445730504924, + "language_loss": 0.85308403, + "learning_rate": 3.7267033218413285e-06, + "loss": 0.87574738, + "num_input_tokens_seen": 34069940, + "step": 1608, + "time_per_iteration": 2.6234028339385986 + }, + { + "auxiliary_loss_clip": 0.01255168, + "auxiliary_loss_mlp": 0.01034995, + "balance_loss_clip": 1.04986429, + "balance_loss_mlp": 1.02555954, + "epoch": 0.19347081103829736, + "flos": 13260877741440.0, + "grad_norm": 1.9820630686067575, + "language_loss": 0.81032741, + "learning_rate": 3.726310119665957e-06, + "loss": 0.83322906, + "num_input_tokens_seen": 34086275, + "step": 1609, + "time_per_iteration": 2.7148001194000244 + }, + { + "auxiliary_loss_clip": 0.0122775, + "auxiliary_loss_mlp": 0.01033303, + "balance_loss_clip": 1.06221342, + "balance_loss_mlp": 1.02372479, + "epoch": 0.19359105392893644, + "flos": 20300122788480.0, + "grad_norm": 1.9754364552386467, + "language_loss": 0.85502321, + "learning_rate": 3.725916655610713e-06, + "loss": 0.87763375, + "num_input_tokens_seen": 34105605, + "step": 1610, + "time_per_iteration": 2.6467864513397217 + }, + { + "auxiliary_loss_clip": 0.01228482, + "auxiliary_loss_mlp": 0.01035094, + "balance_loss_clip": 1.06018853, + "balance_loss_mlp": 1.02501512, + "epoch": 0.19371129681957555, + "flos": 20484596062080.0, + "grad_norm": 3.132473494629016, + "language_loss": 0.75530386, + "learning_rate": 3.725522929735284e-06, + "loss": 0.77793962, + "num_input_tokens_seen": 34122540, + "step": 1611, + "time_per_iteration": 2.6607398986816406 + }, + { + "auxiliary_loss_clip": 0.01237527, + "auxiliary_loss_mlp": 0.01033768, + "balance_loss_clip": 1.05643058, + "balance_loss_mlp": 1.02392149, + "epoch": 0.19383153971021463, + "flos": 30445497457920.0, + "grad_norm": 2.2602221488812457, + "language_loss": 0.74028146, + "learning_rate": 3.725128942099399e-06, + "loss": 0.76299447, + "num_input_tokens_seen": 34142940, + "step": 1612, + "time_per_iteration": 2.8473317623138428 + }, + { + "auxiliary_loss_clip": 0.01221785, + "auxiliary_loss_mlp": 0.01031469, + "balance_loss_clip": 1.05520058, + "balance_loss_mlp": 1.02151465, + "epoch": 0.19395178260085372, + "flos": 24569937325440.0, + "grad_norm": 1.8622185129586566, + "language_loss": 0.79854459, + "learning_rate": 3.7247346927628245e-06, + "loss": 0.82107711, + "num_input_tokens_seen": 34162875, + "step": 1613, + "time_per_iteration": 2.722994804382324 + }, + { + "auxiliary_loss_clip": 0.0123393, + "auxiliary_loss_mlp": 0.01380502, + "balance_loss_clip": 1.06050634, + "balance_loss_mlp": 1.00018477, + "epoch": 0.19407202549149283, + "flos": 28950608211840.0, + "grad_norm": 1.9965095101807606, + "language_loss": 0.79261094, + "learning_rate": 3.7243401817853694e-06, + "loss": 0.81875527, + "num_input_tokens_seen": 34183565, + "step": 1614, + "time_per_iteration": 2.760986566543579 + }, + { + "auxiliary_loss_clip": 0.01220147, + "auxiliary_loss_mlp": 0.01032543, + "balance_loss_clip": 1.06155825, + "balance_loss_mlp": 1.02313781, + "epoch": 0.1941922683821319, + "flos": 18004497603840.0, + "grad_norm": 1.866274882751797, + "language_loss": 0.71942049, + "learning_rate": 3.723945409226879e-06, + "loss": 0.74194741, + "num_input_tokens_seen": 34202055, + "step": 1615, + "time_per_iteration": 2.636820077896118 + }, + { + "auxiliary_loss_clip": 0.01224865, + "auxiliary_loss_mlp": 0.01035753, + "balance_loss_clip": 1.05906701, + "balance_loss_mlp": 1.02565598, + "epoch": 0.194312511272771, + "flos": 9720337034880.0, + "grad_norm": 2.12697807689995, + "language_loss": 0.80033565, + "learning_rate": 3.723550375147241e-06, + "loss": 0.82294184, + "num_input_tokens_seen": 34216830, + "step": 1616, + "time_per_iteration": 2.5849545001983643 + }, + { + "auxiliary_loss_clip": 0.01224067, + "auxiliary_loss_mlp": 0.01037357, + "balance_loss_clip": 1.05246103, + "balance_loss_mlp": 1.02802336, + "epoch": 0.19443275416341008, + "flos": 27016208150400.0, + "grad_norm": 2.175562667226884, + "language_loss": 0.79866803, + "learning_rate": 3.7231550796063816e-06, + "loss": 0.82128227, + "num_input_tokens_seen": 34236840, + "step": 1617, + "time_per_iteration": 2.7586328983306885 + }, + { + "auxiliary_loss_clip": 0.01245718, + "auxiliary_loss_mlp": 0.01032948, + "balance_loss_clip": 1.06360054, + "balance_loss_mlp": 1.02322686, + "epoch": 0.1945529970540492, + "flos": 15846625077120.0, + "grad_norm": 1.8552160835263343, + "language_loss": 0.65068763, + "learning_rate": 3.722759522664266e-06, + "loss": 0.67347431, + "num_input_tokens_seen": 34254140, + "step": 1618, + "time_per_iteration": 2.650352954864502 + }, + { + "auxiliary_loss_clip": 0.01249934, + "auxiliary_loss_mlp": 0.01040134, + "balance_loss_clip": 1.05651367, + "balance_loss_mlp": 1.03082347, + "epoch": 0.19467323994468827, + "flos": 19314985403520.0, + "grad_norm": 1.8278541842053033, + "language_loss": 0.81798661, + "learning_rate": 3.7223637043809016e-06, + "loss": 0.84088731, + "num_input_tokens_seen": 34273120, + "step": 1619, + "time_per_iteration": 2.73252534866333 + }, + { + "auxiliary_loss_clip": 0.01247978, + "auxiliary_loss_mlp": 0.0103608, + "balance_loss_clip": 1.06131387, + "balance_loss_mlp": 1.02662098, + "epoch": 0.19479348283532735, + "flos": 24133227770880.0, + "grad_norm": 1.9543410870454088, + "language_loss": 0.86454386, + "learning_rate": 3.7219676248163322e-06, + "loss": 0.88738441, + "num_input_tokens_seen": 34290285, + "step": 1620, + "time_per_iteration": 2.684662103652954 + }, + { + "auxiliary_loss_clip": 0.01233987, + "auxiliary_loss_mlp": 0.01036372, + "balance_loss_clip": 1.06490898, + "balance_loss_mlp": 1.02718711, + "epoch": 0.19491372572596646, + "flos": 25775638174080.0, + "grad_norm": 1.9927077998184075, + "language_loss": 0.93286157, + "learning_rate": 3.721571284030643e-06, + "loss": 0.95556509, + "num_input_tokens_seen": 34310095, + "step": 1621, + "time_per_iteration": 2.68229603767395 + }, + { + "auxiliary_loss_clip": 0.01228913, + "auxiliary_loss_mlp": 0.01037716, + "balance_loss_clip": 1.06050777, + "balance_loss_mlp": 1.02832866, + "epoch": 0.19503396861660555, + "flos": 19645220067840.0, + "grad_norm": 2.1588065392881823, + "language_loss": 0.78859895, + "learning_rate": 3.7211746820839587e-06, + "loss": 0.81126523, + "num_input_tokens_seen": 34327190, + "step": 1622, + "time_per_iteration": 2.6712565422058105 + }, + { + "auxiliary_loss_clip": 0.01247445, + "auxiliary_loss_mlp": 0.01035055, + "balance_loss_clip": 1.04978716, + "balance_loss_mlp": 1.02551234, + "epoch": 0.19515421150724463, + "flos": 21033023892480.0, + "grad_norm": 3.6143984674655987, + "language_loss": 0.80880535, + "learning_rate": 3.7207778190364437e-06, + "loss": 0.83163035, + "num_input_tokens_seen": 34345615, + "step": 1623, + "time_per_iteration": 2.8369176387786865 + }, + { + "auxiliary_loss_clip": 0.01240214, + "auxiliary_loss_mlp": 0.01034255, + "balance_loss_clip": 1.05077481, + "balance_loss_mlp": 1.02576137, + "epoch": 0.1952744543978837, + "flos": 32961255143040.0, + "grad_norm": 1.5150554031438894, + "language_loss": 0.74002874, + "learning_rate": 3.720380694948302e-06, + "loss": 0.76277339, + "num_input_tokens_seen": 34368500, + "step": 1624, + "time_per_iteration": 2.7988524436950684 + }, + { + "auxiliary_loss_clip": 0.01173368, + "auxiliary_loss_mlp": 0.01011985, + "balance_loss_clip": 1.03590786, + "balance_loss_mlp": 1.00915956, + "epoch": 0.19539469728852282, + "flos": 64044312030720.0, + "grad_norm": 1.0409402868163478, + "language_loss": 0.71248221, + "learning_rate": 3.719983309879777e-06, + "loss": 0.73433578, + "num_input_tokens_seen": 34428280, + "step": 1625, + "time_per_iteration": 3.2525441646575928 + }, + { + "auxiliary_loss_clip": 0.01232998, + "auxiliary_loss_mlp": 0.01036202, + "balance_loss_clip": 1.05656099, + "balance_loss_mlp": 1.02637327, + "epoch": 0.1955149401791619, + "flos": 13370908078080.0, + "grad_norm": 1.7338519920284, + "language_loss": 0.7751379, + "learning_rate": 3.719585663891151e-06, + "loss": 0.79782987, + "num_input_tokens_seen": 34445815, + "step": 1626, + "time_per_iteration": 3.6269006729125977 + }, + { + "auxiliary_loss_clip": 0.01242213, + "auxiliary_loss_mlp": 0.01037056, + "balance_loss_clip": 1.05667937, + "balance_loss_mlp": 1.02750111, + "epoch": 0.195635183069801, + "flos": 18728887184640.0, + "grad_norm": 1.9930521433683435, + "language_loss": 0.79214835, + "learning_rate": 3.719187757042747e-06, + "loss": 0.81494105, + "num_input_tokens_seen": 34463635, + "step": 1627, + "time_per_iteration": 2.683145046234131 + }, + { + "auxiliary_loss_clip": 0.01138786, + "auxiliary_loss_mlp": 0.01006671, + "balance_loss_clip": 1.02953339, + "balance_loss_mlp": 1.00406075, + "epoch": 0.1957554259604401, + "flos": 69313952615040.0, + "grad_norm": 0.730716019762997, + "language_loss": 0.54990387, + "learning_rate": 3.7187895893949275e-06, + "loss": 0.57135844, + "num_input_tokens_seen": 34530105, + "step": 1628, + "time_per_iteration": 4.296249628067017 + }, + { + "auxiliary_loss_clip": 0.0123529, + "auxiliary_loss_mlp": 0.01029721, + "balance_loss_clip": 1.05207741, + "balance_loss_mlp": 1.01992226, + "epoch": 0.19587566885107918, + "flos": 21069257736960.0, + "grad_norm": 2.342759413777079, + "language_loss": 0.76437765, + "learning_rate": 3.7183911610080937e-06, + "loss": 0.78702778, + "num_input_tokens_seen": 34546970, + "step": 1629, + "time_per_iteration": 2.740668773651123 + }, + { + "auxiliary_loss_clip": 0.01243526, + "auxiliary_loss_mlp": 0.01051191, + "balance_loss_clip": 1.05838227, + "balance_loss_mlp": 1.04078412, + "epoch": 0.19599591174171827, + "flos": 22194661731840.0, + "grad_norm": 2.5604181907298162, + "language_loss": 0.74775803, + "learning_rate": 3.7179924719426872e-06, + "loss": 0.77070522, + "num_input_tokens_seen": 34564865, + "step": 1630, + "time_per_iteration": 3.649298906326294 + }, + { + "auxiliary_loss_clip": 0.01230347, + "auxiliary_loss_mlp": 0.01037125, + "balance_loss_clip": 1.06300676, + "balance_loss_mlp": 1.02778482, + "epoch": 0.19611615463235738, + "flos": 23768375374080.0, + "grad_norm": 2.7754256193128795, + "language_loss": 0.75668454, + "learning_rate": 3.7175935222591885e-06, + "loss": 0.77935928, + "num_input_tokens_seen": 34584165, + "step": 1631, + "time_per_iteration": 3.623095750808716 + }, + { + "auxiliary_loss_clip": 0.01243262, + "auxiliary_loss_mlp": 0.01041771, + "balance_loss_clip": 1.06633925, + "balance_loss_mlp": 1.03179944, + "epoch": 0.19623639752299646, + "flos": 28618218731520.0, + "grad_norm": 1.9155747082807386, + "language_loss": 0.74296618, + "learning_rate": 3.717194312018118e-06, + "loss": 0.76581651, + "num_input_tokens_seen": 34603150, + "step": 1632, + "time_per_iteration": 2.735193967819214 + }, + { + "auxiliary_loss_clip": 0.01232229, + "auxiliary_loss_mlp": 0.01036044, + "balance_loss_clip": 1.06355274, + "balance_loss_mlp": 1.02594078, + "epoch": 0.19635664041363554, + "flos": 21032700670080.0, + "grad_norm": 2.0652617980767154, + "language_loss": 0.76297671, + "learning_rate": 3.716794841280036e-06, + "loss": 0.78565943, + "num_input_tokens_seen": 34621855, + "step": 1633, + "time_per_iteration": 2.6123859882354736 + }, + { + "auxiliary_loss_clip": 0.01234858, + "auxiliary_loss_mlp": 0.01038752, + "balance_loss_clip": 1.06411552, + "balance_loss_mlp": 1.02829695, + "epoch": 0.19647688330427462, + "flos": 18879748306560.0, + "grad_norm": 2.78039900251183, + "language_loss": 0.77197015, + "learning_rate": 3.7163951101055407e-06, + "loss": 0.79470623, + "num_input_tokens_seen": 34639915, + "step": 1634, + "time_per_iteration": 2.6013548374176025 + }, + { + "auxiliary_loss_clip": 0.01232327, + "auxiliary_loss_mlp": 0.01042416, + "balance_loss_clip": 1.06108809, + "balance_loss_mlp": 1.03229535, + "epoch": 0.19659712619491373, + "flos": 24242503921920.0, + "grad_norm": 1.770295289905249, + "language_loss": 0.79070055, + "learning_rate": 3.715995118555273e-06, + "loss": 0.81344795, + "num_input_tokens_seen": 34659890, + "step": 1635, + "time_per_iteration": 2.7070958614349365 + }, + { + "auxiliary_loss_clip": 0.01245908, + "auxiliary_loss_mlp": 0.01032811, + "balance_loss_clip": 1.05420446, + "balance_loss_mlp": 1.02280927, + "epoch": 0.19671736908555282, + "flos": 24717422568960.0, + "grad_norm": 2.1912630791044507, + "language_loss": 0.85766804, + "learning_rate": 3.71559486668991e-06, + "loss": 0.8804552, + "num_input_tokens_seen": 34678750, + "step": 1636, + "time_per_iteration": 2.801345109939575 + }, + { + "auxiliary_loss_clip": 0.01235093, + "auxiliary_loss_mlp": 0.01380797, + "balance_loss_clip": 1.06509852, + "balance_loss_mlp": 1.00026631, + "epoch": 0.1968376119761919, + "flos": 23842279607040.0, + "grad_norm": 1.729715058174772, + "language_loss": 0.77289075, + "learning_rate": 3.715194354570169e-06, + "loss": 0.79904968, + "num_input_tokens_seen": 34698755, + "step": 1637, + "time_per_iteration": 2.668491840362549 + }, + { + "auxiliary_loss_clip": 0.01227568, + "auxiliary_loss_mlp": 0.01041761, + "balance_loss_clip": 1.06470394, + "balance_loss_mlp": 1.03185487, + "epoch": 0.196957854866831, + "flos": 18113917409280.0, + "grad_norm": 2.0144225917746286, + "language_loss": 0.83215052, + "learning_rate": 3.714793582256809e-06, + "loss": 0.85484374, + "num_input_tokens_seen": 34715820, + "step": 1638, + "time_per_iteration": 2.5985331535339355 + }, + { + "auxiliary_loss_clip": 0.01219867, + "auxiliary_loss_mlp": 0.01034463, + "balance_loss_clip": 1.06368256, + "balance_loss_mlp": 1.02520645, + "epoch": 0.1970780977574701, + "flos": 21653129312640.0, + "grad_norm": 2.6825845099836534, + "language_loss": 0.85167044, + "learning_rate": 3.7143925498106253e-06, + "loss": 0.87421376, + "num_input_tokens_seen": 34734360, + "step": 1639, + "time_per_iteration": 2.607614278793335 + }, + { + "auxiliary_loss_clip": 0.012294, + "auxiliary_loss_mlp": 0.01038742, + "balance_loss_clip": 1.05463815, + "balance_loss_mlp": 1.02904463, + "epoch": 0.19719834064810918, + "flos": 20811813984000.0, + "grad_norm": 1.8486742063935753, + "language_loss": 0.79132992, + "learning_rate": 3.7139912572924558e-06, + "loss": 0.81401139, + "num_input_tokens_seen": 34753390, + "step": 1640, + "time_per_iteration": 2.6889488697052 + }, + { + "auxiliary_loss_clip": 0.01220707, + "auxiliary_loss_mlp": 0.01036209, + "balance_loss_clip": 1.05739856, + "balance_loss_mlp": 1.02592731, + "epoch": 0.19731858353874826, + "flos": 23434800744960.0, + "grad_norm": 2.676962082636155, + "language_loss": 0.80715585, + "learning_rate": 3.7135897047631744e-06, + "loss": 0.82972515, + "num_input_tokens_seen": 34771275, + "step": 1641, + "time_per_iteration": 2.6758341789245605 + }, + { + "auxiliary_loss_clip": 0.01238374, + "auxiliary_loss_mlp": 0.01034567, + "balance_loss_clip": 1.05797744, + "balance_loss_mlp": 1.02521539, + "epoch": 0.19743882642938737, + "flos": 23988184652160.0, + "grad_norm": 2.86669776686613, + "language_loss": 0.75839996, + "learning_rate": 3.713187892283698e-06, + "loss": 0.78112936, + "num_input_tokens_seen": 34790885, + "step": 1642, + "time_per_iteration": 2.650284767150879 + }, + { + "auxiliary_loss_clip": 0.01248691, + "auxiliary_loss_mlp": 0.01031227, + "balance_loss_clip": 1.05457187, + "balance_loss_mlp": 1.02151108, + "epoch": 0.19755906932002645, + "flos": 15004340081280.0, + "grad_norm": 2.442052709981298, + "language_loss": 0.87195504, + "learning_rate": 3.71278581991498e-06, + "loss": 0.89475423, + "num_input_tokens_seen": 34806745, + "step": 1643, + "time_per_iteration": 2.748849391937256 + }, + { + "auxiliary_loss_clip": 0.01246376, + "auxiliary_loss_mlp": 0.0138081, + "balance_loss_clip": 1.05993295, + "balance_loss_mlp": 1.00039458, + "epoch": 0.19767931221066554, + "flos": 19494466686720.0, + "grad_norm": 1.9425644591765463, + "language_loss": 0.78855014, + "learning_rate": 3.712383487718015e-06, + "loss": 0.81482202, + "num_input_tokens_seen": 34824985, + "step": 1644, + "time_per_iteration": 2.6798648834228516 + }, + { + "auxiliary_loss_clip": 0.01255128, + "auxiliary_loss_mlp": 0.01028598, + "balance_loss_clip": 1.05642724, + "balance_loss_mlp": 1.02009881, + "epoch": 0.19779955510130465, + "flos": 25737895958400.0, + "grad_norm": 1.9070436744528623, + "language_loss": 0.86569548, + "learning_rate": 3.7119808957538365e-06, + "loss": 0.88853276, + "num_input_tokens_seen": 34843980, + "step": 1645, + "time_per_iteration": 2.8334999084472656 + }, + { + "auxiliary_loss_clip": 0.01226849, + "auxiliary_loss_mlp": 0.01035305, + "balance_loss_clip": 1.05498159, + "balance_loss_mlp": 1.02467763, + "epoch": 0.19791979799194373, + "flos": 20777699041920.0, + "grad_norm": 2.254545688151635, + "language_loss": 0.80437744, + "learning_rate": 3.711578044083517e-06, + "loss": 0.82699895, + "num_input_tokens_seen": 34860780, + "step": 1646, + "time_per_iteration": 2.6386163234710693 + }, + { + "auxiliary_loss_clip": 0.01232602, + "auxiliary_loss_mlp": 0.01037451, + "balance_loss_clip": 1.05769491, + "balance_loss_mlp": 1.02786613, + "epoch": 0.1980400408825828, + "flos": 25589010084480.0, + "grad_norm": 2.010258771376046, + "language_loss": 0.7458325, + "learning_rate": 3.7111749327681698e-06, + "loss": 0.76853299, + "num_input_tokens_seen": 34880815, + "step": 1647, + "time_per_iteration": 2.7027082443237305 + }, + { + "auxiliary_loss_clip": 0.01230666, + "auxiliary_loss_mlp": 0.0104213, + "balance_loss_clip": 1.06192327, + "balance_loss_mlp": 1.03304052, + "epoch": 0.1981602837732219, + "flos": 23513840622720.0, + "grad_norm": 2.2932894142072375, + "language_loss": 0.86332679, + "learning_rate": 3.7107715618689455e-06, + "loss": 0.88605475, + "num_input_tokens_seen": 34899790, + "step": 1648, + "time_per_iteration": 2.615725517272949 + }, + { + "auxiliary_loss_clip": 0.01224239, + "auxiliary_loss_mlp": 0.0103447, + "balance_loss_clip": 1.06180525, + "balance_loss_mlp": 1.02427721, + "epoch": 0.198280526663861, + "flos": 23185365724800.0, + "grad_norm": 1.5449702904088911, + "language_loss": 0.83269906, + "learning_rate": 3.710367931447035e-06, + "loss": 0.85528618, + "num_input_tokens_seen": 34921570, + "step": 1649, + "time_per_iteration": 2.6880719661712646 + }, + { + "auxiliary_loss_clip": 0.01236031, + "auxiliary_loss_mlp": 0.01042041, + "balance_loss_clip": 1.06358242, + "balance_loss_mlp": 1.03175282, + "epoch": 0.1984007695545001, + "flos": 21689470897920.0, + "grad_norm": 2.3559987114557597, + "language_loss": 0.86486101, + "learning_rate": 3.70996404156367e-06, + "loss": 0.88764179, + "num_input_tokens_seen": 34941205, + "step": 1650, + "time_per_iteration": 2.6465117931365967 + }, + { + "auxiliary_loss_clip": 0.01234659, + "auxiliary_loss_mlp": 0.01035908, + "balance_loss_clip": 1.05216146, + "balance_loss_mlp": 1.02701461, + "epoch": 0.19852101244513917, + "flos": 36064008887040.0, + "grad_norm": 1.6400659577008434, + "language_loss": 0.72685027, + "learning_rate": 3.7095598922801187e-06, + "loss": 0.74955589, + "num_input_tokens_seen": 34963280, + "step": 1651, + "time_per_iteration": 2.8650054931640625 + }, + { + "auxiliary_loss_clip": 0.01219446, + "auxiliary_loss_mlp": 0.01032889, + "balance_loss_clip": 1.06346273, + "balance_loss_mlp": 1.02315545, + "epoch": 0.19864125533577828, + "flos": 23105894883840.0, + "grad_norm": 2.661372893606717, + "language_loss": 0.76025093, + "learning_rate": 3.7091554836576914e-06, + "loss": 0.78277421, + "num_input_tokens_seen": 34979955, + "step": 1652, + "time_per_iteration": 2.6495234966278076 + }, + { + "auxiliary_loss_clip": 0.01225155, + "auxiliary_loss_mlp": 0.01380447, + "balance_loss_clip": 1.06171393, + "balance_loss_mlp": 1.00027895, + "epoch": 0.19876149822641737, + "flos": 24608505553920.0, + "grad_norm": 1.71196753043, + "language_loss": 0.82937002, + "learning_rate": 3.708750815757736e-06, + "loss": 0.85542607, + "num_input_tokens_seen": 35000725, + "step": 1653, + "time_per_iteration": 3.5514254570007324 + }, + { + "auxiliary_loss_clip": 0.01225735, + "auxiliary_loss_mlp": 0.01034262, + "balance_loss_clip": 1.05921412, + "balance_loss_mlp": 1.02414083, + "epoch": 0.19888174111705645, + "flos": 32196645308160.0, + "grad_norm": 3.013679731256889, + "language_loss": 0.72944367, + "learning_rate": 3.7083458886416407e-06, + "loss": 0.75204366, + "num_input_tokens_seen": 35019920, + "step": 1654, + "time_per_iteration": 3.6243081092834473 + }, + { + "auxiliary_loss_clip": 0.01259484, + "auxiliary_loss_mlp": 0.01036146, + "balance_loss_clip": 1.0542196, + "balance_loss_mlp": 1.02603734, + "epoch": 0.19900198400769553, + "flos": 24608469640320.0, + "grad_norm": 10.125159527898852, + "language_loss": 0.88201761, + "learning_rate": 3.707940702370832e-06, + "loss": 0.90497398, + "num_input_tokens_seen": 35040765, + "step": 1655, + "time_per_iteration": 2.8040454387664795 + }, + { + "auxiliary_loss_clip": 0.01124922, + "auxiliary_loss_mlp": 0.01012523, + "balance_loss_clip": 1.03055072, + "balance_loss_mlp": 1.01003134, + "epoch": 0.19912222689833464, + "flos": 67915805673600.0, + "grad_norm": 0.7846027073569387, + "language_loss": 0.58259952, + "learning_rate": 3.707535257006777e-06, + "loss": 0.60397398, + "num_input_tokens_seen": 35106390, + "step": 1656, + "time_per_iteration": 4.2402026653289795 + }, + { + "auxiliary_loss_clip": 0.01234375, + "auxiliary_loss_mlp": 0.01034292, + "balance_loss_clip": 1.05786026, + "balance_loss_mlp": 1.02396226, + "epoch": 0.19924246978897373, + "flos": 15742340916480.0, + "grad_norm": 2.2048621382284663, + "language_loss": 0.88686711, + "learning_rate": 3.707129552610981e-06, + "loss": 0.90955377, + "num_input_tokens_seen": 35125040, + "step": 1657, + "time_per_iteration": 3.5690529346466064 + }, + { + "auxiliary_loss_clip": 0.01227664, + "auxiliary_loss_mlp": 0.01034472, + "balance_loss_clip": 1.06014919, + "balance_loss_mlp": 1.02451169, + "epoch": 0.1993627126796128, + "flos": 17566566986880.0, + "grad_norm": 3.2564881965139745, + "language_loss": 0.73954415, + "learning_rate": 3.70672358924499e-06, + "loss": 0.76216543, + "num_input_tokens_seen": 35144280, + "step": 1658, + "time_per_iteration": 2.6471779346466064 + }, + { + "auxiliary_loss_clip": 0.01243415, + "auxiliary_loss_mlp": 0.01038485, + "balance_loss_clip": 1.06142831, + "balance_loss_mlp": 1.02891207, + "epoch": 0.19948295557025192, + "flos": 40843826680320.0, + "grad_norm": 1.9338400373153888, + "language_loss": 0.78879768, + "learning_rate": 3.706317366970386e-06, + "loss": 0.81161666, + "num_input_tokens_seen": 35165280, + "step": 1659, + "time_per_iteration": 2.8085505962371826 + }, + { + "auxiliary_loss_clip": 0.01218295, + "auxiliary_loss_mlp": 0.01381187, + "balance_loss_clip": 1.05878806, + "balance_loss_mlp": 1.00030887, + "epoch": 0.199603198460891, + "flos": 25082418620160.0, + "grad_norm": 1.9572042566139147, + "language_loss": 0.83487892, + "learning_rate": 3.705910885848795e-06, + "loss": 0.8608737, + "num_input_tokens_seen": 35183655, + "step": 1660, + "time_per_iteration": 2.6777572631835938 + }, + { + "auxiliary_loss_clip": 0.0122607, + "auxiliary_loss_mlp": 0.01034773, + "balance_loss_clip": 1.06229472, + "balance_loss_mlp": 1.0251708, + "epoch": 0.19972344135153008, + "flos": 20084120352000.0, + "grad_norm": 2.7498748648324725, + "language_loss": 0.84960043, + "learning_rate": 3.705504145941879e-06, + "loss": 0.87220883, + "num_input_tokens_seen": 35201825, + "step": 1661, + "time_per_iteration": 2.5919580459594727 + }, + { + "auxiliary_loss_clip": 0.0121967, + "auxiliary_loss_mlp": 0.01031792, + "balance_loss_clip": 1.0643909, + "balance_loss_mlp": 1.02154613, + "epoch": 0.1998436842421692, + "flos": 23727472761600.0, + "grad_norm": 1.9982737678339968, + "language_loss": 0.78621328, + "learning_rate": 3.7050971473113403e-06, + "loss": 0.80872798, + "num_input_tokens_seen": 35221600, + "step": 1662, + "time_per_iteration": 2.580399513244629 + }, + { + "auxiliary_loss_clip": 0.01220506, + "auxiliary_loss_mlp": 0.0138044, + "balance_loss_clip": 1.05963683, + "balance_loss_mlp": 1.00022149, + "epoch": 0.19996392713280828, + "flos": 36102361633920.0, + "grad_norm": 1.6903916171875035, + "language_loss": 0.79960871, + "learning_rate": 3.7046898900189196e-06, + "loss": 0.82561815, + "num_input_tokens_seen": 35245935, + "step": 1663, + "time_per_iteration": 2.737147331237793 + }, + { + "auxiliary_loss_clip": 0.01245016, + "auxiliary_loss_mlp": 0.01035443, + "balance_loss_clip": 1.05886936, + "balance_loss_mlp": 1.02561998, + "epoch": 0.20008417002344736, + "flos": 23657662679040.0, + "grad_norm": 3.0252687045507685, + "language_loss": 0.82916248, + "learning_rate": 3.704282374126398e-06, + "loss": 0.8519671, + "num_input_tokens_seen": 35265615, + "step": 1664, + "time_per_iteration": 2.7129218578338623 + }, + { + "auxiliary_loss_clip": 0.0123932, + "auxiliary_loss_mlp": 0.01040767, + "balance_loss_clip": 1.054528, + "balance_loss_mlp": 1.03142095, + "epoch": 0.20020441291408644, + "flos": 21872076664320.0, + "grad_norm": 1.8591094368409005, + "language_loss": 0.87562644, + "learning_rate": 3.7038745996955954e-06, + "loss": 0.89842731, + "num_input_tokens_seen": 35284960, + "step": 1665, + "time_per_iteration": 277.529926776886 + }, + { + "auxiliary_loss_clip": 0.01243982, + "auxiliary_loss_mlp": 0.01040348, + "balance_loss_clip": 1.05627179, + "balance_loss_mlp": 1.03022707, + "epoch": 0.20032465580472555, + "flos": 23179691376000.0, + "grad_norm": 2.7586137843045373, + "language_loss": 0.7219156, + "learning_rate": 3.703466566788371e-06, + "loss": 0.74475896, + "num_input_tokens_seen": 35304090, + "step": 1666, + "time_per_iteration": 2.750622272491455 + }, + { + "auxiliary_loss_clip": 0.01225229, + "auxiliary_loss_mlp": 0.01044917, + "balance_loss_clip": 1.05843711, + "balance_loss_mlp": 1.03511262, + "epoch": 0.20044489869536464, + "flos": 23873521461120.0, + "grad_norm": 2.015834359309433, + "language_loss": 0.7432031, + "learning_rate": 3.703058275466622e-06, + "loss": 0.76590455, + "num_input_tokens_seen": 35323325, + "step": 1667, + "time_per_iteration": 2.7204527854919434 + }, + { + "auxiliary_loss_clip": 0.01232408, + "auxiliary_loss_mlp": 0.01033392, + "balance_loss_clip": 1.05794883, + "balance_loss_mlp": 1.02432632, + "epoch": 0.20056514158600372, + "flos": 21945226711680.0, + "grad_norm": 2.0056680023330054, + "language_loss": 0.78045976, + "learning_rate": 3.7026497257922877e-06, + "loss": 0.80311775, + "num_input_tokens_seen": 35343635, + "step": 1668, + "time_per_iteration": 2.712399959564209 + }, + { + "auxiliary_loss_clip": 0.01245075, + "auxiliary_loss_mlp": 0.01025728, + "balance_loss_clip": 1.05284059, + "balance_loss_mlp": 1.01640582, + "epoch": 0.20068538447664283, + "flos": 23879159896320.0, + "grad_norm": 1.6789736507015434, + "language_loss": 0.85446078, + "learning_rate": 3.7022409178273436e-06, + "loss": 0.87716877, + "num_input_tokens_seen": 35364615, + "step": 1669, + "time_per_iteration": 2.7640249729156494 + }, + { + "auxiliary_loss_clip": 0.0122352, + "auxiliary_loss_mlp": 0.01031895, + "balance_loss_clip": 1.06012511, + "balance_loss_mlp": 1.02223337, + "epoch": 0.2008056273672819, + "flos": 18442823270400.0, + "grad_norm": 1.9226468599114468, + "language_loss": 0.78589433, + "learning_rate": 3.7018318516338054e-06, + "loss": 0.80844849, + "num_input_tokens_seen": 35383775, + "step": 1670, + "time_per_iteration": 2.636781930923462 + }, + { + "auxiliary_loss_clip": 0.0123245, + "auxiliary_loss_mlp": 0.01038093, + "balance_loss_clip": 1.06247711, + "balance_loss_mlp": 1.02857995, + "epoch": 0.200925870257921, + "flos": 23659530186240.0, + "grad_norm": 2.388634166340792, + "language_loss": 0.81731629, + "learning_rate": 3.7014225272737284e-06, + "loss": 0.84002173, + "num_input_tokens_seen": 35403000, + "step": 1671, + "time_per_iteration": 2.749871015548706 + }, + { + "auxiliary_loss_clip": 0.01215075, + "auxiliary_loss_mlp": 0.010335, + "balance_loss_clip": 1.05689812, + "balance_loss_mlp": 1.02428555, + "epoch": 0.20104611314856008, + "flos": 16217115909120.0, + "grad_norm": 2.217936964655928, + "language_loss": 0.74283832, + "learning_rate": 3.701012944809207e-06, + "loss": 0.76532412, + "num_input_tokens_seen": 35420115, + "step": 1672, + "time_per_iteration": 2.633826494216919 + }, + { + "auxiliary_loss_clip": 0.01236283, + "auxiliary_loss_mlp": 0.0138044, + "balance_loss_clip": 1.05970645, + "balance_loss_mlp": 1.00027728, + "epoch": 0.2011663560391992, + "flos": 21397373498880.0, + "grad_norm": 1.8941494324218542, + "language_loss": 0.78799224, + "learning_rate": 3.700603104302374e-06, + "loss": 0.81415951, + "num_input_tokens_seen": 35439925, + "step": 1673, + "time_per_iteration": 2.799649715423584 + }, + { + "auxiliary_loss_clip": 0.01133856, + "auxiliary_loss_mlp": 0.01001536, + "balance_loss_clip": 1.02698839, + "balance_loss_mlp": 0.99883014, + "epoch": 0.20128659892983827, + "flos": 62229459409920.0, + "grad_norm": 0.9076350058862757, + "language_loss": 0.55958998, + "learning_rate": 3.7001930058154027e-06, + "loss": 0.58094394, + "num_input_tokens_seen": 35504885, + "step": 1674, + "time_per_iteration": 3.3164212703704834 + }, + { + "auxiliary_loss_clip": 0.01242119, + "auxiliary_loss_mlp": 0.01038253, + "balance_loss_clip": 1.05802858, + "balance_loss_mlp": 1.02825713, + "epoch": 0.20140684182047736, + "flos": 28438737448320.0, + "grad_norm": 4.234432526886847, + "language_loss": 0.79775077, + "learning_rate": 3.6997826494105037e-06, + "loss": 0.82055444, + "num_input_tokens_seen": 35525330, + "step": 1675, + "time_per_iteration": 2.8419737815856934 + }, + { + "auxiliary_loss_clip": 0.01234857, + "auxiliary_loss_mlp": 0.01031999, + "balance_loss_clip": 1.05789769, + "balance_loss_mlp": 1.02261758, + "epoch": 0.20152708471111647, + "flos": 28074064619520.0, + "grad_norm": 1.9592818599803763, + "language_loss": 0.69138801, + "learning_rate": 3.6993720351499286e-06, + "loss": 0.71405661, + "num_input_tokens_seen": 35546455, + "step": 1676, + "time_per_iteration": 2.9169130325317383 + }, + { + "auxiliary_loss_clip": 0.01228449, + "auxiliary_loss_mlp": 0.0103915, + "balance_loss_clip": 1.06007528, + "balance_loss_mlp": 1.02977395, + "epoch": 0.20164732760175555, + "flos": 23549751244800.0, + "grad_norm": 1.8550218475019122, + "language_loss": 0.76913816, + "learning_rate": 3.6989611630959666e-06, + "loss": 0.79181421, + "num_input_tokens_seen": 35565010, + "step": 1677, + "time_per_iteration": 2.8446269035339355 + }, + { + "auxiliary_loss_clip": 0.01122099, + "auxiliary_loss_mlp": 0.01001588, + "balance_loss_clip": 1.02775609, + "balance_loss_mlp": 0.99927503, + "epoch": 0.20176757049239463, + "flos": 71100616037760.0, + "grad_norm": 0.6778552646000285, + "language_loss": 0.58318317, + "learning_rate": 3.6985500333109474e-06, + "loss": 0.60442007, + "num_input_tokens_seen": 35633340, + "step": 1678, + "time_per_iteration": 4.29540753364563 + }, + { + "auxiliary_loss_clip": 0.01228866, + "auxiliary_loss_mlp": 0.01034505, + "balance_loss_clip": 1.05293894, + "balance_loss_mlp": 1.02521873, + "epoch": 0.20188781338303372, + "flos": 21430159637760.0, + "grad_norm": 2.1649983962118577, + "language_loss": 0.76609743, + "learning_rate": 3.6981386458572385e-06, + "loss": 0.7887311, + "num_input_tokens_seen": 35651315, + "step": 1679, + "time_per_iteration": 2.7413582801818848 + }, + { + "auxiliary_loss_clip": 0.01232036, + "auxiliary_loss_mlp": 0.01035768, + "balance_loss_clip": 1.05292308, + "balance_loss_mlp": 1.02562308, + "epoch": 0.20200805627367283, + "flos": 11546215130880.0, + "grad_norm": 3.898385547022079, + "language_loss": 0.76408583, + "learning_rate": 3.6977270007972468e-06, + "loss": 0.78676391, + "num_input_tokens_seen": 35668850, + "step": 1680, + "time_per_iteration": 3.664987802505493 + }, + { + "auxiliary_loss_clip": 0.01237821, + "auxiliary_loss_mlp": 0.01040982, + "balance_loss_clip": 1.06086648, + "balance_loss_mlp": 1.03158832, + "epoch": 0.2021282991643119, + "flos": 28545391906560.0, + "grad_norm": 2.479805586089166, + "language_loss": 0.7201001, + "learning_rate": 3.6973150981934196e-06, + "loss": 0.74288821, + "num_input_tokens_seen": 35690080, + "step": 1681, + "time_per_iteration": 3.639193534851074 + }, + { + "auxiliary_loss_clip": 0.01221902, + "auxiliary_loss_mlp": 0.01038951, + "balance_loss_clip": 1.06261957, + "balance_loss_mlp": 1.02903926, + "epoch": 0.202248542054951, + "flos": 17923446564480.0, + "grad_norm": 2.4264873949052053, + "language_loss": 0.83876407, + "learning_rate": 3.6969029381082415e-06, + "loss": 0.86137259, + "num_input_tokens_seen": 35706075, + "step": 1682, + "time_per_iteration": 2.606851100921631 + }, + { + "auxiliary_loss_clip": 0.01227211, + "auxiliary_loss_mlp": 0.01029882, + "balance_loss_clip": 1.05596638, + "balance_loss_mlp": 1.02107227, + "epoch": 0.2023687849455901, + "flos": 19864634296320.0, + "grad_norm": 1.6670621609005976, + "language_loss": 0.79652119, + "learning_rate": 3.696490520604237e-06, + "loss": 0.81909209, + "num_input_tokens_seen": 35724765, + "step": 1683, + "time_per_iteration": 3.5791752338409424 + }, + { + "auxiliary_loss_clip": 0.01225956, + "auxiliary_loss_mlp": 0.01036642, + "balance_loss_clip": 1.0623132, + "balance_loss_mlp": 1.0272541, + "epoch": 0.20248902783622919, + "flos": 22564721600640.0, + "grad_norm": 1.7603287591421686, + "language_loss": 0.80475867, + "learning_rate": 3.696077845743968e-06, + "loss": 0.82738459, + "num_input_tokens_seen": 35744355, + "step": 1684, + "time_per_iteration": 2.7022175788879395 + }, + { + "auxiliary_loss_clip": 0.01217952, + "auxiliary_loss_mlp": 0.01036661, + "balance_loss_clip": 1.0604794, + "balance_loss_mlp": 1.02650404, + "epoch": 0.20260927072686827, + "flos": 22709728805760.0, + "grad_norm": 2.3396159529551475, + "language_loss": 0.73235446, + "learning_rate": 3.69566491359004e-06, + "loss": 0.75490069, + "num_input_tokens_seen": 35761000, + "step": 1685, + "time_per_iteration": 2.5887811183929443 + }, + { + "auxiliary_loss_clip": 0.01228965, + "auxiliary_loss_mlp": 0.01038813, + "balance_loss_clip": 1.05619454, + "balance_loss_mlp": 1.02859712, + "epoch": 0.20272951361750738, + "flos": 51023998650240.0, + "grad_norm": 2.1809676890701595, + "language_loss": 0.69295597, + "learning_rate": 3.695251724205092e-06, + "loss": 0.71563375, + "num_input_tokens_seen": 35785360, + "step": 1686, + "time_per_iteration": 2.901552677154541 + }, + { + "auxiliary_loss_clip": 0.01217059, + "auxiliary_loss_mlp": 0.01027224, + "balance_loss_clip": 1.06175601, + "balance_loss_mlp": 1.01767516, + "epoch": 0.20284975650814646, + "flos": 26578133879040.0, + "grad_norm": 1.6171705831380037, + "language_loss": 0.86438549, + "learning_rate": 3.6948382776518054e-06, + "loss": 0.8868283, + "num_input_tokens_seen": 35806065, + "step": 1687, + "time_per_iteration": 2.640838384628296 + }, + { + "auxiliary_loss_clip": 0.01250369, + "auxiliary_loss_mlp": 0.01036985, + "balance_loss_clip": 1.05765474, + "balance_loss_mlp": 1.02583885, + "epoch": 0.20296999939878554, + "flos": 16034222833920.0, + "grad_norm": 2.1039872025755773, + "language_loss": 0.78857511, + "learning_rate": 3.6944245739929e-06, + "loss": 0.81144857, + "num_input_tokens_seen": 35822225, + "step": 1688, + "time_per_iteration": 2.6923139095306396 + }, + { + "auxiliary_loss_clip": 0.01223688, + "auxiliary_loss_mlp": 0.01041108, + "balance_loss_clip": 1.06157827, + "balance_loss_mlp": 1.03078485, + "epoch": 0.20309024228942463, + "flos": 19203374868480.0, + "grad_norm": 2.255326100666505, + "language_loss": 0.71920216, + "learning_rate": 3.6940106132911332e-06, + "loss": 0.74185014, + "num_input_tokens_seen": 35839410, + "step": 1689, + "time_per_iteration": 2.659721851348877 + }, + { + "auxiliary_loss_clip": 0.01228612, + "auxiliary_loss_mlp": 0.01032945, + "balance_loss_clip": 1.0618335, + "balance_loss_mlp": 1.02371836, + "epoch": 0.20321048518006374, + "flos": 22821087945600.0, + "grad_norm": 2.142657860788256, + "language_loss": 0.88778222, + "learning_rate": 3.6935963956093037e-06, + "loss": 0.91039771, + "num_input_tokens_seen": 35859495, + "step": 1690, + "time_per_iteration": 2.6283631324768066 + }, + { + "auxiliary_loss_clip": 0.0121376, + "auxiliary_loss_mlp": 0.01034623, + "balance_loss_clip": 1.05774713, + "balance_loss_mlp": 1.02541995, + "epoch": 0.20333072807070282, + "flos": 19096397187840.0, + "grad_norm": 1.9047043909547288, + "language_loss": 0.687325, + "learning_rate": 3.6931819210102474e-06, + "loss": 0.70980895, + "num_input_tokens_seen": 35878890, + "step": 1691, + "time_per_iteration": 2.7386157512664795 + }, + { + "auxiliary_loss_clip": 0.01217622, + "auxiliary_loss_mlp": 0.01040246, + "balance_loss_clip": 1.06054568, + "balance_loss_mlp": 1.03013134, + "epoch": 0.2034509709613419, + "flos": 18180962144640.0, + "grad_norm": 2.2478670355030403, + "language_loss": 0.84511673, + "learning_rate": 3.6927671895568402e-06, + "loss": 0.86769539, + "num_input_tokens_seen": 35897950, + "step": 1692, + "time_per_iteration": 2.634437084197998 + }, + { + "auxiliary_loss_clip": 0.01220807, + "auxiliary_loss_mlp": 0.01040713, + "balance_loss_clip": 1.06439555, + "balance_loss_mlp": 1.0307827, + "epoch": 0.20357121385198101, + "flos": 22923899648640.0, + "grad_norm": 1.9812615398918063, + "language_loss": 0.86604822, + "learning_rate": 3.692352201311996e-06, + "loss": 0.88866335, + "num_input_tokens_seen": 35916800, + "step": 1693, + "time_per_iteration": 2.6483383178710938 + }, + { + "auxiliary_loss_clip": 0.01238346, + "auxiliary_loss_mlp": 0.01032594, + "balance_loss_clip": 1.05709624, + "balance_loss_mlp": 1.02283716, + "epoch": 0.2036914567426201, + "flos": 20922131629440.0, + "grad_norm": 1.7467526333725507, + "language_loss": 0.77029735, + "learning_rate": 3.6919369563386687e-06, + "loss": 0.79300678, + "num_input_tokens_seen": 35936600, + "step": 1694, + "time_per_iteration": 2.7208871841430664 + }, + { + "auxiliary_loss_clip": 0.01230975, + "auxiliary_loss_mlp": 0.01026946, + "balance_loss_clip": 1.05941808, + "balance_loss_mlp": 1.01745117, + "epoch": 0.20381169963325918, + "flos": 15519155760000.0, + "grad_norm": 2.1378370528632735, + "language_loss": 0.79064697, + "learning_rate": 3.69152145469985e-06, + "loss": 0.81322616, + "num_input_tokens_seen": 35953645, + "step": 1695, + "time_per_iteration": 2.699223756790161 + }, + { + "auxiliary_loss_clip": 0.01251413, + "auxiliary_loss_mlp": 0.01034877, + "balance_loss_clip": 1.05547667, + "balance_loss_mlp": 1.02469647, + "epoch": 0.20393194252389826, + "flos": 28833143760000.0, + "grad_norm": 1.8801143390402086, + "language_loss": 0.8203637, + "learning_rate": 3.691105696458572e-06, + "loss": 0.84322655, + "num_input_tokens_seen": 35970940, + "step": 1696, + "time_per_iteration": 2.912236213684082 + }, + { + "auxiliary_loss_clip": 0.01219555, + "auxiliary_loss_mlp": 0.01038691, + "balance_loss_clip": 1.06550348, + "balance_loss_mlp": 1.02923787, + "epoch": 0.20405218541453737, + "flos": 22488554810880.0, + "grad_norm": 3.1518445154686705, + "language_loss": 0.67796993, + "learning_rate": 3.690689681677904e-06, + "loss": 0.7005524, + "num_input_tokens_seen": 35989410, + "step": 1697, + "time_per_iteration": 2.650482177734375 + }, + { + "auxiliary_loss_clip": 0.01233658, + "auxiliary_loss_mlp": 0.01033541, + "balance_loss_clip": 1.05681372, + "balance_loss_mlp": 1.02366424, + "epoch": 0.20417242830517646, + "flos": 25374408278400.0, + "grad_norm": 1.6651219876520442, + "language_loss": 0.88887984, + "learning_rate": 3.690273410420956e-06, + "loss": 0.91155183, + "num_input_tokens_seen": 36009175, + "step": 1698, + "time_per_iteration": 2.8341336250305176 + }, + { + "auxiliary_loss_clip": 0.01221492, + "auxiliary_loss_mlp": 0.01032193, + "balance_loss_clip": 1.0585531, + "balance_loss_mlp": 1.02217984, + "epoch": 0.20429267119581554, + "flos": 14793078240000.0, + "grad_norm": 2.834182827320403, + "language_loss": 0.76452076, + "learning_rate": 3.689856882750875e-06, + "loss": 0.78705764, + "num_input_tokens_seen": 36024375, + "step": 1699, + "time_per_iteration": 2.6322388648986816 + }, + { + "auxiliary_loss_clip": 0.01221326, + "auxiliary_loss_mlp": 0.01036719, + "balance_loss_clip": 1.06163001, + "balance_loss_mlp": 1.02746844, + "epoch": 0.20441291408645465, + "flos": 17781851151360.0, + "grad_norm": 1.932425438201807, + "language_loss": 0.78666496, + "learning_rate": 3.6894400987308486e-06, + "loss": 0.80924547, + "num_input_tokens_seen": 36041895, + "step": 1700, + "time_per_iteration": 2.681504726409912 + }, + { + "auxiliary_loss_clip": 0.01228091, + "auxiliary_loss_mlp": 0.01027369, + "balance_loss_clip": 1.06022191, + "balance_loss_mlp": 1.01771283, + "epoch": 0.20453315697709373, + "flos": 16435668211200.0, + "grad_norm": 2.2880532641566593, + "language_loss": 0.84825933, + "learning_rate": 3.6890230584241024e-06, + "loss": 0.87081397, + "num_input_tokens_seen": 36058825, + "step": 1701, + "time_per_iteration": 2.634615421295166 + }, + { + "auxiliary_loss_clip": 0.01119957, + "auxiliary_loss_mlp": 0.01003414, + "balance_loss_clip": 1.03426099, + "balance_loss_mlp": 1.00096977, + "epoch": 0.20465339986773282, + "flos": 66713085653760.0, + "grad_norm": 1.0636838914386384, + "language_loss": 0.6639154, + "learning_rate": 3.6886057618939016e-06, + "loss": 0.68514919, + "num_input_tokens_seen": 36121645, + "step": 1702, + "time_per_iteration": 3.2744362354278564 + }, + { + "auxiliary_loss_clip": 0.0123525, + "auxiliary_loss_mlp": 0.01035598, + "balance_loss_clip": 1.05687356, + "balance_loss_mlp": 1.02582324, + "epoch": 0.2047736427583719, + "flos": 41974114924800.0, + "grad_norm": 2.301661246354434, + "language_loss": 0.69222152, + "learning_rate": 3.6881882092035492e-06, + "loss": 0.71492994, + "num_input_tokens_seen": 36143030, + "step": 1703, + "time_per_iteration": 2.8601651191711426 + }, + { + "auxiliary_loss_clip": 0.01158652, + "auxiliary_loss_mlp": 0.01376598, + "balance_loss_clip": 1.03247046, + "balance_loss_mlp": 1.00017893, + "epoch": 0.204893885649011, + "flos": 69940878641280.0, + "grad_norm": 0.9226362253644688, + "language_loss": 0.61215991, + "learning_rate": 3.6877704004163873e-06, + "loss": 0.63751245, + "num_input_tokens_seen": 36203435, + "step": 1704, + "time_per_iteration": 4.332745552062988 + }, + { + "auxiliary_loss_clip": 0.01221498, + "auxiliary_loss_mlp": 0.0103757, + "balance_loss_clip": 1.06496406, + "balance_loss_mlp": 1.02739549, + "epoch": 0.2050141285396501, + "flos": 22200012858240.0, + "grad_norm": 2.135956364633467, + "language_loss": 0.77455062, + "learning_rate": 3.6873523355957984e-06, + "loss": 0.79714131, + "num_input_tokens_seen": 36222435, + "step": 1705, + "time_per_iteration": 3.5970699787139893 + }, + { + "auxiliary_loss_clip": 0.01118845, + "auxiliary_loss_mlp": 0.0100133, + "balance_loss_clip": 1.03340983, + "balance_loss_mlp": 0.99883902, + "epoch": 0.20513437143028918, + "flos": 46283721730560.0, + "grad_norm": 0.9908141678472148, + "language_loss": 0.6408447, + "learning_rate": 3.686934014805201e-06, + "loss": 0.66204643, + "num_input_tokens_seen": 36273065, + "step": 1706, + "time_per_iteration": 3.094595432281494 + }, + { + "auxiliary_loss_clip": 0.01225222, + "auxiliary_loss_mlp": 0.01037233, + "balance_loss_clip": 1.06170213, + "balance_loss_mlp": 1.02717209, + "epoch": 0.20525461432092829, + "flos": 21904324099200.0, + "grad_norm": 1.8977681641728754, + "language_loss": 0.80746645, + "learning_rate": 3.6865154381080552e-06, + "loss": 0.830091, + "num_input_tokens_seen": 36293750, + "step": 1707, + "time_per_iteration": 3.6925742626190186 + }, + { + "auxiliary_loss_clip": 0.01269674, + "auxiliary_loss_mlp": 0.01025594, + "balance_loss_clip": 1.05211926, + "balance_loss_mlp": 1.01618242, + "epoch": 0.20537485721156737, + "flos": 21214264942080.0, + "grad_norm": 2.007740871796611, + "language_loss": 0.82865125, + "learning_rate": 3.6860966055678585e-06, + "loss": 0.85160393, + "num_input_tokens_seen": 36310105, + "step": 1708, + "time_per_iteration": 2.7983479499816895 + }, + { + "auxiliary_loss_clip": 0.01229057, + "auxiliary_loss_mlp": 0.01030308, + "balance_loss_clip": 1.06314278, + "balance_loss_mlp": 1.02070534, + "epoch": 0.20549510010220645, + "flos": 20191205773440.0, + "grad_norm": 1.6978486542703328, + "language_loss": 0.86765945, + "learning_rate": 3.685677517248147e-06, + "loss": 0.89025313, + "num_input_tokens_seen": 36328995, + "step": 1709, + "time_per_iteration": 3.539417266845703 + }, + { + "auxiliary_loss_clip": 0.01233763, + "auxiliary_loss_mlp": 0.013808, + "balance_loss_clip": 1.06262732, + "balance_loss_mlp": 1.00030756, + "epoch": 0.20561534299284553, + "flos": 17016702612480.0, + "grad_norm": 1.9873915651220049, + "language_loss": 0.80301595, + "learning_rate": 3.6852581732124967e-06, + "loss": 0.82916164, + "num_input_tokens_seen": 36346340, + "step": 1710, + "time_per_iteration": 2.7027246952056885 + }, + { + "auxiliary_loss_clip": 0.01228576, + "auxiliary_loss_mlp": 0.01034171, + "balance_loss_clip": 1.06353998, + "balance_loss_mlp": 1.02424145, + "epoch": 0.20573558588348465, + "flos": 22890467064960.0, + "grad_norm": 1.896722613110049, + "language_loss": 0.76402438, + "learning_rate": 3.6848385735245213e-06, + "loss": 0.78665185, + "num_input_tokens_seen": 36365430, + "step": 1711, + "time_per_iteration": 2.702627420425415 + }, + { + "auxiliary_loss_clip": 0.01208545, + "auxiliary_loss_mlp": 0.01030298, + "balance_loss_clip": 1.05584276, + "balance_loss_mlp": 1.02095747, + "epoch": 0.20585582877412373, + "flos": 24643123286400.0, + "grad_norm": 1.896663949799562, + "language_loss": 0.86007988, + "learning_rate": 3.6844187182478734e-06, + "loss": 0.88246834, + "num_input_tokens_seen": 36386285, + "step": 1712, + "time_per_iteration": 2.7977559566497803 + }, + { + "auxiliary_loss_clip": 0.01222174, + "auxiliary_loss_mlp": 0.01037303, + "balance_loss_clip": 1.0556376, + "balance_loss_mlp": 1.02737319, + "epoch": 0.2059760716647628, + "flos": 24206952435840.0, + "grad_norm": 2.0083361016846397, + "language_loss": 0.74766433, + "learning_rate": 3.683998607446246e-06, + "loss": 0.77025914, + "num_input_tokens_seen": 36404935, + "step": 1713, + "time_per_iteration": 2.715747356414795 + }, + { + "auxiliary_loss_clip": 0.01228183, + "auxiliary_loss_mlp": 0.01034245, + "balance_loss_clip": 1.06234097, + "balance_loss_mlp": 1.02526271, + "epoch": 0.20609631455540192, + "flos": 20229522606720.0, + "grad_norm": 1.8068363673514725, + "language_loss": 0.74763376, + "learning_rate": 3.6835782411833686e-06, + "loss": 0.77025807, + "num_input_tokens_seen": 36424455, + "step": 1714, + "time_per_iteration": 2.594743490219116 + }, + { + "auxiliary_loss_clip": 0.01226564, + "auxiliary_loss_mlp": 0.01030555, + "balance_loss_clip": 1.05471539, + "balance_loss_mlp": 1.02095246, + "epoch": 0.206216557446041, + "flos": 19864957518720.0, + "grad_norm": 1.6810350026765888, + "language_loss": 0.73952848, + "learning_rate": 3.68315761952301e-06, + "loss": 0.76209962, + "num_input_tokens_seen": 36441685, + "step": 1715, + "time_per_iteration": 2.708423614501953 + }, + { + "auxiliary_loss_clip": 0.01217108, + "auxiliary_loss_mlp": 0.01028919, + "balance_loss_clip": 1.06272566, + "balance_loss_mlp": 1.01900053, + "epoch": 0.2063368003366801, + "flos": 24096311568000.0, + "grad_norm": 2.1377173898495703, + "language_loss": 0.83098292, + "learning_rate": 3.6827367425289797e-06, + "loss": 0.85344321, + "num_input_tokens_seen": 36461460, + "step": 1716, + "time_per_iteration": 2.6099979877471924 + }, + { + "auxiliary_loss_clip": 0.01232801, + "auxiliary_loss_mlp": 0.01030417, + "balance_loss_clip": 1.05878901, + "balance_loss_mlp": 1.02070737, + "epoch": 0.2064570432273192, + "flos": 20340163474560.0, + "grad_norm": 2.4859874371174455, + "language_loss": 0.7288062, + "learning_rate": 3.6823156102651225e-06, + "loss": 0.75143838, + "num_input_tokens_seen": 36479615, + "step": 1717, + "time_per_iteration": 2.7115001678466797 + }, + { + "auxiliary_loss_clip": 0.01245985, + "auxiliary_loss_mlp": 0.01032882, + "balance_loss_clip": 1.05374217, + "balance_loss_mlp": 1.0229218, + "epoch": 0.20657728611795828, + "flos": 20520363029760.0, + "grad_norm": 4.351238686825377, + "language_loss": 0.71024454, + "learning_rate": 3.6818942227953257e-06, + "loss": 0.73303318, + "num_input_tokens_seen": 36500160, + "step": 1718, + "time_per_iteration": 2.753232955932617 + }, + { + "auxiliary_loss_clip": 0.01244171, + "auxiliary_loss_mlp": 0.01030159, + "balance_loss_clip": 1.05881429, + "balance_loss_mlp": 1.02090216, + "epoch": 0.20669752900859736, + "flos": 21799285752960.0, + "grad_norm": 1.985286160191967, + "language_loss": 0.69080371, + "learning_rate": 3.681472580183512e-06, + "loss": 0.71354699, + "num_input_tokens_seen": 36518810, + "step": 1719, + "time_per_iteration": 2.726015567779541 + }, + { + "auxiliary_loss_clip": 0.01220456, + "auxiliary_loss_mlp": 0.01032363, + "balance_loss_clip": 1.06083369, + "balance_loss_mlp": 1.0231781, + "epoch": 0.20681777189923645, + "flos": 15122020014720.0, + "grad_norm": 1.8364068734968442, + "language_loss": 0.86344755, + "learning_rate": 3.6810506824936455e-06, + "loss": 0.88597578, + "num_input_tokens_seen": 36536890, + "step": 1720, + "time_per_iteration": 2.6808347702026367 + }, + { + "auxiliary_loss_clip": 0.01132732, + "auxiliary_loss_mlp": 0.01002177, + "balance_loss_clip": 1.02557278, + "balance_loss_mlp": 0.9996382, + "epoch": 0.20693801478987556, + "flos": 56481021509760.0, + "grad_norm": 1.0509848131591029, + "language_loss": 0.62471408, + "learning_rate": 3.680628529789726e-06, + "loss": 0.64606315, + "num_input_tokens_seen": 36589300, + "step": 1721, + "time_per_iteration": 3.0779993534088135 + }, + { + "auxiliary_loss_clip": 0.01222654, + "auxiliary_loss_mlp": 0.01037806, + "balance_loss_clip": 1.06362307, + "balance_loss_mlp": 1.02805448, + "epoch": 0.20705825768051464, + "flos": 21614201948160.0, + "grad_norm": 2.2447965132094883, + "language_loss": 0.86339521, + "learning_rate": 3.680206122135796e-06, + "loss": 0.8859998, + "num_input_tokens_seen": 36609905, + "step": 1722, + "time_per_iteration": 2.6886162757873535 + }, + { + "auxiliary_loss_clip": 0.01257624, + "auxiliary_loss_mlp": 0.0103953, + "balance_loss_clip": 1.05960381, + "balance_loss_mlp": 1.03117347, + "epoch": 0.20717850057115372, + "flos": 25848895962240.0, + "grad_norm": 1.7931324149347274, + "language_loss": 0.78592509, + "learning_rate": 3.6797834595959323e-06, + "loss": 0.80889666, + "num_input_tokens_seen": 36629805, + "step": 1723, + "time_per_iteration": 2.871598720550537 + }, + { + "auxiliary_loss_clip": 0.01236821, + "auxiliary_loss_mlp": 0.01033182, + "balance_loss_clip": 1.05223334, + "balance_loss_mlp": 1.02297211, + "epoch": 0.20729874346179283, + "flos": 29130807767040.0, + "grad_norm": 3.0215445609276705, + "language_loss": 0.77768546, + "learning_rate": 3.679360542234254e-06, + "loss": 0.80038548, + "num_input_tokens_seen": 36649150, + "step": 1724, + "time_per_iteration": 2.881413698196411 + }, + { + "auxiliary_loss_clip": 0.01220987, + "auxiliary_loss_mlp": 0.01380533, + "balance_loss_clip": 1.05317068, + "balance_loss_mlp": 1.00028729, + "epoch": 0.20741898635243192, + "flos": 29023363209600.0, + "grad_norm": 2.229693836552283, + "language_loss": 0.72339016, + "learning_rate": 3.678937370114916e-06, + "loss": 0.74940538, + "num_input_tokens_seen": 36668955, + "step": 1725, + "time_per_iteration": 2.828139066696167 + }, + { + "auxiliary_loss_clip": 0.01229418, + "auxiliary_loss_mlp": 0.01030561, + "balance_loss_clip": 1.06013489, + "balance_loss_mlp": 1.02133393, + "epoch": 0.207539229243071, + "flos": 15559447841280.0, + "grad_norm": 2.3011190599313336, + "language_loss": 0.78759241, + "learning_rate": 3.678513943302114e-06, + "loss": 0.81019223, + "num_input_tokens_seen": 36685730, + "step": 1726, + "time_per_iteration": 2.692218780517578 + }, + { + "auxiliary_loss_clip": 0.01215244, + "auxiliary_loss_mlp": 0.01029565, + "balance_loss_clip": 1.06122386, + "balance_loss_mlp": 1.02039182, + "epoch": 0.20765947213371008, + "flos": 20521081301760.0, + "grad_norm": 1.7961623068102042, + "language_loss": 0.84823549, + "learning_rate": 3.678090261860082e-06, + "loss": 0.87068355, + "num_input_tokens_seen": 36705460, + "step": 1727, + "time_per_iteration": 2.8322393894195557 + }, + { + "auxiliary_loss_clip": 0.0124003, + "auxiliary_loss_mlp": 0.01032573, + "balance_loss_clip": 1.0508889, + "balance_loss_mlp": 1.02367139, + "epoch": 0.2077797150243492, + "flos": 19354415558400.0, + "grad_norm": 2.9934962045390554, + "language_loss": 0.77788132, + "learning_rate": 3.6776663258530906e-06, + "loss": 0.80060738, + "num_input_tokens_seen": 36724110, + "step": 1728, + "time_per_iteration": 2.8313746452331543 + }, + { + "auxiliary_loss_clip": 0.01224008, + "auxiliary_loss_mlp": 0.01031955, + "balance_loss_clip": 1.0586555, + "balance_loss_mlp": 1.02337766, + "epoch": 0.20789995791498828, + "flos": 21829952989440.0, + "grad_norm": 2.0443501698373185, + "language_loss": 0.71404755, + "learning_rate": 3.6772421353454516e-06, + "loss": 0.73660719, + "num_input_tokens_seen": 36742705, + "step": 1729, + "time_per_iteration": 2.763543128967285 + }, + { + "auxiliary_loss_clip": 0.01222934, + "auxiliary_loss_mlp": 0.01036929, + "balance_loss_clip": 1.06122661, + "balance_loss_mlp": 1.02782106, + "epoch": 0.20802020080562736, + "flos": 23148844571520.0, + "grad_norm": 2.112236037526336, + "language_loss": 0.887586, + "learning_rate": 3.6768176904015153e-06, + "loss": 0.91018456, + "num_input_tokens_seen": 36762510, + "step": 1730, + "time_per_iteration": 3.6078357696533203 + }, + { + "auxiliary_loss_clip": 0.01224119, + "auxiliary_loss_mlp": 0.01038128, + "balance_loss_clip": 1.0575068, + "balance_loss_mlp": 1.0285852, + "epoch": 0.20814044369626647, + "flos": 23072677781760.0, + "grad_norm": 2.571166942554245, + "language_loss": 0.60000157, + "learning_rate": 3.6763929910856674e-06, + "loss": 0.62262404, + "num_input_tokens_seen": 36780960, + "step": 1731, + "time_per_iteration": 3.697690963745117 + }, + { + "auxiliary_loss_clip": 0.01221671, + "auxiliary_loss_mlp": 0.01037716, + "balance_loss_clip": 1.05956364, + "balance_loss_mlp": 1.0280242, + "epoch": 0.20826068658690555, + "flos": 19608016556160.0, + "grad_norm": 3.0309997895140053, + "language_loss": 0.77646172, + "learning_rate": 3.6759680374623365e-06, + "loss": 0.79905558, + "num_input_tokens_seen": 36798875, + "step": 1732, + "time_per_iteration": 2.6495907306671143 + }, + { + "auxiliary_loss_clip": 0.01215082, + "auxiliary_loss_mlp": 0.01037152, + "balance_loss_clip": 1.06244564, + "balance_loss_mlp": 1.02722228, + "epoch": 0.20838092947754464, + "flos": 25374049142400.0, + "grad_norm": 2.9463054671417575, + "language_loss": 0.75585186, + "learning_rate": 3.675542829595986e-06, + "loss": 0.77837425, + "num_input_tokens_seen": 36818540, + "step": 1733, + "time_per_iteration": 3.5363049507141113 + }, + { + "auxiliary_loss_clip": 0.01232805, + "auxiliary_loss_mlp": 0.01030592, + "balance_loss_clip": 1.05776668, + "balance_loss_mlp": 1.02158594, + "epoch": 0.20850117236818372, + "flos": 24061729749120.0, + "grad_norm": 1.6498950994402546, + "language_loss": 0.79324609, + "learning_rate": 3.6751173675511213e-06, + "loss": 0.81588006, + "num_input_tokens_seen": 36840585, + "step": 1734, + "time_per_iteration": 2.772258758544922 + }, + { + "auxiliary_loss_clip": 0.01223242, + "auxiliary_loss_mlp": 0.01034546, + "balance_loss_clip": 1.05251026, + "balance_loss_mlp": 1.02488422, + "epoch": 0.20862141525882283, + "flos": 20077799558400.0, + "grad_norm": 2.044598485763772, + "language_loss": 0.87536836, + "learning_rate": 3.674691651392283e-06, + "loss": 0.89794624, + "num_input_tokens_seen": 36858255, + "step": 1735, + "time_per_iteration": 3.575350522994995 + }, + { + "auxiliary_loss_clip": 0.01237822, + "auxiliary_loss_mlp": 0.01036681, + "balance_loss_clip": 1.06055665, + "balance_loss_mlp": 1.02759707, + "epoch": 0.2087416581494619, + "flos": 39015183237120.0, + "grad_norm": 2.0230579263205204, + "language_loss": 0.75629401, + "learning_rate": 3.674265681184053e-06, + "loss": 0.77903903, + "num_input_tokens_seen": 36881515, + "step": 1736, + "time_per_iteration": 2.838794231414795 + }, + { + "auxiliary_loss_clip": 0.01231088, + "auxiliary_loss_mlp": 0.01028807, + "balance_loss_clip": 1.05718422, + "balance_loss_mlp": 1.01899016, + "epoch": 0.208861901040101, + "flos": 26101994169600.0, + "grad_norm": 2.1919260749680136, + "language_loss": 0.86297548, + "learning_rate": 3.6738394569910504e-06, + "loss": 0.88557434, + "num_input_tokens_seen": 36902055, + "step": 1737, + "time_per_iteration": 2.8875184059143066 + }, + { + "auxiliary_loss_clip": 0.01222236, + "auxiliary_loss_mlp": 0.01041861, + "balance_loss_clip": 1.0598762, + "balance_loss_mlp": 1.03246093, + "epoch": 0.2089821439307401, + "flos": 28398732675840.0, + "grad_norm": 1.9662434342175301, + "language_loss": 0.82743174, + "learning_rate": 3.6734129788779333e-06, + "loss": 0.85007274, + "num_input_tokens_seen": 36921230, + "step": 1738, + "time_per_iteration": 2.6714961528778076 + }, + { + "auxiliary_loss_clip": 0.0123616, + "auxiliary_loss_mlp": 0.01034568, + "balance_loss_clip": 1.05604124, + "balance_loss_mlp": 1.02544808, + "epoch": 0.2091023868213792, + "flos": 21069616872960.0, + "grad_norm": 2.079249973299627, + "language_loss": 0.9017753, + "learning_rate": 3.6729862469093976e-06, + "loss": 0.92448258, + "num_input_tokens_seen": 36940325, + "step": 1739, + "time_per_iteration": 2.790480852127075 + }, + { + "auxiliary_loss_clip": 0.01216215, + "auxiliary_loss_mlp": 0.01038087, + "balance_loss_clip": 1.05389333, + "balance_loss_mlp": 1.02948022, + "epoch": 0.20922262971201827, + "flos": 22455481363200.0, + "grad_norm": 2.455299128668109, + "language_loss": 0.82581925, + "learning_rate": 3.6725592611501782e-06, + "loss": 0.84836227, + "num_input_tokens_seen": 36959000, + "step": 1740, + "time_per_iteration": 2.683302164077759 + }, + { + "auxiliary_loss_clip": 0.01220143, + "auxiliary_loss_mlp": 0.01030295, + "balance_loss_clip": 1.0566299, + "balance_loss_mlp": 1.02129471, + "epoch": 0.20934287260265738, + "flos": 27852244179840.0, + "grad_norm": 1.7416393594677342, + "language_loss": 0.76498348, + "learning_rate": 3.6721320216650496e-06, + "loss": 0.78748786, + "num_input_tokens_seen": 36979615, + "step": 1741, + "time_per_iteration": 2.8758912086486816 + }, + { + "auxiliary_loss_clip": 0.01228455, + "auxiliary_loss_mlp": 0.01038873, + "balance_loss_clip": 1.05751371, + "balance_loss_mlp": 1.02959299, + "epoch": 0.20946311549329646, + "flos": 16435309075200.0, + "grad_norm": 1.829888515979776, + "language_loss": 0.83142847, + "learning_rate": 3.6717045285188215e-06, + "loss": 0.85410178, + "num_input_tokens_seen": 36997310, + "step": 1742, + "time_per_iteration": 2.628352403640747 + }, + { + "auxiliary_loss_clip": 0.01228107, + "auxiliary_loss_mlp": 0.01038867, + "balance_loss_clip": 1.04944181, + "balance_loss_mlp": 1.0283941, + "epoch": 0.20958335838393555, + "flos": 22492720788480.0, + "grad_norm": 2.186255444461003, + "language_loss": 0.86787164, + "learning_rate": 3.671276781776346e-06, + "loss": 0.89054143, + "num_input_tokens_seen": 37015965, + "step": 1743, + "time_per_iteration": 2.7862184047698975 + }, + { + "auxiliary_loss_clip": 0.01247651, + "auxiliary_loss_mlp": 0.01027117, + "balance_loss_clip": 1.05567586, + "balance_loss_mlp": 1.01831973, + "epoch": 0.20970360127457463, + "flos": 25224768218880.0, + "grad_norm": 1.9365225997752433, + "language_loss": 0.67002469, + "learning_rate": 3.6708487815025128e-06, + "loss": 0.69277239, + "num_input_tokens_seen": 37036545, + "step": 1744, + "time_per_iteration": 2.73175048828125 + }, + { + "auxiliary_loss_clip": 0.01239857, + "auxiliary_loss_mlp": 0.01027289, + "balance_loss_clip": 1.05524611, + "balance_loss_mlp": 1.01799107, + "epoch": 0.20982384416521374, + "flos": 18479164855680.0, + "grad_norm": 2.3013552465615614, + "language_loss": 0.7443161, + "learning_rate": 3.6704205277622463e-06, + "loss": 0.76698756, + "num_input_tokens_seen": 37054985, + "step": 1745, + "time_per_iteration": 2.675903558731079 + }, + { + "auxiliary_loss_clip": 0.01232533, + "auxiliary_loss_mlp": 0.01026832, + "balance_loss_clip": 1.05457854, + "balance_loss_mlp": 1.01758707, + "epoch": 0.20994408705585282, + "flos": 25373546352000.0, + "grad_norm": 1.7272373906768081, + "language_loss": 0.80286402, + "learning_rate": 3.6699920206205146e-06, + "loss": 0.82545769, + "num_input_tokens_seen": 37075725, + "step": 1746, + "time_per_iteration": 2.756683826446533 + }, + { + "auxiliary_loss_clip": 0.01222178, + "auxiliary_loss_mlp": 0.01031556, + "balance_loss_clip": 1.05656481, + "balance_loss_mlp": 1.02256119, + "epoch": 0.2100643299464919, + "flos": 21320955313920.0, + "grad_norm": 1.680525683122723, + "language_loss": 0.81890762, + "learning_rate": 3.669563260142321e-06, + "loss": 0.84144497, + "num_input_tokens_seen": 37094615, + "step": 1747, + "time_per_iteration": 2.63472318649292 + }, + { + "auxiliary_loss_clip": 0.01228556, + "auxiliary_loss_mlp": 0.01034949, + "balance_loss_clip": 1.05881703, + "balance_loss_mlp": 1.02557302, + "epoch": 0.21018457283713102, + "flos": 19354379644800.0, + "grad_norm": 2.3140205730345285, + "language_loss": 0.84336317, + "learning_rate": 3.6691342463927083e-06, + "loss": 0.86599815, + "num_input_tokens_seen": 37113610, + "step": 1748, + "time_per_iteration": 2.6958675384521484 + }, + { + "auxiliary_loss_clip": 0.01241993, + "auxiliary_loss_mlp": 0.01033656, + "balance_loss_clip": 1.05706799, + "balance_loss_mlp": 1.02422094, + "epoch": 0.2103048157277701, + "flos": 28330035914880.0, + "grad_norm": 1.745409069460149, + "language_loss": 0.81553441, + "learning_rate": 3.668704979436758e-06, + "loss": 0.83829087, + "num_input_tokens_seen": 37133705, + "step": 1749, + "time_per_iteration": 2.899087429046631 + }, + { + "auxiliary_loss_clip": 0.01219994, + "auxiliary_loss_mlp": 0.01038438, + "balance_loss_clip": 1.05369759, + "balance_loss_mlp": 1.02896714, + "epoch": 0.21042505861840918, + "flos": 17457290835840.0, + "grad_norm": 2.2458771167558496, + "language_loss": 0.78824258, + "learning_rate": 3.668275459339588e-06, + "loss": 0.8108269, + "num_input_tokens_seen": 37152185, + "step": 1750, + "time_per_iteration": 2.704240322113037 + }, + { + "auxiliary_loss_clip": 0.01214985, + "auxiliary_loss_mlp": 0.0103339, + "balance_loss_clip": 1.06172228, + "balance_loss_mlp": 1.0238353, + "epoch": 0.21054530150904827, + "flos": 14209817195520.0, + "grad_norm": 1.8406860962143758, + "language_loss": 0.80470181, + "learning_rate": 3.667845686166358e-06, + "loss": 0.82718557, + "num_input_tokens_seen": 37169110, + "step": 1751, + "time_per_iteration": 2.6205079555511475 + }, + { + "auxiliary_loss_clip": 0.01217093, + "auxiliary_loss_mlp": 0.01039546, + "balance_loss_clip": 1.05024791, + "balance_loss_mlp": 1.0294075, + "epoch": 0.21066554439968738, + "flos": 18618210403200.0, + "grad_norm": 1.6349596825860317, + "language_loss": 0.85838467, + "learning_rate": 3.6674156599822634e-06, + "loss": 0.88095105, + "num_input_tokens_seen": 37184905, + "step": 1752, + "time_per_iteration": 2.6915578842163086 + }, + { + "auxiliary_loss_clip": 0.01247202, + "auxiliary_loss_mlp": 0.01039286, + "balance_loss_clip": 1.05106592, + "balance_loss_mlp": 1.02918291, + "epoch": 0.21078578729032646, + "flos": 23658883741440.0, + "grad_norm": 1.9400939712196446, + "language_loss": 0.81860054, + "learning_rate": 3.666985380852539e-06, + "loss": 0.84146547, + "num_input_tokens_seen": 37203910, + "step": 1753, + "time_per_iteration": 2.8165125846862793 + }, + { + "auxiliary_loss_clip": 0.01232069, + "auxiliary_loss_mlp": 0.0103214, + "balance_loss_clip": 1.05866754, + "balance_loss_mlp": 1.02166212, + "epoch": 0.21090603018096554, + "flos": 29346379240320.0, + "grad_norm": 3.117570726734395, + "language_loss": 0.74226922, + "learning_rate": 3.6665548488424576e-06, + "loss": 0.76491129, + "num_input_tokens_seen": 37222670, + "step": 1754, + "time_per_iteration": 2.8489954471588135 + }, + { + "auxiliary_loss_clip": 0.01215733, + "auxiliary_loss_mlp": 0.01037708, + "balance_loss_clip": 1.06028247, + "balance_loss_mlp": 1.02770638, + "epoch": 0.21102627307160465, + "flos": 23261245205760.0, + "grad_norm": 1.7812082553558208, + "language_loss": 0.87757283, + "learning_rate": 3.6661240640173307e-06, + "loss": 0.90010726, + "num_input_tokens_seen": 37244140, + "step": 1755, + "time_per_iteration": 2.71250057220459 + }, + { + "auxiliary_loss_clip": 0.0115193, + "auxiliary_loss_mlp": 0.01002928, + "balance_loss_clip": 1.03079259, + "balance_loss_mlp": 1.00056744, + "epoch": 0.21114651596224374, + "flos": 54633454577280.0, + "grad_norm": 1.097682489664134, + "language_loss": 0.5786612, + "learning_rate": 3.6656930264425085e-06, + "loss": 0.60020983, + "num_input_tokens_seen": 37308185, + "step": 1756, + "time_per_iteration": 4.308766841888428 + }, + { + "auxiliary_loss_clip": 0.01213926, + "auxiliary_loss_mlp": 0.01034633, + "balance_loss_clip": 1.06075478, + "balance_loss_mlp": 1.02495885, + "epoch": 0.21126675885288282, + "flos": 21543314457600.0, + "grad_norm": 1.8099711384480919, + "language_loss": 0.75580102, + "learning_rate": 3.665261736183378e-06, + "loss": 0.77828658, + "num_input_tokens_seen": 37328220, + "step": 1757, + "time_per_iteration": 3.487778425216675 + }, + { + "auxiliary_loss_clip": 0.01240477, + "auxiliary_loss_mlp": 0.01034883, + "balance_loss_clip": 1.05633342, + "balance_loss_mlp": 1.02470219, + "epoch": 0.2113870017435219, + "flos": 10961876678400.0, + "grad_norm": 3.538529684033641, + "language_loss": 0.88607764, + "learning_rate": 3.664830193305366e-06, + "loss": 0.90883124, + "num_input_tokens_seen": 37345995, + "step": 1758, + "time_per_iteration": 2.710068941116333 + }, + { + "auxiliary_loss_clip": 0.01230192, + "auxiliary_loss_mlp": 0.01038347, + "balance_loss_clip": 1.05314648, + "balance_loss_mlp": 1.02871525, + "epoch": 0.211507244634161, + "flos": 16653825463680.0, + "grad_norm": 7.461777193621392, + "language_loss": 0.77198309, + "learning_rate": 3.6643983978739373e-06, + "loss": 0.7946685, + "num_input_tokens_seen": 37362610, + "step": 1759, + "time_per_iteration": 3.619699239730835 + }, + { + "auxiliary_loss_clip": 0.01222799, + "auxiliary_loss_mlp": 0.01033974, + "balance_loss_clip": 1.05752265, + "balance_loss_mlp": 1.02397275, + "epoch": 0.2116274875248001, + "flos": 20954091755520.0, + "grad_norm": 2.236918117276062, + "language_loss": 0.82176417, + "learning_rate": 3.663966349954596e-06, + "loss": 0.84433192, + "num_input_tokens_seen": 37382790, + "step": 1760, + "time_per_iteration": 2.695450782775879 + }, + { + "auxiliary_loss_clip": 0.01123868, + "auxiliary_loss_mlp": 0.01004507, + "balance_loss_clip": 1.02749193, + "balance_loss_mlp": 1.00203979, + "epoch": 0.21174773041543918, + "flos": 68196949424640.0, + "grad_norm": 0.7853055894862541, + "language_loss": 0.59746659, + "learning_rate": 3.6635340496128816e-06, + "loss": 0.61875033, + "num_input_tokens_seen": 37439720, + "step": 1761, + "time_per_iteration": 4.008767604827881 + }, + { + "auxiliary_loss_clip": 0.01244429, + "auxiliary_loss_mlp": 0.01028977, + "balance_loss_clip": 1.05399132, + "balance_loss_mlp": 1.01998305, + "epoch": 0.2118679733060783, + "flos": 20668315150080.0, + "grad_norm": 1.7102704287872565, + "language_loss": 0.9279002, + "learning_rate": 3.6631014969143747e-06, + "loss": 0.9506343, + "num_input_tokens_seen": 37459410, + "step": 1762, + "time_per_iteration": 2.8307950496673584 + }, + { + "auxiliary_loss_clip": 0.01224456, + "auxiliary_loss_mlp": 0.0103102, + "balance_loss_clip": 1.06172276, + "balance_loss_mlp": 1.02173364, + "epoch": 0.21198821619671737, + "flos": 23223431162880.0, + "grad_norm": 1.8637160701209043, + "language_loss": 0.88982761, + "learning_rate": 3.662668691924693e-06, + "loss": 0.91238236, + "num_input_tokens_seen": 37480460, + "step": 1763, + "time_per_iteration": 2.789374351501465 + }, + { + "auxiliary_loss_clip": 0.01235881, + "auxiliary_loss_mlp": 0.01040181, + "balance_loss_clip": 1.05337572, + "balance_loss_mlp": 1.02991688, + "epoch": 0.21210845908735645, + "flos": 24498547044480.0, + "grad_norm": 2.3939841786890237, + "language_loss": 0.71371549, + "learning_rate": 3.6622356347094927e-06, + "loss": 0.73647612, + "num_input_tokens_seen": 37502025, + "step": 1764, + "time_per_iteration": 2.748342514038086 + }, + { + "auxiliary_loss_clip": 0.01237703, + "auxiliary_loss_mlp": 0.01034798, + "balance_loss_clip": 1.05335641, + "balance_loss_mlp": 1.02490973, + "epoch": 0.21222870197799554, + "flos": 27089789160960.0, + "grad_norm": 2.2072210267969354, + "language_loss": 0.79009843, + "learning_rate": 3.6618023253344684e-06, + "loss": 0.81282341, + "num_input_tokens_seen": 37520885, + "step": 1765, + "time_per_iteration": 2.7819647789001465 + }, + { + "auxiliary_loss_clip": 0.01220542, + "auxiliary_loss_mlp": 0.01039891, + "balance_loss_clip": 1.058025, + "balance_loss_mlp": 1.0298717, + "epoch": 0.21234894486863465, + "flos": 16873850223360.0, + "grad_norm": 1.488881741644399, + "language_loss": 0.83984256, + "learning_rate": 3.6613687638653527e-06, + "loss": 0.86244696, + "num_input_tokens_seen": 37539055, + "step": 1766, + "time_per_iteration": 2.6272528171539307 + }, + { + "auxiliary_loss_clip": 0.01228121, + "auxiliary_loss_mlp": 0.01029403, + "balance_loss_clip": 1.05773926, + "balance_loss_mlp": 1.01960421, + "epoch": 0.21246918775927373, + "flos": 23474949171840.0, + "grad_norm": 2.0156313298486426, + "language_loss": 0.77913618, + "learning_rate": 3.660934950367916e-06, + "loss": 0.80171144, + "num_input_tokens_seen": 37558300, + "step": 1767, + "time_per_iteration": 2.7634568214416504 + }, + { + "auxiliary_loss_clip": 0.01224632, + "auxiliary_loss_mlp": 0.01028682, + "balance_loss_clip": 1.05776024, + "balance_loss_mlp": 1.01893115, + "epoch": 0.21258943064991281, + "flos": 22382295402240.0, + "grad_norm": 1.8918554797411447, + "language_loss": 0.8345437, + "learning_rate": 3.660500884907968e-06, + "loss": 0.85707688, + "num_input_tokens_seen": 37579040, + "step": 1768, + "time_per_iteration": 2.6720807552337646 + }, + { + "auxiliary_loss_clip": 0.01132716, + "auxiliary_loss_mlp": 0.01005951, + "balance_loss_clip": 1.02523327, + "balance_loss_mlp": 1.00356698, + "epoch": 0.21270967354055192, + "flos": 59440168679040.0, + "grad_norm": 0.8206762473410629, + "language_loss": 0.60044813, + "learning_rate": 3.660066567551356e-06, + "loss": 0.62183487, + "num_input_tokens_seen": 37639185, + "step": 1769, + "time_per_iteration": 3.24772047996521 + }, + { + "auxiliary_loss_clip": 0.01223574, + "auxiliary_loss_mlp": 0.01380351, + "balance_loss_clip": 1.05872178, + "balance_loss_mlp": 1.00004494, + "epoch": 0.212829916431191, + "flos": 21544032729600.0, + "grad_norm": 2.674773976000037, + "language_loss": 0.84477359, + "learning_rate": 3.6596319983639657e-06, + "loss": 0.87081283, + "num_input_tokens_seen": 37657765, + "step": 1770, + "time_per_iteration": 2.94575834274292 + }, + { + "auxiliary_loss_clip": 0.01237517, + "auxiliary_loss_mlp": 0.01380676, + "balance_loss_clip": 1.05553722, + "balance_loss_mlp": 0.99997753, + "epoch": 0.2129501593218301, + "flos": 28987739896320.0, + "grad_norm": 1.728529928077784, + "language_loss": 0.86444926, + "learning_rate": 3.6591971774117214e-06, + "loss": 0.8906312, + "num_input_tokens_seen": 37680740, + "step": 1771, + "time_per_iteration": 2.8383092880249023 + }, + { + "auxiliary_loss_clip": 0.01227571, + "auxiliary_loss_mlp": 0.01033917, + "balance_loss_clip": 1.06039929, + "balance_loss_mlp": 1.02358162, + "epoch": 0.2130704022124692, + "flos": 18806993308800.0, + "grad_norm": 2.4253339721415825, + "language_loss": 0.80235076, + "learning_rate": 3.6587621047605833e-06, + "loss": 0.8249656, + "num_input_tokens_seen": 37697910, + "step": 1772, + "time_per_iteration": 2.791328191757202 + }, + { + "auxiliary_loss_clip": 0.01221666, + "auxiliary_loss_mlp": 0.01035669, + "balance_loss_clip": 1.05904317, + "balance_loss_mlp": 1.02557766, + "epoch": 0.21319064510310828, + "flos": 13918150759680.0, + "grad_norm": 2.086305689218926, + "language_loss": 0.86563015, + "learning_rate": 3.6583267804765542e-06, + "loss": 0.8882035, + "num_input_tokens_seen": 37712245, + "step": 1773, + "time_per_iteration": 2.5965754985809326 + }, + { + "auxiliary_loss_clip": 0.01220545, + "auxiliary_loss_mlp": 0.01034827, + "balance_loss_clip": 1.05895734, + "balance_loss_mlp": 1.02409816, + "epoch": 0.21331088799374737, + "flos": 20959694277120.0, + "grad_norm": 1.916068208843687, + "language_loss": 0.85588855, + "learning_rate": 3.6578912046256702e-06, + "loss": 0.87844229, + "num_input_tokens_seen": 37730765, + "step": 1774, + "time_per_iteration": 2.6819682121276855 + }, + { + "auxiliary_loss_clip": 0.01234051, + "auxiliary_loss_mlp": 0.0103587, + "balance_loss_clip": 1.05392396, + "balance_loss_mlp": 1.02639306, + "epoch": 0.21343113088438645, + "flos": 18624638937600.0, + "grad_norm": 1.9962284224132958, + "language_loss": 0.76466525, + "learning_rate": 3.6574553772740083e-06, + "loss": 0.78736448, + "num_input_tokens_seen": 37748695, + "step": 1775, + "time_per_iteration": 2.6843550205230713 + }, + { + "auxiliary_loss_clip": 0.01119509, + "auxiliary_loss_mlp": 0.01012622, + "balance_loss_clip": 1.02651954, + "balance_loss_mlp": 1.01038063, + "epoch": 0.21355137377502556, + "flos": 67413128791680.0, + "grad_norm": 0.875678523918047, + "language_loss": 0.61874801, + "learning_rate": 3.657019298487684e-06, + "loss": 0.64006931, + "num_input_tokens_seen": 37813705, + "step": 1776, + "time_per_iteration": 3.2038559913635254 + }, + { + "auxiliary_loss_clip": 0.0122794, + "auxiliary_loss_mlp": 0.01380754, + "balance_loss_clip": 1.05775261, + "balance_loss_mlp": 1.00018775, + "epoch": 0.21367161666566464, + "flos": 34532095697280.0, + "grad_norm": 1.6679496080694898, + "language_loss": 0.83642101, + "learning_rate": 3.6565829683328495e-06, + "loss": 0.86250794, + "num_input_tokens_seen": 37836330, + "step": 1777, + "time_per_iteration": 2.799772262573242 + }, + { + "auxiliary_loss_clip": 0.0121882, + "auxiliary_loss_mlp": 0.01039646, + "balance_loss_clip": 1.05794728, + "balance_loss_mlp": 1.02978182, + "epoch": 0.21379185955630373, + "flos": 18989347680000.0, + "grad_norm": 2.5984386321701245, + "language_loss": 0.85855377, + "learning_rate": 3.6561463868756965e-06, + "loss": 0.88113844, + "num_input_tokens_seen": 37855030, + "step": 1778, + "time_per_iteration": 2.624549150466919 + }, + { + "auxiliary_loss_clip": 0.01224844, + "auxiliary_loss_mlp": 0.01026776, + "balance_loss_clip": 1.06271684, + "balance_loss_mlp": 1.017102, + "epoch": 0.21391210244694284, + "flos": 28218497207040.0, + "grad_norm": 1.573267200337749, + "language_loss": 0.77976197, + "learning_rate": 3.655709554182452e-06, + "loss": 0.80227816, + "num_input_tokens_seen": 37875370, + "step": 1779, + "time_per_iteration": 2.739389419555664 + }, + { + "auxiliary_loss_clip": 0.01224879, + "auxiliary_loss_mlp": 0.01030266, + "balance_loss_clip": 1.05898547, + "balance_loss_mlp": 1.02055085, + "epoch": 0.21403234533758192, + "flos": 17455064192640.0, + "grad_norm": 2.687518342181109, + "language_loss": 0.84417909, + "learning_rate": 3.6552724703193855e-06, + "loss": 0.86673057, + "num_input_tokens_seen": 37892560, + "step": 1780, + "time_per_iteration": 2.6107468605041504 + }, + { + "auxiliary_loss_clip": 0.01142878, + "auxiliary_loss_mlp": 0.01010723, + "balance_loss_clip": 1.01648903, + "balance_loss_mlp": 1.00835085, + "epoch": 0.214152588228221, + "flos": 51637606686720.0, + "grad_norm": 0.792306588289245, + "language_loss": 0.55949461, + "learning_rate": 3.654835135352801e-06, + "loss": 0.58103061, + "num_input_tokens_seen": 37947370, + "step": 1781, + "time_per_iteration": 3.242400884628296 + }, + { + "auxiliary_loss_clip": 0.01245194, + "auxiliary_loss_mlp": 0.01028788, + "balance_loss_clip": 1.04969084, + "balance_loss_mlp": 1.01843452, + "epoch": 0.21427283111886009, + "flos": 19496154625920.0, + "grad_norm": 1.839016532686822, + "language_loss": 0.87370718, + "learning_rate": 3.654397549349043e-06, + "loss": 0.89644694, + "num_input_tokens_seen": 37964745, + "step": 1782, + "time_per_iteration": 2.865816593170166 + }, + { + "auxiliary_loss_clip": 0.01230134, + "auxiliary_loss_mlp": 0.0103845, + "balance_loss_clip": 1.05908608, + "balance_loss_mlp": 1.02818632, + "epoch": 0.2143930740094992, + "flos": 20084802710400.0, + "grad_norm": 2.1611332833243235, + "language_loss": 0.75155401, + "learning_rate": 3.653959712374491e-06, + "loss": 0.77423984, + "num_input_tokens_seen": 37982850, + "step": 1783, + "time_per_iteration": 3.5566165447235107 + }, + { + "auxiliary_loss_clip": 0.01234313, + "auxiliary_loss_mlp": 0.01034583, + "balance_loss_clip": 1.05682182, + "balance_loss_mlp": 1.02449799, + "epoch": 0.21451331690013828, + "flos": 21798603394560.0, + "grad_norm": 1.66896348586263, + "language_loss": 0.82644445, + "learning_rate": 3.6535216244955663e-06, + "loss": 0.84913337, + "num_input_tokens_seen": 38002745, + "step": 1784, + "time_per_iteration": 3.6992111206054688 + }, + { + "auxiliary_loss_clip": 0.01230998, + "auxiliary_loss_mlp": 0.01038526, + "balance_loss_clip": 1.05749655, + "balance_loss_mlp": 1.02841103, + "epoch": 0.21463355979077736, + "flos": 32853882412800.0, + "grad_norm": 1.9838888055322719, + "language_loss": 0.7100879, + "learning_rate": 3.653083285778726e-06, + "loss": 0.7327832, + "num_input_tokens_seen": 38024115, + "step": 1785, + "time_per_iteration": 3.709148406982422 + }, + { + "auxiliary_loss_clip": 0.01226585, + "auxiliary_loss_mlp": 0.01038649, + "balance_loss_clip": 1.05874133, + "balance_loss_mlp": 1.02753234, + "epoch": 0.21475380268141647, + "flos": 21543817248000.0, + "grad_norm": 1.996781254650364, + "language_loss": 0.81474811, + "learning_rate": 3.6526446962904653e-06, + "loss": 0.83740044, + "num_input_tokens_seen": 38042830, + "step": 1786, + "time_per_iteration": 2.6390113830566406 + }, + { + "auxiliary_loss_clip": 0.01217821, + "auxiliary_loss_mlp": 0.01036884, + "balance_loss_clip": 1.06009865, + "balance_loss_mlp": 1.02663195, + "epoch": 0.21487404557205556, + "flos": 32159082660480.0, + "grad_norm": 1.5149928798394132, + "language_loss": 0.74106508, + "learning_rate": 3.652205856097318e-06, + "loss": 0.76361215, + "num_input_tokens_seen": 38066015, + "step": 1787, + "time_per_iteration": 3.681015729904175 + }, + { + "auxiliary_loss_clip": 0.01249541, + "auxiliary_loss_mlp": 0.01380459, + "balance_loss_clip": 1.05448842, + "balance_loss_mlp": 1.0001235, + "epoch": 0.21499428846269464, + "flos": 12673091583360.0, + "grad_norm": 1.9177496654258854, + "language_loss": 0.78996873, + "learning_rate": 3.651766765265856e-06, + "loss": 0.81626868, + "num_input_tokens_seen": 38083025, + "step": 1788, + "time_per_iteration": 2.742708683013916 + }, + { + "auxiliary_loss_clip": 0.01226814, + "auxiliary_loss_mlp": 0.01036086, + "balance_loss_clip": 1.05531454, + "balance_loss_mlp": 1.02545798, + "epoch": 0.21511453135333372, + "flos": 23471573293440.0, + "grad_norm": 2.364569209965115, + "language_loss": 0.80951738, + "learning_rate": 3.65132742386269e-06, + "loss": 0.83214641, + "num_input_tokens_seen": 38098245, + "step": 1789, + "time_per_iteration": 2.7135396003723145 + }, + { + "auxiliary_loss_clip": 0.01213824, + "auxiliary_loss_mlp": 0.01035841, + "balance_loss_clip": 1.0591414, + "balance_loss_mlp": 1.02608991, + "epoch": 0.21523477424397283, + "flos": 26943560893440.0, + "grad_norm": 1.9995874536574045, + "language_loss": 0.84983164, + "learning_rate": 3.6508878319544656e-06, + "loss": 0.87232828, + "num_input_tokens_seen": 38118460, + "step": 1790, + "time_per_iteration": 2.6712234020233154 + }, + { + "auxiliary_loss_clip": 0.01223889, + "auxiliary_loss_mlp": 0.01048136, + "balance_loss_clip": 1.0602324, + "balance_loss_mlp": 1.03900445, + "epoch": 0.21535501713461191, + "flos": 18916161719040.0, + "grad_norm": 6.629789956461775, + "language_loss": 0.80908161, + "learning_rate": 3.65044798960787e-06, + "loss": 0.83180189, + "num_input_tokens_seen": 38136800, + "step": 1791, + "time_per_iteration": 2.709784507751465 + }, + { + "auxiliary_loss_clip": 0.01230374, + "auxiliary_loss_mlp": 0.01034935, + "balance_loss_clip": 1.05429339, + "balance_loss_mlp": 1.02509975, + "epoch": 0.215475260025251, + "flos": 17895113712000.0, + "grad_norm": 1.8714719095983225, + "language_loss": 0.7836591, + "learning_rate": 3.650007896889627e-06, + "loss": 0.8063122, + "num_input_tokens_seen": 38155380, + "step": 1792, + "time_per_iteration": 2.682584524154663 + }, + { + "auxiliary_loss_clip": 0.01213964, + "auxiliary_loss_mlp": 0.01037048, + "balance_loss_clip": 1.06262267, + "balance_loss_mlp": 1.02771938, + "epoch": 0.2155955029158901, + "flos": 16654292340480.0, + "grad_norm": 1.7884027180019662, + "language_loss": 0.80595791, + "learning_rate": 3.6495675538664974e-06, + "loss": 0.82846797, + "num_input_tokens_seen": 38174395, + "step": 1793, + "time_per_iteration": 2.5891809463500977 + }, + { + "auxiliary_loss_clip": 0.01228585, + "auxiliary_loss_mlp": 0.01038425, + "balance_loss_clip": 1.05400312, + "balance_loss_mlp": 1.02838135, + "epoch": 0.2157157458065292, + "flos": 23621213352960.0, + "grad_norm": 8.478647292323497, + "language_loss": 0.82539439, + "learning_rate": 3.649126960605282e-06, + "loss": 0.84806448, + "num_input_tokens_seen": 38195380, + "step": 1794, + "time_per_iteration": 2.7079482078552246 + }, + { + "auxiliary_loss_clip": 0.01226553, + "auxiliary_loss_mlp": 0.01039218, + "balance_loss_clip": 1.05698323, + "balance_loss_mlp": 1.029073, + "epoch": 0.21583598869716827, + "flos": 22127078292480.0, + "grad_norm": 2.2474526571204945, + "language_loss": 0.83955151, + "learning_rate": 3.6486861171728174e-06, + "loss": 0.8622092, + "num_input_tokens_seen": 38213775, + "step": 1795, + "time_per_iteration": 2.6305670738220215 + }, + { + "auxiliary_loss_clip": 0.01238848, + "auxiliary_loss_mlp": 0.01037259, + "balance_loss_clip": 1.05231643, + "balance_loss_mlp": 1.02741194, + "epoch": 0.21595623158780738, + "flos": 23441229279360.0, + "grad_norm": 1.6273066111732044, + "language_loss": 0.78352046, + "learning_rate": 3.6482450236359803e-06, + "loss": 0.80628157, + "num_input_tokens_seen": 38235630, + "step": 1796, + "time_per_iteration": 2.766083002090454 + }, + { + "auxiliary_loss_clip": 0.01222411, + "auxiliary_loss_mlp": 0.01035295, + "balance_loss_clip": 1.061306, + "balance_loss_mlp": 1.02598453, + "epoch": 0.21607647447844647, + "flos": 26906501036160.0, + "grad_norm": 2.3444771031456413, + "language_loss": 0.77795768, + "learning_rate": 3.647803680061683e-06, + "loss": 0.80053473, + "num_input_tokens_seen": 38256045, + "step": 1797, + "time_per_iteration": 2.688809633255005 + }, + { + "auxiliary_loss_clip": 0.0123236, + "auxiliary_loss_mlp": 0.01033415, + "balance_loss_clip": 1.05775642, + "balance_loss_mlp": 1.02330041, + "epoch": 0.21619671736908555, + "flos": 14495378319360.0, + "grad_norm": 2.4101488401576368, + "language_loss": 0.74483651, + "learning_rate": 3.6473620865168776e-06, + "loss": 0.76749426, + "num_input_tokens_seen": 38272915, + "step": 1798, + "time_per_iteration": 2.7175748348236084 + }, + { + "auxiliary_loss_clip": 0.01231989, + "auxiliary_loss_mlp": 0.01036121, + "balance_loss_clip": 1.05978894, + "balance_loss_mlp": 1.0255177, + "epoch": 0.21631696025972463, + "flos": 17931096161280.0, + "grad_norm": 2.3461208987934077, + "language_loss": 0.81727052, + "learning_rate": 3.646920243068554e-06, + "loss": 0.83995163, + "num_input_tokens_seen": 38290810, + "step": 1799, + "time_per_iteration": 2.7032835483551025 + }, + { + "auxiliary_loss_clip": 0.01214002, + "auxiliary_loss_mlp": 0.01040479, + "balance_loss_clip": 1.05528986, + "balance_loss_mlp": 1.03085256, + "epoch": 0.21643720315036374, + "flos": 24462385027200.0, + "grad_norm": 1.7884623799066193, + "language_loss": 0.74599463, + "learning_rate": 3.6464781497837384e-06, + "loss": 0.76853949, + "num_input_tokens_seen": 38312785, + "step": 1800, + "time_per_iteration": 2.7159290313720703 + }, + { + "auxiliary_loss_clip": 0.01231663, + "auxiliary_loss_mlp": 0.01036082, + "balance_loss_clip": 1.05522871, + "balance_loss_mlp": 1.02612805, + "epoch": 0.21655744604100283, + "flos": 28474432588800.0, + "grad_norm": 1.6792682311448321, + "language_loss": 0.72715193, + "learning_rate": 3.6460358067294965e-06, + "loss": 0.74982941, + "num_input_tokens_seen": 38334015, + "step": 1801, + "time_per_iteration": 2.719424247741699 + }, + { + "auxiliary_loss_clip": 0.01217808, + "auxiliary_loss_mlp": 0.0103471, + "balance_loss_clip": 1.06095433, + "balance_loss_mlp": 1.0245775, + "epoch": 0.2166776889316419, + "flos": 20152960767360.0, + "grad_norm": 1.8126694068421363, + "language_loss": 0.77393556, + "learning_rate": 3.645593213972932e-06, + "loss": 0.79646081, + "num_input_tokens_seen": 38352920, + "step": 1802, + "time_per_iteration": 2.6482834815979004 + }, + { + "auxiliary_loss_clip": 0.01218152, + "auxiliary_loss_mlp": 0.01031885, + "balance_loss_clip": 1.05908751, + "balance_loss_mlp": 1.02217555, + "epoch": 0.21679793182228102, + "flos": 15193482122880.0, + "grad_norm": 2.3964726937793155, + "language_loss": 0.79838693, + "learning_rate": 3.6451503715811852e-06, + "loss": 0.82088733, + "num_input_tokens_seen": 38371230, + "step": 1803, + "time_per_iteration": 2.6681711673736572 + }, + { + "auxiliary_loss_clip": 0.01230026, + "auxiliary_loss_mlp": 0.01031077, + "balance_loss_clip": 1.05886054, + "balance_loss_mlp": 1.02215469, + "epoch": 0.2169181747129201, + "flos": 17384464010880.0, + "grad_norm": 2.142186005209377, + "language_loss": 0.80549204, + "learning_rate": 3.6447072796214345e-06, + "loss": 0.82810307, + "num_input_tokens_seen": 38389795, + "step": 1804, + "time_per_iteration": 2.6386828422546387 + }, + { + "auxiliary_loss_clip": 0.01148593, + "auxiliary_loss_mlp": 0.01002243, + "balance_loss_clip": 1.02420855, + "balance_loss_mlp": 0.9996441, + "epoch": 0.21703841760355919, + "flos": 58760955429120.0, + "grad_norm": 0.9120387846908584, + "language_loss": 0.63127416, + "learning_rate": 3.644263938160898e-06, + "loss": 0.6527825, + "num_input_tokens_seen": 38445760, + "step": 1805, + "time_per_iteration": 3.1741764545440674 + }, + { + "auxiliary_loss_clip": 0.01238848, + "auxiliary_loss_mlp": 0.0103459, + "balance_loss_clip": 1.0566045, + "balance_loss_mlp": 1.02507746, + "epoch": 0.21715866049419827, + "flos": 22418457419520.0, + "grad_norm": 1.9111943440611494, + "language_loss": 0.71846998, + "learning_rate": 3.6438203472668293e-06, + "loss": 0.74120438, + "num_input_tokens_seen": 38465405, + "step": 1806, + "time_per_iteration": 2.6971139907836914 + }, + { + "auxiliary_loss_clip": 0.01234225, + "auxiliary_loss_mlp": 0.01035463, + "balance_loss_clip": 1.05948997, + "balance_loss_mlp": 1.02581906, + "epoch": 0.21727890338483738, + "flos": 17237732952960.0, + "grad_norm": 2.0067108382295418, + "language_loss": 0.8188709, + "learning_rate": 3.6433765070065206e-06, + "loss": 0.84156775, + "num_input_tokens_seen": 38483195, + "step": 1807, + "time_per_iteration": 2.6115517616271973 + }, + { + "auxiliary_loss_clip": 0.01213822, + "auxiliary_loss_mlp": 0.01032124, + "balance_loss_clip": 1.06048012, + "balance_loss_mlp": 1.02191401, + "epoch": 0.21739914627547646, + "flos": 13434792416640.0, + "grad_norm": 2.7123664396370972, + "language_loss": 0.87230527, + "learning_rate": 3.6429324174473025e-06, + "loss": 0.89476472, + "num_input_tokens_seen": 38496735, + "step": 1808, + "time_per_iteration": 2.502225637435913 + }, + { + "auxiliary_loss_clip": 0.0122421, + "auxiliary_loss_mlp": 0.01034076, + "balance_loss_clip": 1.05919218, + "balance_loss_mlp": 1.02417517, + "epoch": 0.21751938916611555, + "flos": 20959514709120.0, + "grad_norm": 3.4040300829923646, + "language_loss": 0.84659004, + "learning_rate": 3.6424880786565425e-06, + "loss": 0.86917287, + "num_input_tokens_seen": 38512880, + "step": 1809, + "time_per_iteration": 3.430041551589966 + }, + { + "auxiliary_loss_clip": 0.0124448, + "auxiliary_loss_mlp": 0.01030769, + "balance_loss_clip": 1.05462551, + "balance_loss_mlp": 1.02132773, + "epoch": 0.21763963205675466, + "flos": 27599936071680.0, + "grad_norm": 3.394627060540741, + "language_loss": 0.79788733, + "learning_rate": 3.6420434907016482e-06, + "loss": 0.82063985, + "num_input_tokens_seen": 38532570, + "step": 1810, + "time_per_iteration": 3.5614655017852783 + }, + { + "auxiliary_loss_clip": 0.01224233, + "auxiliary_loss_mlp": 0.01037632, + "balance_loss_clip": 1.0625422, + "balance_loss_mlp": 1.02842903, + "epoch": 0.21775987494739374, + "flos": 21430411032960.0, + "grad_norm": 1.5534643719014511, + "language_loss": 0.80831099, + "learning_rate": 3.6415986536500606e-06, + "loss": 0.83092964, + "num_input_tokens_seen": 38550900, + "step": 1811, + "time_per_iteration": 3.360713005065918 + }, + { + "auxiliary_loss_clip": 0.01241504, + "auxiliary_loss_mlp": 0.01035007, + "balance_loss_clip": 1.05879188, + "balance_loss_mlp": 1.02635252, + "epoch": 0.21788011783803282, + "flos": 18332972501760.0, + "grad_norm": 1.7735762648023883, + "language_loss": 0.80655235, + "learning_rate": 3.641153567569263e-06, + "loss": 0.82931745, + "num_input_tokens_seen": 38569215, + "step": 1812, + "time_per_iteration": 2.7568161487579346 + }, + { + "auxiliary_loss_clip": 0.01218883, + "auxiliary_loss_mlp": 0.01030498, + "balance_loss_clip": 1.05965078, + "balance_loss_mlp": 1.02052009, + "epoch": 0.2180003607286719, + "flos": 30262748037120.0, + "grad_norm": 1.9888282447155705, + "language_loss": 0.95609486, + "learning_rate": 3.640708232526774e-06, + "loss": 0.97858864, + "num_input_tokens_seen": 38587870, + "step": 1813, + "time_per_iteration": 3.6460788249969482 + }, + { + "auxiliary_loss_clip": 0.01249406, + "auxiliary_loss_mlp": 0.01032642, + "balance_loss_clip": 1.04811287, + "balance_loss_mlp": 1.02337956, + "epoch": 0.21812060361931102, + "flos": 25480272637440.0, + "grad_norm": 1.6829955532791023, + "language_loss": 0.78203952, + "learning_rate": 3.6402626485901504e-06, + "loss": 0.80486, + "num_input_tokens_seen": 38606965, + "step": 1814, + "time_per_iteration": 2.824772596359253 + }, + { + "auxiliary_loss_clip": 0.01222375, + "auxiliary_loss_mlp": 0.01029358, + "balance_loss_clip": 1.06337488, + "balance_loss_mlp": 1.02038777, + "epoch": 0.2182408465099501, + "flos": 21908166854400.0, + "grad_norm": 1.9666407615711694, + "language_loss": 0.78114522, + "learning_rate": 3.639816815826988e-06, + "loss": 0.80366254, + "num_input_tokens_seen": 38626290, + "step": 1815, + "time_per_iteration": 2.635852098464966 + }, + { + "auxiliary_loss_clip": 0.01229592, + "auxiliary_loss_mlp": 0.01039331, + "balance_loss_clip": 1.0595299, + "balance_loss_mlp": 1.02990186, + "epoch": 0.21836108940058918, + "flos": 23657339456640.0, + "grad_norm": 2.02662866014454, + "language_loss": 0.78172982, + "learning_rate": 3.6393707343049176e-06, + "loss": 0.80441904, + "num_input_tokens_seen": 38646620, + "step": 1816, + "time_per_iteration": 2.6369311809539795 + }, + { + "auxiliary_loss_clip": 0.01226272, + "auxiliary_loss_mlp": 0.01032663, + "balance_loss_clip": 1.06041098, + "balance_loss_mlp": 1.02288818, + "epoch": 0.2184813322912283, + "flos": 24681009156480.0, + "grad_norm": 2.2300580939168855, + "language_loss": 0.73430979, + "learning_rate": 3.6389244040916104e-06, + "loss": 0.75689912, + "num_input_tokens_seen": 38665695, + "step": 1817, + "time_per_iteration": 2.6371335983276367 + }, + { + "auxiliary_loss_clip": 0.01218305, + "auxiliary_loss_mlp": 0.01380323, + "balance_loss_clip": 1.05509114, + "balance_loss_mlp": 1.00009799, + "epoch": 0.21860157518186737, + "flos": 26574650259840.0, + "grad_norm": 2.0048156493737066, + "language_loss": 0.79128444, + "learning_rate": 3.6384778252547747e-06, + "loss": 0.81727076, + "num_input_tokens_seen": 38681575, + "step": 1818, + "time_per_iteration": 2.6170859336853027 + }, + { + "auxiliary_loss_clip": 0.01230528, + "auxiliary_loss_mlp": 0.01380702, + "balance_loss_clip": 1.0604583, + "balance_loss_mlp": 1.00021291, + "epoch": 0.21872181807250646, + "flos": 20886292834560.0, + "grad_norm": 2.5518566340734354, + "language_loss": 0.78475165, + "learning_rate": 3.638030997862155e-06, + "loss": 0.81086385, + "num_input_tokens_seen": 38700510, + "step": 1819, + "time_per_iteration": 2.6603944301605225 + }, + { + "auxiliary_loss_clip": 0.01133024, + "auxiliary_loss_mlp": 0.01006019, + "balance_loss_clip": 1.03175759, + "balance_loss_mlp": 1.00351608, + "epoch": 0.21884206096314554, + "flos": 61209452897280.0, + "grad_norm": 0.7653876622628085, + "language_loss": 0.59455204, + "learning_rate": 3.6375839219815356e-06, + "loss": 0.61594248, + "num_input_tokens_seen": 38758310, + "step": 1820, + "time_per_iteration": 3.16528058052063 + }, + { + "auxiliary_loss_clip": 0.01212802, + "auxiliary_loss_mlp": 0.01035194, + "balance_loss_clip": 1.06007421, + "balance_loss_mlp": 1.02527547, + "epoch": 0.21896230385378465, + "flos": 23473835850240.0, + "grad_norm": 2.144114376443238, + "language_loss": 0.82552481, + "learning_rate": 3.6371365976807375e-06, + "loss": 0.8480047, + "num_input_tokens_seen": 38778705, + "step": 1821, + "time_per_iteration": 2.634838104248047 + }, + { + "auxiliary_loss_clip": 0.01236849, + "auxiliary_loss_mlp": 0.01040357, + "balance_loss_clip": 1.05615807, + "balance_loss_mlp": 1.03052235, + "epoch": 0.21908254674442373, + "flos": 25081915829760.0, + "grad_norm": 1.8987272428076536, + "language_loss": 0.83427286, + "learning_rate": 3.6366890250276185e-06, + "loss": 0.85704494, + "num_input_tokens_seen": 38799660, + "step": 1822, + "time_per_iteration": 2.7581543922424316 + }, + { + "auxiliary_loss_clip": 0.01214788, + "auxiliary_loss_mlp": 0.01033481, + "balance_loss_clip": 1.06231046, + "balance_loss_mlp": 1.02421808, + "epoch": 0.21920278963506282, + "flos": 23513768795520.0, + "grad_norm": 3.0520760356257433, + "language_loss": 0.89902949, + "learning_rate": 3.6362412040900764e-06, + "loss": 0.92151213, + "num_input_tokens_seen": 38819450, + "step": 1823, + "time_per_iteration": 2.635899782180786 + }, + { + "auxiliary_loss_clip": 0.0122532, + "auxiliary_loss_mlp": 0.0103932, + "balance_loss_clip": 1.05989647, + "balance_loss_mlp": 1.02996194, + "epoch": 0.21932303252570193, + "flos": 29242238734080.0, + "grad_norm": 1.8846757792798514, + "language_loss": 0.80420518, + "learning_rate": 3.635793134936044e-06, + "loss": 0.82685161, + "num_input_tokens_seen": 38840460, + "step": 1824, + "time_per_iteration": 2.656358242034912 + }, + { + "auxiliary_loss_clip": 0.01223899, + "auxiliary_loss_mlp": 0.01032152, + "balance_loss_clip": 1.06201994, + "balance_loss_mlp": 1.02275848, + "epoch": 0.219443275416341, + "flos": 20806857907200.0, + "grad_norm": 1.9810333788117138, + "language_loss": 0.73312223, + "learning_rate": 3.635344817633494e-06, + "loss": 0.75568271, + "num_input_tokens_seen": 38859775, + "step": 1825, + "time_per_iteration": 2.611393690109253 + }, + { + "auxiliary_loss_clip": 0.01218646, + "auxiliary_loss_mlp": 0.01036369, + "balance_loss_clip": 1.05906022, + "balance_loss_mlp": 1.0275116, + "epoch": 0.2195635183069801, + "flos": 14501555458560.0, + "grad_norm": 3.020219479228437, + "language_loss": 0.7506054, + "learning_rate": 3.634896252250436e-06, + "loss": 0.77315557, + "num_input_tokens_seen": 38876540, + "step": 1826, + "time_per_iteration": 2.6047234535217285 + }, + { + "auxiliary_loss_clip": 0.01218579, + "auxiliary_loss_mlp": 0.01029412, + "balance_loss_clip": 1.06389093, + "balance_loss_mlp": 1.01946378, + "epoch": 0.2196837611976192, + "flos": 24243473589120.0, + "grad_norm": 1.8224827567596211, + "language_loss": 0.82308006, + "learning_rate": 3.6344474388549157e-06, + "loss": 0.84556001, + "num_input_tokens_seen": 38896195, + "step": 1827, + "time_per_iteration": 2.6136176586151123 + }, + { + "auxiliary_loss_clip": 0.01225355, + "auxiliary_loss_mlp": 0.01037052, + "balance_loss_clip": 1.06331396, + "balance_loss_mlp": 1.02721691, + "epoch": 0.2198040040882583, + "flos": 18074523168000.0, + "grad_norm": 2.2263934628190554, + "language_loss": 0.79972458, + "learning_rate": 3.6339983775150183e-06, + "loss": 0.82234871, + "num_input_tokens_seen": 38912755, + "step": 1828, + "time_per_iteration": 2.6268868446350098 + }, + { + "auxiliary_loss_clip": 0.01224464, + "auxiliary_loss_mlp": 0.01035463, + "balance_loss_clip": 1.062024, + "balance_loss_mlp": 1.02620649, + "epoch": 0.21992424697889737, + "flos": 17784185535360.0, + "grad_norm": 2.8958043962713025, + "language_loss": 0.83956087, + "learning_rate": 3.6335490682988664e-06, + "loss": 0.86216015, + "num_input_tokens_seen": 38928365, + "step": 1829, + "time_per_iteration": 2.5816988945007324 + }, + { + "auxiliary_loss_clip": 0.01244697, + "auxiliary_loss_mlp": 0.01037232, + "balance_loss_clip": 1.05134416, + "balance_loss_mlp": 1.0271349, + "epoch": 0.22004448986953645, + "flos": 17638495971840.0, + "grad_norm": 2.3882078673389313, + "language_loss": 0.82537216, + "learning_rate": 3.63309951127462e-06, + "loss": 0.8481915, + "num_input_tokens_seen": 38945275, + "step": 1830, + "time_per_iteration": 2.7889583110809326 + }, + { + "auxiliary_loss_clip": 0.01239535, + "auxiliary_loss_mlp": 0.01032646, + "balance_loss_clip": 1.0585357, + "balance_loss_mlp": 1.02235818, + "epoch": 0.22016473276017556, + "flos": 22275533203200.0, + "grad_norm": 2.18636506612596, + "language_loss": 0.75058365, + "learning_rate": 3.6326497065104757e-06, + "loss": 0.77330548, + "num_input_tokens_seen": 38965740, + "step": 1831, + "time_per_iteration": 2.734578847885132 + }, + { + "auxiliary_loss_clip": 0.01227105, + "auxiliary_loss_mlp": 0.01033277, + "balance_loss_clip": 1.06097353, + "balance_loss_mlp": 1.02396703, + "epoch": 0.22028497565081465, + "flos": 25556259859200.0, + "grad_norm": 2.050341873341893, + "language_loss": 0.77814817, + "learning_rate": 3.6321996540746697e-06, + "loss": 0.80075198, + "num_input_tokens_seen": 38984815, + "step": 1832, + "time_per_iteration": 2.719008445739746 + }, + { + "auxiliary_loss_clip": 0.0123821, + "auxiliary_loss_mlp": 0.01034488, + "balance_loss_clip": 1.05715275, + "balance_loss_mlp": 1.02421844, + "epoch": 0.22040521854145373, + "flos": 36247332925440.0, + "grad_norm": 2.092586293246555, + "language_loss": 0.8058182, + "learning_rate": 3.6317493540354733e-06, + "loss": 0.82854521, + "num_input_tokens_seen": 39008230, + "step": 1833, + "time_per_iteration": 2.912940502166748 + }, + { + "auxiliary_loss_clip": 0.01214715, + "auxiliary_loss_mlp": 0.01037496, + "balance_loss_clip": 1.05723095, + "balance_loss_mlp": 1.02760172, + "epoch": 0.22052546143209284, + "flos": 11838420270720.0, + "grad_norm": 2.6169492758498896, + "language_loss": 0.7682848, + "learning_rate": 3.6312988064611976e-06, + "loss": 0.79080689, + "num_input_tokens_seen": 39026540, + "step": 1834, + "time_per_iteration": 3.5469205379486084 + }, + { + "auxiliary_loss_clip": 0.01240553, + "auxiliary_loss_mlp": 0.01035826, + "balance_loss_clip": 1.05358839, + "balance_loss_mlp": 1.02602124, + "epoch": 0.22064570432273192, + "flos": 24209250906240.0, + "grad_norm": 1.647917634094187, + "language_loss": 0.81270993, + "learning_rate": 3.6308480114201896e-06, + "loss": 0.83547372, + "num_input_tokens_seen": 39048460, + "step": 1835, + "time_per_iteration": 2.7447636127471924 + }, + { + "auxiliary_loss_clip": 0.01218137, + "auxiliary_loss_mlp": 0.01035856, + "balance_loss_clip": 1.06505907, + "balance_loss_mlp": 1.02578235, + "epoch": 0.220765947213371, + "flos": 17931347556480.0, + "grad_norm": 1.9086838270643005, + "language_loss": 0.76802307, + "learning_rate": 3.630396968980835e-06, + "loss": 0.79056299, + "num_input_tokens_seen": 39066335, + "step": 1836, + "time_per_iteration": 4.574431657791138 + }, + { + "auxiliary_loss_clip": 0.01231823, + "auxiliary_loss_mlp": 0.01034721, + "balance_loss_clip": 1.0596962, + "balance_loss_mlp": 1.02512407, + "epoch": 0.2208861901040101, + "flos": 26757040544640.0, + "grad_norm": 2.2395008816197555, + "language_loss": 0.83705151, + "learning_rate": 3.6299456792115575e-06, + "loss": 0.85971689, + "num_input_tokens_seen": 39087590, + "step": 1837, + "time_per_iteration": 2.710528612136841 + }, + { + "auxiliary_loss_clip": 0.01240529, + "auxiliary_loss_mlp": 0.01038108, + "balance_loss_clip": 1.04752064, + "balance_loss_mlp": 1.02836299, + "epoch": 0.2210064329946492, + "flos": 17817977255040.0, + "grad_norm": 2.7105229592964415, + "language_loss": 0.81211805, + "learning_rate": 3.629494142180815e-06, + "loss": 0.83490443, + "num_input_tokens_seen": 39106335, + "step": 1838, + "time_per_iteration": 2.911562204360962 + }, + { + "auxiliary_loss_clip": 0.01215721, + "auxiliary_loss_mlp": 0.01036734, + "balance_loss_clip": 1.06278443, + "balance_loss_mlp": 1.02728677, + "epoch": 0.22112667588528828, + "flos": 17967401832960.0, + "grad_norm": 2.27685269323822, + "language_loss": 0.85201263, + "learning_rate": 3.6290423579571075e-06, + "loss": 0.87453717, + "num_input_tokens_seen": 39122875, + "step": 1839, + "time_per_iteration": 3.9680047035217285 + }, + { + "auxiliary_loss_clip": 0.01221764, + "auxiliary_loss_mlp": 0.01031153, + "balance_loss_clip": 1.06188881, + "balance_loss_mlp": 1.02149081, + "epoch": 0.22124691877592736, + "flos": 18369206346240.0, + "grad_norm": 1.7284756961399488, + "language_loss": 0.80406034, + "learning_rate": 3.6285903266089694e-06, + "loss": 0.82658947, + "num_input_tokens_seen": 39142150, + "step": 1840, + "time_per_iteration": 2.7910149097442627 + }, + { + "auxiliary_loss_clip": 0.01232582, + "auxiliary_loss_mlp": 0.01034895, + "balance_loss_clip": 1.05832481, + "balance_loss_mlp": 1.02510762, + "epoch": 0.22136716166656648, + "flos": 20813286441600.0, + "grad_norm": 2.4424882830343644, + "language_loss": 0.77151859, + "learning_rate": 3.628138048204974e-06, + "loss": 0.79419339, + "num_input_tokens_seen": 39162835, + "step": 1841, + "time_per_iteration": 2.7453904151916504 + }, + { + "auxiliary_loss_clip": 0.01242114, + "auxiliary_loss_mlp": 0.01035926, + "balance_loss_clip": 1.05691457, + "balance_loss_mlp": 1.0250479, + "epoch": 0.22148740455720556, + "flos": 17675699483520.0, + "grad_norm": 2.0286321503392917, + "language_loss": 0.75899142, + "learning_rate": 3.6276855228137304e-06, + "loss": 0.78177184, + "num_input_tokens_seen": 39181040, + "step": 1842, + "time_per_iteration": 2.6869170665740967 + }, + { + "auxiliary_loss_clip": 0.01216908, + "auxiliary_loss_mlp": 0.01380859, + "balance_loss_clip": 1.06365454, + "balance_loss_mlp": 1.00016487, + "epoch": 0.22160764744784464, + "flos": 21726710323200.0, + "grad_norm": 2.1028085869050255, + "language_loss": 0.81803972, + "learning_rate": 3.6272327505038874e-06, + "loss": 0.84401739, + "num_input_tokens_seen": 39197505, + "step": 1843, + "time_per_iteration": 2.7521629333496094 + }, + { + "auxiliary_loss_clip": 0.01253611, + "auxiliary_loss_mlp": 0.01026483, + "balance_loss_clip": 1.0572865, + "balance_loss_mlp": 1.01738739, + "epoch": 0.22172789033848372, + "flos": 23764712186880.0, + "grad_norm": 1.8767553588855141, + "language_loss": 0.7874282, + "learning_rate": 3.626779731344131e-06, + "loss": 0.81022918, + "num_input_tokens_seen": 39217295, + "step": 1844, + "time_per_iteration": 2.8430240154266357 + }, + { + "auxiliary_loss_clip": 0.01210871, + "auxiliary_loss_mlp": 0.01033266, + "balance_loss_clip": 1.06021297, + "balance_loss_mlp": 1.02394366, + "epoch": 0.22184813322912283, + "flos": 16982300361600.0, + "grad_norm": 2.243329398726319, + "language_loss": 0.85208935, + "learning_rate": 3.6263264654031814e-06, + "loss": 0.87453067, + "num_input_tokens_seen": 39234195, + "step": 1845, + "time_per_iteration": 2.6374824047088623 + }, + { + "auxiliary_loss_clip": 0.01140523, + "auxiliary_loss_mlp": 0.01003724, + "balance_loss_clip": 1.026366, + "balance_loss_mlp": 1.00141144, + "epoch": 0.22196837611976192, + "flos": 61823740314240.0, + "grad_norm": 0.7064007139871679, + "language_loss": 0.59169209, + "learning_rate": 3.6258729527498008e-06, + "loss": 0.61313456, + "num_input_tokens_seen": 39295040, + "step": 1846, + "time_per_iteration": 3.2211215496063232 + }, + { + "auxiliary_loss_clip": 0.01238581, + "auxiliary_loss_mlp": 0.01038465, + "balance_loss_clip": 1.06089115, + "balance_loss_mlp": 1.02845788, + "epoch": 0.222088619010401, + "flos": 25558019625600.0, + "grad_norm": 3.0336772351406336, + "language_loss": 0.653126, + "learning_rate": 3.6254191934527854e-06, + "loss": 0.67589641, + "num_input_tokens_seen": 39314395, + "step": 1847, + "time_per_iteration": 2.746190309524536 + }, + { + "auxiliary_loss_clip": 0.01239678, + "auxiliary_loss_mlp": 0.01035938, + "balance_loss_clip": 1.05949759, + "balance_loss_mlp": 1.02599585, + "epoch": 0.2222088619010401, + "flos": 19318612677120.0, + "grad_norm": 1.8570758910216978, + "language_loss": 0.64653575, + "learning_rate": 3.6249651875809715e-06, + "loss": 0.66929197, + "num_input_tokens_seen": 39334275, + "step": 1848, + "time_per_iteration": 2.671419620513916 + }, + { + "auxiliary_loss_clip": 0.01224169, + "auxiliary_loss_mlp": 0.01033929, + "balance_loss_clip": 1.05812633, + "balance_loss_mlp": 1.02455318, + "epoch": 0.2223291047916792, + "flos": 19099342103040.0, + "grad_norm": 2.0597657891916605, + "language_loss": 0.8917101, + "learning_rate": 3.62451093520323e-06, + "loss": 0.91429108, + "num_input_tokens_seen": 39352180, + "step": 1849, + "time_per_iteration": 2.6758062839508057 + }, + { + "auxiliary_loss_clip": 0.01237999, + "auxiliary_loss_mlp": 0.01033936, + "balance_loss_clip": 1.05111337, + "balance_loss_mlp": 1.02431571, + "epoch": 0.22244934768231828, + "flos": 20850418126080.0, + "grad_norm": 2.0891394742344414, + "language_loss": 0.90555125, + "learning_rate": 3.6240564363884714e-06, + "loss": 0.92827058, + "num_input_tokens_seen": 39372125, + "step": 1850, + "time_per_iteration": 2.7535204887390137 + }, + { + "auxiliary_loss_clip": 0.01223591, + "auxiliary_loss_mlp": 0.0102879, + "balance_loss_clip": 1.05642176, + "balance_loss_mlp": 1.01864505, + "epoch": 0.2225695905729574, + "flos": 15632921111040.0, + "grad_norm": 3.9376361451846753, + "language_loss": 0.7083019, + "learning_rate": 3.623601691205643e-06, + "loss": 0.73082572, + "num_input_tokens_seen": 39391200, + "step": 1851, + "time_per_iteration": 2.6266987323760986 + }, + { + "auxiliary_loss_clip": 0.0121833, + "auxiliary_loss_mlp": 0.01035435, + "balance_loss_clip": 1.05864525, + "balance_loss_mlp": 1.02584505, + "epoch": 0.22268983346359647, + "flos": 25373582265600.0, + "grad_norm": 3.505342155845072, + "language_loss": 0.81549215, + "learning_rate": 3.623146699723729e-06, + "loss": 0.83802974, + "num_input_tokens_seen": 39410660, + "step": 1852, + "time_per_iteration": 2.746691942214966 + }, + { + "auxiliary_loss_clip": 0.01231189, + "auxiliary_loss_mlp": 0.01037108, + "balance_loss_clip": 1.06138146, + "balance_loss_mlp": 1.02748168, + "epoch": 0.22281007635423555, + "flos": 13261452359040.0, + "grad_norm": 2.3433694011628727, + "language_loss": 0.77949202, + "learning_rate": 3.6226914620117507e-06, + "loss": 0.80217493, + "num_input_tokens_seen": 39429280, + "step": 1853, + "time_per_iteration": 2.7114717960357666 + }, + { + "auxiliary_loss_clip": 0.01234488, + "auxiliary_loss_mlp": 0.01027786, + "balance_loss_clip": 1.05208266, + "balance_loss_mlp": 1.01922071, + "epoch": 0.22293031924487464, + "flos": 15340536403200.0, + "grad_norm": 2.230525255635408, + "language_loss": 0.81008387, + "learning_rate": 3.622235978138768e-06, + "loss": 0.83270663, + "num_input_tokens_seen": 39446905, + "step": 1854, + "time_per_iteration": 2.7301244735717773 + }, + { + "auxiliary_loss_clip": 0.01220301, + "auxiliary_loss_mlp": 0.01034543, + "balance_loss_clip": 1.06119549, + "balance_loss_mlp": 1.02514923, + "epoch": 0.22305056213551375, + "flos": 22564649773440.0, + "grad_norm": 1.960822266738686, + "language_loss": 0.81302869, + "learning_rate": 3.621780248173877e-06, + "loss": 0.83557713, + "num_input_tokens_seen": 39465105, + "step": 1855, + "time_per_iteration": 2.707562208175659 + }, + { + "auxiliary_loss_clip": 0.01133335, + "auxiliary_loss_mlp": 0.01004493, + "balance_loss_clip": 1.03803039, + "balance_loss_mlp": 1.00235963, + "epoch": 0.22317080502615283, + "flos": 64880419887360.0, + "grad_norm": 0.8630985431220574, + "language_loss": 0.61042905, + "learning_rate": 3.6213242721862125e-06, + "loss": 0.63180733, + "num_input_tokens_seen": 39523560, + "step": 1856, + "time_per_iteration": 3.280025005340576 + }, + { + "auxiliary_loss_clip": 0.01217722, + "auxiliary_loss_mlp": 0.01031699, + "balance_loss_clip": 1.05831325, + "balance_loss_mlp": 1.02299631, + "epoch": 0.2232910479167919, + "flos": 25775997310080.0, + "grad_norm": 1.5575175722682097, + "language_loss": 0.75518513, + "learning_rate": 3.620868050244945e-06, + "loss": 0.77767932, + "num_input_tokens_seen": 39544040, + "step": 1857, + "time_per_iteration": 2.7667670249938965 + }, + { + "auxiliary_loss_clip": 0.01220779, + "auxiliary_loss_mlp": 0.01036113, + "balance_loss_clip": 1.05627477, + "balance_loss_mlp": 1.02642143, + "epoch": 0.22341129080743102, + "flos": 23251799928960.0, + "grad_norm": 2.053244827060857, + "language_loss": 0.7748521, + "learning_rate": 3.6204115824192817e-06, + "loss": 0.79742098, + "num_input_tokens_seen": 39561515, + "step": 1858, + "time_per_iteration": 2.8045952320098877 + }, + { + "auxiliary_loss_clip": 0.0121514, + "auxiliary_loss_mlp": 0.01032533, + "balance_loss_clip": 1.05380857, + "balance_loss_mlp": 1.02262652, + "epoch": 0.2235315336980701, + "flos": 21214552250880.0, + "grad_norm": 2.6641803231221495, + "language_loss": 0.76982838, + "learning_rate": 3.619954868778471e-06, + "loss": 0.79230511, + "num_input_tokens_seen": 39578210, + "step": 1859, + "time_per_iteration": 2.7193763256073 + }, + { + "auxiliary_loss_clip": 0.01225058, + "auxiliary_loss_mlp": 0.01033549, + "balance_loss_clip": 1.05567467, + "balance_loss_mlp": 1.02436435, + "epoch": 0.2236517765887092, + "flos": 19901945548800.0, + "grad_norm": 1.854938099222829, + "language_loss": 0.82981622, + "learning_rate": 3.6194979093917944e-06, + "loss": 0.85240221, + "num_input_tokens_seen": 39597625, + "step": 1860, + "time_per_iteration": 2.731153964996338 + }, + { + "auxiliary_loss_clip": 0.012229, + "auxiliary_loss_mlp": 0.01036055, + "balance_loss_clip": 1.05820775, + "balance_loss_mlp": 1.02712011, + "epoch": 0.22377201947934827, + "flos": 23214847812480.0, + "grad_norm": 1.9399630802408137, + "language_loss": 0.86761343, + "learning_rate": 3.6190407043285724e-06, + "loss": 0.890203, + "num_input_tokens_seen": 39615360, + "step": 1861, + "time_per_iteration": 4.608783006668091 + }, + { + "auxiliary_loss_clip": 0.01212846, + "auxiliary_loss_mlp": 0.01033164, + "balance_loss_clip": 1.06050348, + "balance_loss_mlp": 1.02364528, + "epoch": 0.22389226236998738, + "flos": 26794244056320.0, + "grad_norm": 1.8477688276324054, + "language_loss": 0.75757354, + "learning_rate": 3.618583253658163e-06, + "loss": 0.78003371, + "num_input_tokens_seen": 39635460, + "step": 1862, + "time_per_iteration": 3.5861153602600098 + }, + { + "auxiliary_loss_clip": 0.01245857, + "auxiliary_loss_mlp": 0.01380626, + "balance_loss_clip": 1.05604935, + "balance_loss_mlp": 1.00019836, + "epoch": 0.22401250526062647, + "flos": 24170359455360.0, + "grad_norm": 2.066235766599223, + "language_loss": 0.86517251, + "learning_rate": 3.618125557449961e-06, + "loss": 0.89143735, + "num_input_tokens_seen": 39653515, + "step": 1863, + "time_per_iteration": 2.8234288692474365 + }, + { + "auxiliary_loss_clip": 0.01217637, + "auxiliary_loss_mlp": 0.01034369, + "balance_loss_clip": 1.05891049, + "balance_loss_mlp": 1.0250113, + "epoch": 0.22413274815126555, + "flos": 16759761649920.0, + "grad_norm": 2.4680802825345993, + "language_loss": 0.83074301, + "learning_rate": 3.6176676157733983e-06, + "loss": 0.85326302, + "num_input_tokens_seen": 39668525, + "step": 1864, + "time_per_iteration": 2.6233479976654053 + }, + { + "auxiliary_loss_clip": 0.01226537, + "auxiliary_loss_mlp": 0.01035291, + "balance_loss_clip": 1.05357933, + "balance_loss_mlp": 1.0255816, + "epoch": 0.22425299104190466, + "flos": 21360205900800.0, + "grad_norm": 2.3130435587007185, + "language_loss": 0.75858527, + "learning_rate": 3.6172094286979443e-06, + "loss": 0.78120357, + "num_input_tokens_seen": 39685895, + "step": 1865, + "time_per_iteration": 2.7605199813842773 + }, + { + "auxiliary_loss_clip": 0.01226473, + "auxiliary_loss_mlp": 0.01038463, + "balance_loss_clip": 1.05578947, + "balance_loss_mlp": 1.02982092, + "epoch": 0.22437323393254374, + "flos": 32165547108480.0, + "grad_norm": 1.4689568599037652, + "language_loss": 0.81564951, + "learning_rate": 3.6167509962931064e-06, + "loss": 0.83829892, + "num_input_tokens_seen": 39711595, + "step": 1866, + "time_per_iteration": 3.666019916534424 + }, + { + "auxiliary_loss_clip": 0.01254432, + "auxiliary_loss_mlp": 0.01030834, + "balance_loss_clip": 1.05733359, + "balance_loss_mlp": 1.02082086, + "epoch": 0.22449347682318282, + "flos": 18002809664640.0, + "grad_norm": 2.5473975456193285, + "language_loss": 0.76524603, + "learning_rate": 3.6162923186284276e-06, + "loss": 0.78809869, + "num_input_tokens_seen": 39727555, + "step": 1867, + "time_per_iteration": 2.773867607116699 + }, + { + "auxiliary_loss_clip": 0.01225363, + "auxiliary_loss_mlp": 0.01033684, + "balance_loss_clip": 1.0561049, + "balance_loss_mlp": 1.02480936, + "epoch": 0.2246137197138219, + "flos": 18697286194560.0, + "grad_norm": 2.101939620892703, + "language_loss": 0.85991228, + "learning_rate": 3.6158333957734888e-06, + "loss": 0.88250273, + "num_input_tokens_seen": 39746145, + "step": 1868, + "time_per_iteration": 2.6392643451690674 + }, + { + "auxiliary_loss_clip": 0.01239459, + "auxiliary_loss_mlp": 0.01032352, + "balance_loss_clip": 1.05426288, + "balance_loss_mlp": 1.02327406, + "epoch": 0.22473396260446102, + "flos": 15590653781760.0, + "grad_norm": 2.264361782788723, + "language_loss": 0.82867444, + "learning_rate": 3.6153742277979088e-06, + "loss": 0.85139257, + "num_input_tokens_seen": 39763575, + "step": 1869, + "time_per_iteration": 2.755798578262329 + }, + { + "auxiliary_loss_clip": 0.01227549, + "auxiliary_loss_mlp": 0.01031831, + "balance_loss_clip": 1.05585909, + "balance_loss_mlp": 1.0227952, + "epoch": 0.2248542054951001, + "flos": 14465501182080.0, + "grad_norm": 2.18316171048681, + "language_loss": 0.78038263, + "learning_rate": 3.6149148147713434e-06, + "loss": 0.80297637, + "num_input_tokens_seen": 39781810, + "step": 1870, + "time_per_iteration": 2.7040047645568848 + }, + { + "auxiliary_loss_clip": 0.01228205, + "auxiliary_loss_mlp": 0.01043513, + "balance_loss_clip": 1.06365204, + "balance_loss_mlp": 1.03428054, + "epoch": 0.22497444838573918, + "flos": 19243882431360.0, + "grad_norm": 2.004101144156891, + "language_loss": 0.86442137, + "learning_rate": 3.614455156763484e-06, + "loss": 0.88713849, + "num_input_tokens_seen": 39800115, + "step": 1871, + "time_per_iteration": 2.7147107124328613 + }, + { + "auxiliary_loss_clip": 0.01238294, + "auxiliary_loss_mlp": 0.01041004, + "balance_loss_clip": 1.04977643, + "balance_loss_mlp": 1.03207481, + "epoch": 0.2250946912763783, + "flos": 16910299549440.0, + "grad_norm": 2.125956707851509, + "language_loss": 0.71664184, + "learning_rate": 3.613995253844061e-06, + "loss": 0.73943484, + "num_input_tokens_seen": 39817795, + "step": 1872, + "time_per_iteration": 2.759983777999878 + }, + { + "auxiliary_loss_clip": 0.01217998, + "auxiliary_loss_mlp": 0.01031517, + "balance_loss_clip": 1.05906618, + "balance_loss_mlp": 1.02196813, + "epoch": 0.22521493416701738, + "flos": 24681368292480.0, + "grad_norm": 2.409063993322615, + "language_loss": 0.80723423, + "learning_rate": 3.6135351060828414e-06, + "loss": 0.82972938, + "num_input_tokens_seen": 39838270, + "step": 1873, + "time_per_iteration": 2.6802423000335693 + }, + { + "auxiliary_loss_clip": 0.01218556, + "auxiliary_loss_mlp": 0.01037219, + "balance_loss_clip": 1.06354332, + "balance_loss_mlp": 1.02639484, + "epoch": 0.22533517705765646, + "flos": 17821963664640.0, + "grad_norm": 2.355613255333517, + "language_loss": 0.6965853, + "learning_rate": 3.6130747135496285e-06, + "loss": 0.71914303, + "num_input_tokens_seen": 39857270, + "step": 1874, + "time_per_iteration": 2.5761191844940186 + }, + { + "auxiliary_loss_clip": 0.01212631, + "auxiliary_loss_mlp": 0.01033937, + "balance_loss_clip": 1.06053853, + "balance_loss_mlp": 1.02363706, + "epoch": 0.22545541994829554, + "flos": 33691390899840.0, + "grad_norm": 1.8338551550711382, + "language_loss": 0.66452754, + "learning_rate": 3.6126140763142646e-06, + "loss": 0.6869933, + "num_input_tokens_seen": 39882300, + "step": 1875, + "time_per_iteration": 2.7722694873809814 + }, + { + "auxiliary_loss_clip": 0.01211361, + "auxiliary_loss_mlp": 0.0103041, + "balance_loss_clip": 1.05937052, + "balance_loss_mlp": 1.0202297, + "epoch": 0.22557566283893465, + "flos": 19171594310400.0, + "grad_norm": 2.7589280198123967, + "language_loss": 0.85941613, + "learning_rate": 3.6121531944466275e-06, + "loss": 0.88183379, + "num_input_tokens_seen": 39899625, + "step": 1876, + "time_per_iteration": 2.6631827354431152 + }, + { + "auxiliary_loss_clip": 0.01215115, + "auxiliary_loss_mlp": 0.01035135, + "balance_loss_clip": 1.05762732, + "balance_loss_mlp": 1.02572954, + "epoch": 0.22569590572957374, + "flos": 20773281669120.0, + "grad_norm": 2.254796564939821, + "language_loss": 0.78226662, + "learning_rate": 3.611692068016633e-06, + "loss": 0.8047691, + "num_input_tokens_seen": 39915955, + "step": 1877, + "time_per_iteration": 2.6709275245666504 + }, + { + "auxiliary_loss_clip": 0.0123181, + "auxiliary_loss_mlp": 0.01046021, + "balance_loss_clip": 1.05256617, + "balance_loss_mlp": 1.03524446, + "epoch": 0.22581614862021282, + "flos": 18442715529600.0, + "grad_norm": 2.3013540850922434, + "language_loss": 0.75038934, + "learning_rate": 3.611230697094233e-06, + "loss": 0.77316767, + "num_input_tokens_seen": 39932655, + "step": 1878, + "time_per_iteration": 2.7321250438690186 + }, + { + "auxiliary_loss_clip": 0.01232559, + "auxiliary_loss_mlp": 0.01029695, + "balance_loss_clip": 1.05910659, + "balance_loss_mlp": 1.02026582, + "epoch": 0.22593639151085193, + "flos": 20048389297920.0, + "grad_norm": 2.022658403517566, + "language_loss": 0.87013847, + "learning_rate": 3.6107690817494173e-06, + "loss": 0.89276105, + "num_input_tokens_seen": 39952875, + "step": 1879, + "time_per_iteration": 2.645967960357666 + }, + { + "auxiliary_loss_clip": 0.01235403, + "auxiliary_loss_mlp": 0.01035139, + "balance_loss_clip": 1.04972124, + "balance_loss_mlp": 1.02536404, + "epoch": 0.226056634401491, + "flos": 13115116350720.0, + "grad_norm": 2.393392483327454, + "language_loss": 0.7054528, + "learning_rate": 3.6103072220522117e-06, + "loss": 0.72815818, + "num_input_tokens_seen": 39968405, + "step": 1880, + "time_per_iteration": 2.6967451572418213 + }, + { + "auxiliary_loss_clip": 0.01238324, + "auxiliary_loss_mlp": 0.01030044, + "balance_loss_clip": 1.05482578, + "balance_loss_mlp": 1.02067435, + "epoch": 0.2261768772921301, + "flos": 18988378012800.0, + "grad_norm": 1.9708910870005676, + "language_loss": 0.91842604, + "learning_rate": 3.609845118072682e-06, + "loss": 0.94110972, + "num_input_tokens_seen": 39987075, + "step": 1881, + "time_per_iteration": 2.6447103023529053 + }, + { + "auxiliary_loss_clip": 0.01225154, + "auxiliary_loss_mlp": 0.01380626, + "balance_loss_clip": 1.06073427, + "balance_loss_mlp": 1.00026834, + "epoch": 0.2262971201827692, + "flos": 19974054101760.0, + "grad_norm": 1.9943735665018703, + "language_loss": 0.80094165, + "learning_rate": 3.6093827698809276e-06, + "loss": 0.82699943, + "num_input_tokens_seen": 40006175, + "step": 1882, + "time_per_iteration": 2.6827969551086426 + }, + { + "auxiliary_loss_clip": 0.01216624, + "auxiliary_loss_mlp": 0.01033879, + "balance_loss_clip": 1.05650604, + "balance_loss_mlp": 1.02422285, + "epoch": 0.2264173630734083, + "flos": 16654543735680.0, + "grad_norm": 2.2138993093262487, + "language_loss": 0.84745759, + "learning_rate": 3.6089201775470864e-06, + "loss": 0.86996257, + "num_input_tokens_seen": 40021630, + "step": 1883, + "time_per_iteration": 2.5881152153015137 + }, + { + "auxiliary_loss_clip": 0.01225127, + "auxiliary_loss_mlp": 0.01031525, + "balance_loss_clip": 1.05345345, + "balance_loss_mlp": 1.02258384, + "epoch": 0.22653760596404737, + "flos": 24389809597440.0, + "grad_norm": 1.3314465392694725, + "language_loss": 0.77698559, + "learning_rate": 3.6084573411413334e-06, + "loss": 0.79955208, + "num_input_tokens_seen": 40041025, + "step": 1884, + "time_per_iteration": 2.761610269546509 + }, + { + "auxiliary_loss_clip": 0.01233523, + "auxiliary_loss_mlp": 0.01041769, + "balance_loss_clip": 1.05646229, + "balance_loss_mlp": 1.0326612, + "epoch": 0.22665784885468646, + "flos": 18332541538560.0, + "grad_norm": 1.9145909200420317, + "language_loss": 0.81397146, + "learning_rate": 3.607994260733881e-06, + "loss": 0.83672434, + "num_input_tokens_seen": 40060265, + "step": 1885, + "time_per_iteration": 2.67734694480896 + }, + { + "auxiliary_loss_clip": 0.01207559, + "auxiliary_loss_mlp": 0.0103646, + "balance_loss_clip": 1.05469608, + "balance_loss_mlp": 1.0274297, + "epoch": 0.22677809174532557, + "flos": 24058102475520.0, + "grad_norm": 1.6888856664685912, + "language_loss": 0.74448216, + "learning_rate": 3.6075309363949776e-06, + "loss": 0.7669223, + "num_input_tokens_seen": 40079435, + "step": 1886, + "time_per_iteration": 2.672311782836914 + }, + { + "auxiliary_loss_clip": 0.01215145, + "auxiliary_loss_mlp": 0.01031241, + "balance_loss_clip": 1.062024, + "balance_loss_mlp": 1.02145982, + "epoch": 0.22689833463596465, + "flos": 20374242503040.0, + "grad_norm": 1.8715278031810956, + "language_loss": 0.81284887, + "learning_rate": 3.6070673681949094e-06, + "loss": 0.83531272, + "num_input_tokens_seen": 40097800, + "step": 1887, + "time_per_iteration": 4.335385322570801 + }, + { + "auxiliary_loss_clip": 0.01230869, + "auxiliary_loss_mlp": 0.01380382, + "balance_loss_clip": 1.05836773, + "balance_loss_mlp": 1.00019789, + "epoch": 0.22701857752660373, + "flos": 30120398438400.0, + "grad_norm": 4.190696261203444, + "language_loss": 0.81512833, + "learning_rate": 3.606603556203999e-06, + "loss": 0.84124082, + "num_input_tokens_seen": 40122745, + "step": 1888, + "time_per_iteration": 2.776937484741211 + }, + { + "auxiliary_loss_clip": 0.01218876, + "auxiliary_loss_mlp": 0.01031475, + "balance_loss_clip": 1.05670989, + "balance_loss_mlp": 1.02221227, + "epoch": 0.22713882041724284, + "flos": 22492182084480.0, + "grad_norm": 2.1281795063819104, + "language_loss": 0.8356123, + "learning_rate": 3.6061395004926066e-06, + "loss": 0.85811579, + "num_input_tokens_seen": 40141680, + "step": 1889, + "time_per_iteration": 3.574983596801758 + }, + { + "auxiliary_loss_clip": 0.01212453, + "auxiliary_loss_mlp": 0.01029924, + "balance_loss_clip": 1.06001759, + "balance_loss_mlp": 1.02010655, + "epoch": 0.22725906330788193, + "flos": 20521548178560.0, + "grad_norm": 7.613222605013229, + "language_loss": 0.85093415, + "learning_rate": 3.605675201131129e-06, + "loss": 0.87335789, + "num_input_tokens_seen": 40160140, + "step": 1890, + "time_per_iteration": 2.6665210723876953 + }, + { + "auxiliary_loss_clip": 0.01227825, + "auxiliary_loss_mlp": 0.01032177, + "balance_loss_clip": 1.0633657, + "balance_loss_mlp": 1.02271795, + "epoch": 0.227379306198521, + "flos": 18989922297600.0, + "grad_norm": 2.165221632632187, + "language_loss": 0.79624289, + "learning_rate": 3.60521065819e-06, + "loss": 0.81884289, + "num_input_tokens_seen": 40177450, + "step": 1891, + "time_per_iteration": 2.6946353912353516 + }, + { + "auxiliary_loss_clip": 0.01229233, + "auxiliary_loss_mlp": 0.01031927, + "balance_loss_clip": 1.05560672, + "balance_loss_mlp": 1.02271819, + "epoch": 0.2274995490891601, + "flos": 21798351999360.0, + "grad_norm": 1.9950910225936698, + "language_loss": 0.87445563, + "learning_rate": 3.60474587173969e-06, + "loss": 0.89706731, + "num_input_tokens_seen": 40195935, + "step": 1892, + "time_per_iteration": 3.5892302989959717 + }, + { + "auxiliary_loss_clip": 0.01222296, + "auxiliary_loss_mlp": 0.01031924, + "balance_loss_clip": 1.06281042, + "balance_loss_mlp": 1.02238095, + "epoch": 0.2276197919797992, + "flos": 19058654972160.0, + "grad_norm": 2.030330326167801, + "language_loss": 0.84293061, + "learning_rate": 3.6042808418507084e-06, + "loss": 0.86547279, + "num_input_tokens_seen": 40213620, + "step": 1893, + "time_per_iteration": 2.6019959449768066 + }, + { + "auxiliary_loss_clip": 0.01224306, + "auxiliary_loss_mlp": 0.01037846, + "balance_loss_clip": 1.064044, + "balance_loss_mlp": 1.02751088, + "epoch": 0.22774003487043828, + "flos": 18806777827200.0, + "grad_norm": 2.0957571764037772, + "language_loss": 0.77042401, + "learning_rate": 3.6038155685935976e-06, + "loss": 0.79304552, + "num_input_tokens_seen": 40230190, + "step": 1894, + "time_per_iteration": 2.6008734703063965 + }, + { + "auxiliary_loss_clip": 0.01218591, + "auxiliary_loss_mlp": 0.01033946, + "balance_loss_clip": 1.05936849, + "balance_loss_mlp": 1.02462435, + "epoch": 0.22786027776107737, + "flos": 23002544476800.0, + "grad_norm": 1.893466610208893, + "language_loss": 0.70741171, + "learning_rate": 3.6033500520389404e-06, + "loss": 0.72993708, + "num_input_tokens_seen": 40246860, + "step": 1895, + "time_per_iteration": 2.636052131652832 + }, + { + "auxiliary_loss_clip": 0.01160599, + "auxiliary_loss_mlp": 0.01004735, + "balance_loss_clip": 1.02977347, + "balance_loss_mlp": 1.00242233, + "epoch": 0.22798052065171648, + "flos": 66706872600960.0, + "grad_norm": 0.9487111266503282, + "language_loss": 0.64748192, + "learning_rate": 3.6028842922573553e-06, + "loss": 0.66913527, + "num_input_tokens_seen": 40311005, + "step": 1896, + "time_per_iteration": 3.3623104095458984 + }, + { + "auxiliary_loss_clip": 0.01147001, + "auxiliary_loss_mlp": 0.01376189, + "balance_loss_clip": 1.02790952, + "balance_loss_mlp": 0.99981469, + "epoch": 0.22810076354235556, + "flos": 62080896758400.0, + "grad_norm": 0.8611714272186012, + "language_loss": 0.62897241, + "learning_rate": 3.602418289319497e-06, + "loss": 0.65420431, + "num_input_tokens_seen": 40369560, + "step": 1897, + "time_per_iteration": 3.230609893798828 + }, + { + "auxiliary_loss_clip": 0.01239194, + "auxiliary_loss_mlp": 0.01034826, + "balance_loss_clip": 1.05297673, + "balance_loss_mlp": 1.02501547, + "epoch": 0.22822100643299464, + "flos": 23876358635520.0, + "grad_norm": 1.741781793431203, + "language_loss": 0.73345208, + "learning_rate": 3.601952043296059e-06, + "loss": 0.75619233, + "num_input_tokens_seen": 40389555, + "step": 1898, + "time_per_iteration": 2.759788751602173 + }, + { + "auxiliary_loss_clip": 0.01233564, + "auxiliary_loss_mlp": 0.01037291, + "balance_loss_clip": 1.05736184, + "balance_loss_mlp": 1.02823734, + "epoch": 0.22834124932363373, + "flos": 20991331180800.0, + "grad_norm": 2.382377656161725, + "language_loss": 0.80642128, + "learning_rate": 3.6014855542577696e-06, + "loss": 0.82912982, + "num_input_tokens_seen": 40406765, + "step": 1899, + "time_per_iteration": 2.737020254135132 + }, + { + "auxiliary_loss_clip": 0.01230246, + "auxiliary_loss_mlp": 0.0103866, + "balance_loss_clip": 1.06002748, + "balance_loss_mlp": 1.02936125, + "epoch": 0.22846149221427284, + "flos": 24901572620160.0, + "grad_norm": 2.251097359349189, + "language_loss": 0.84399104, + "learning_rate": 3.6010188222753943e-06, + "loss": 0.86668003, + "num_input_tokens_seen": 40427535, + "step": 1900, + "time_per_iteration": 2.690023899078369 + }, + { + "auxiliary_loss_clip": 0.01128622, + "auxiliary_loss_mlp": 0.01000544, + "balance_loss_clip": 1.02568257, + "balance_loss_mlp": 0.99842197, + "epoch": 0.22858173510491192, + "flos": 56132294319360.0, + "grad_norm": 0.9017420691444975, + "language_loss": 0.64131641, + "learning_rate": 3.6005518474197372e-06, + "loss": 0.66260803, + "num_input_tokens_seen": 40479580, + "step": 1901, + "time_per_iteration": 3.145679235458374 + }, + { + "auxiliary_loss_clip": 0.0122061, + "auxiliary_loss_mlp": 0.01031516, + "balance_loss_clip": 1.06054807, + "balance_loss_mlp": 1.0225153, + "epoch": 0.228701977995551, + "flos": 24170826332160.0, + "grad_norm": 4.121986572115232, + "language_loss": 0.7847054, + "learning_rate": 3.6000846297616373e-06, + "loss": 0.80722672, + "num_input_tokens_seen": 40497880, + "step": 1902, + "time_per_iteration": 2.8262276649475098 + }, + { + "auxiliary_loss_clip": 0.01220092, + "auxiliary_loss_mlp": 0.01035954, + "balance_loss_clip": 1.06664348, + "balance_loss_mlp": 1.02573788, + "epoch": 0.22882222088619011, + "flos": 21387892308480.0, + "grad_norm": 2.3463590726470396, + "language_loss": 0.72500658, + "learning_rate": 3.5996171693719717e-06, + "loss": 0.74756706, + "num_input_tokens_seen": 40513975, + "step": 1903, + "time_per_iteration": 2.598330020904541 + }, + { + "auxiliary_loss_clip": 0.01122838, + "auxiliary_loss_mlp": 0.01000855, + "balance_loss_clip": 1.02776766, + "balance_loss_mlp": 0.99878043, + "epoch": 0.2289424637768292, + "flos": 64589615377920.0, + "grad_norm": 0.8420207645036583, + "language_loss": 0.64866376, + "learning_rate": 3.5991494663216528e-06, + "loss": 0.66990077, + "num_input_tokens_seen": 40576960, + "step": 1904, + "time_per_iteration": 3.2514255046844482 + }, + { + "auxiliary_loss_clip": 0.01211118, + "auxiliary_loss_mlp": 0.01031108, + "balance_loss_clip": 1.06077409, + "balance_loss_mlp": 1.02174973, + "epoch": 0.22906270666746828, + "flos": 22163419877760.0, + "grad_norm": 1.8859907907379228, + "language_loss": 0.87786669, + "learning_rate": 3.5986815206816314e-06, + "loss": 0.900289, + "num_input_tokens_seen": 40595780, + "step": 1905, + "time_per_iteration": 2.6002414226531982 + }, + { + "auxiliary_loss_clip": 0.01209029, + "auxiliary_loss_mlp": 0.0103058, + "balance_loss_clip": 1.05850697, + "balance_loss_mlp": 1.02138925, + "epoch": 0.2291829495581074, + "flos": 25772334122880.0, + "grad_norm": 2.871210260649478, + "language_loss": 0.74631625, + "learning_rate": 3.598213332522895e-06, + "loss": 0.7687124, + "num_input_tokens_seen": 40615810, + "step": 1906, + "time_per_iteration": 2.6786789894104004 + }, + { + "auxiliary_loss_clip": 0.01217305, + "auxiliary_loss_mlp": 0.01035901, + "balance_loss_clip": 1.0579102, + "balance_loss_mlp": 1.02712762, + "epoch": 0.22930319244874647, + "flos": 31172760126720.0, + "grad_norm": 1.8906274224195658, + "language_loss": 0.77288008, + "learning_rate": 3.597744901916466e-06, + "loss": 0.79541212, + "num_input_tokens_seen": 40637095, + "step": 1907, + "time_per_iteration": 2.646880626678467 + }, + { + "auxiliary_loss_clip": 0.01215013, + "auxiliary_loss_mlp": 0.01030048, + "balance_loss_clip": 1.05958962, + "balance_loss_mlp": 1.02017164, + "epoch": 0.22942343533938556, + "flos": 23254098399360.0, + "grad_norm": 2.036387428776994, + "language_loss": 0.76766622, + "learning_rate": 3.5972762289334058e-06, + "loss": 0.79011679, + "num_input_tokens_seen": 40656725, + "step": 1908, + "time_per_iteration": 2.6155190467834473 + }, + { + "auxiliary_loss_clip": 0.01247356, + "auxiliary_loss_mlp": 0.01034912, + "balance_loss_clip": 1.05222511, + "balance_loss_mlp": 1.02566683, + "epoch": 0.22954367823002464, + "flos": 14610903436800.0, + "grad_norm": 2.4601622677287067, + "language_loss": 0.85482717, + "learning_rate": 3.5968073136448116e-06, + "loss": 0.87764984, + "num_input_tokens_seen": 40674745, + "step": 1909, + "time_per_iteration": 2.732517957687378 + }, + { + "auxiliary_loss_clip": 0.01223112, + "auxiliary_loss_mlp": 0.0102998, + "balance_loss_clip": 1.0580765, + "balance_loss_mlp": 1.02046072, + "epoch": 0.22966392112066375, + "flos": 16763604405120.0, + "grad_norm": 1.7015350678480303, + "language_loss": 0.9144237, + "learning_rate": 3.596338156121818e-06, + "loss": 0.93695462, + "num_input_tokens_seen": 40693630, + "step": 1910, + "time_per_iteration": 2.6711559295654297 + }, + { + "auxiliary_loss_clip": 0.01127937, + "auxiliary_loss_mlp": 0.01007408, + "balance_loss_clip": 1.02293205, + "balance_loss_mlp": 1.00541675, + "epoch": 0.22978416401130283, + "flos": 67474247783040.0, + "grad_norm": 0.7433014969078577, + "language_loss": 0.59314191, + "learning_rate": 3.595868756435595e-06, + "loss": 0.61449534, + "num_input_tokens_seen": 40761310, + "step": 1911, + "time_per_iteration": 3.344005584716797 + }, + { + "auxiliary_loss_clip": 0.01243684, + "auxiliary_loss_mlp": 0.01034002, + "balance_loss_clip": 1.060498, + "balance_loss_mlp": 1.02497792, + "epoch": 0.22990440690194192, + "flos": 19865137086720.0, + "grad_norm": 2.636931491315663, + "language_loss": 0.80234963, + "learning_rate": 3.5953991146573504e-06, + "loss": 0.82512653, + "num_input_tokens_seen": 40779955, + "step": 1912, + "time_per_iteration": 3.6460657119750977 + }, + { + "auxiliary_loss_clip": 0.01223671, + "auxiliary_loss_mlp": 0.01032781, + "balance_loss_clip": 1.05824101, + "balance_loss_mlp": 1.0232029, + "epoch": 0.23002464979258103, + "flos": 13289246507520.0, + "grad_norm": 9.871668307567216, + "language_loss": 0.83156466, + "learning_rate": 3.5949292308583294e-06, + "loss": 0.8541292, + "num_input_tokens_seen": 40793200, + "step": 1913, + "time_per_iteration": 3.512970447540283 + }, + { + "auxiliary_loss_clip": 0.01212094, + "auxiliary_loss_mlp": 0.01037278, + "balance_loss_clip": 1.06144381, + "balance_loss_mlp": 1.0276165, + "epoch": 0.2301448926832201, + "flos": 22163779013760.0, + "grad_norm": 2.4682230785232195, + "language_loss": 0.80642331, + "learning_rate": 3.594459105109811e-06, + "loss": 0.82891703, + "num_input_tokens_seen": 40812380, + "step": 1914, + "time_per_iteration": 3.5719804763793945 + }, + { + "auxiliary_loss_clip": 0.01225075, + "auxiliary_loss_mlp": 0.01031251, + "balance_loss_clip": 1.06203842, + "balance_loss_mlp": 1.02154756, + "epoch": 0.2302651355738592, + "flos": 20704477167360.0, + "grad_norm": 2.248075492869372, + "language_loss": 0.81565666, + "learning_rate": 3.593988737483115e-06, + "loss": 0.83821994, + "num_input_tokens_seen": 40832320, + "step": 1915, + "time_per_iteration": 2.695765256881714 + }, + { + "auxiliary_loss_clip": 0.01227593, + "auxiliary_loss_mlp": 0.01044826, + "balance_loss_clip": 1.05653095, + "balance_loss_mlp": 1.03512824, + "epoch": 0.23038537846449827, + "flos": 18588943797120.0, + "grad_norm": 1.8421738838466362, + "language_loss": 0.78425467, + "learning_rate": 3.5935181280495947e-06, + "loss": 0.80697882, + "num_input_tokens_seen": 40850900, + "step": 1916, + "time_per_iteration": 2.6958515644073486 + }, + { + "auxiliary_loss_clip": 0.01117242, + "auxiliary_loss_mlp": 0.01001198, + "balance_loss_clip": 1.01762414, + "balance_loss_mlp": 0.99911231, + "epoch": 0.23050562135513739, + "flos": 64224260190720.0, + "grad_norm": 0.8606421924743785, + "language_loss": 0.54254544, + "learning_rate": 3.5930472768806412e-06, + "loss": 0.56372982, + "num_input_tokens_seen": 40909570, + "step": 1917, + "time_per_iteration": 3.1916491985321045 + }, + { + "auxiliary_loss_clip": 0.01211941, + "auxiliary_loss_mlp": 0.01033377, + "balance_loss_clip": 1.06329131, + "balance_loss_mlp": 1.02419186, + "epoch": 0.23062586424577647, + "flos": 17313396952320.0, + "grad_norm": 2.0902108159851087, + "language_loss": 0.77166045, + "learning_rate": 3.5925761840476826e-06, + "loss": 0.79411358, + "num_input_tokens_seen": 40928180, + "step": 1918, + "time_per_iteration": 3.6347286701202393 + }, + { + "auxiliary_loss_clip": 0.01225965, + "auxiliary_loss_mlp": 0.01024759, + "balance_loss_clip": 1.05958533, + "balance_loss_mlp": 1.01605034, + "epoch": 0.23074610713641555, + "flos": 27855979194240.0, + "grad_norm": 2.0408631432291906, + "language_loss": 0.81577718, + "learning_rate": 3.592104849622183e-06, + "loss": 0.83828443, + "num_input_tokens_seen": 40950435, + "step": 1919, + "time_per_iteration": 2.7597618103027344 + }, + { + "auxiliary_loss_clip": 0.0123513, + "auxiliary_loss_mlp": 0.01033277, + "balance_loss_clip": 1.05289698, + "balance_loss_mlp": 1.02395475, + "epoch": 0.23086635002705466, + "flos": 28841798937600.0, + "grad_norm": 1.4867150658447075, + "language_loss": 0.73046559, + "learning_rate": 3.591633273675644e-06, + "loss": 0.75314963, + "num_input_tokens_seen": 40972670, + "step": 1920, + "time_per_iteration": 2.8238697052001953 + }, + { + "auxiliary_loss_clip": 0.01126007, + "auxiliary_loss_mlp": 0.01000917, + "balance_loss_clip": 1.02246308, + "balance_loss_mlp": 0.99884301, + "epoch": 0.23098659291769374, + "flos": 62923681566720.0, + "grad_norm": 0.9051527389991283, + "language_loss": 0.58203518, + "learning_rate": 3.591161456279602e-06, + "loss": 0.60330445, + "num_input_tokens_seen": 41018215, + "step": 1921, + "time_per_iteration": 3.0383753776550293 + }, + { + "auxiliary_loss_clip": 0.01233544, + "auxiliary_loss_mlp": 0.01029829, + "balance_loss_clip": 1.05757475, + "balance_loss_mlp": 1.0205605, + "epoch": 0.23110683580833283, + "flos": 23476816679040.0, + "grad_norm": 1.5087178922356483, + "language_loss": 0.80248463, + "learning_rate": 3.590689397505633e-06, + "loss": 0.82511842, + "num_input_tokens_seen": 41039125, + "step": 1922, + "time_per_iteration": 2.6617865562438965 + }, + { + "auxiliary_loss_clip": 0.0120878, + "auxiliary_loss_mlp": 0.01036568, + "balance_loss_clip": 1.06032109, + "balance_loss_mlp": 1.02820563, + "epoch": 0.2312270786989719, + "flos": 27271066124160.0, + "grad_norm": 2.2532003941360514, + "language_loss": 0.8663581, + "learning_rate": 3.590217097425347e-06, + "loss": 0.88881159, + "num_input_tokens_seen": 41059025, + "step": 1923, + "time_per_iteration": 2.784956932067871 + }, + { + "auxiliary_loss_clip": 0.0121128, + "auxiliary_loss_mlp": 0.01034591, + "balance_loss_clip": 1.06131494, + "balance_loss_mlp": 1.02460718, + "epoch": 0.23134732158961102, + "flos": 13261344618240.0, + "grad_norm": 2.2235104936208567, + "language_loss": 0.71230197, + "learning_rate": 3.589744556110391e-06, + "loss": 0.73476064, + "num_input_tokens_seen": 41077015, + "step": 1924, + "time_per_iteration": 2.6436374187469482 + }, + { + "auxiliary_loss_clip": 0.01221174, + "auxiliary_loss_mlp": 0.01035742, + "balance_loss_clip": 1.05418873, + "balance_loss_mlp": 1.02549565, + "epoch": 0.2314675644802501, + "flos": 36977648250240.0, + "grad_norm": 1.6063983860490432, + "language_loss": 0.84265542, + "learning_rate": 3.58927177363245e-06, + "loss": 0.8652246, + "num_input_tokens_seen": 41099840, + "step": 1925, + "time_per_iteration": 2.7937350273132324 + }, + { + "auxiliary_loss_clip": 0.0122793, + "auxiliary_loss_mlp": 0.01034398, + "balance_loss_clip": 1.0538125, + "balance_loss_mlp": 1.02481318, + "epoch": 0.2315878073708892, + "flos": 23842207779840.0, + "grad_norm": 2.1848583732210507, + "language_loss": 0.72701466, + "learning_rate": 3.5887987500632447e-06, + "loss": 0.74963796, + "num_input_tokens_seen": 41117845, + "step": 1926, + "time_per_iteration": 2.8266348838806152 + }, + { + "auxiliary_loss_clip": 0.01241679, + "auxiliary_loss_mlp": 0.0103029, + "balance_loss_clip": 1.05710387, + "balance_loss_mlp": 1.02133203, + "epoch": 0.2317080502615283, + "flos": 23039424766080.0, + "grad_norm": 1.8589514642941534, + "language_loss": 0.84021652, + "learning_rate": 3.5883254854745325e-06, + "loss": 0.86293614, + "num_input_tokens_seen": 41136235, + "step": 1927, + "time_per_iteration": 2.7085180282592773 + }, + { + "auxiliary_loss_clip": 0.01224571, + "auxiliary_loss_mlp": 0.0103628, + "balance_loss_clip": 1.05927038, + "balance_loss_mlp": 1.02643347, + "epoch": 0.23182829315216738, + "flos": 11254656435840.0, + "grad_norm": 5.418578343393888, + "language_loss": 0.75164348, + "learning_rate": 3.587851979938107e-06, + "loss": 0.774252, + "num_input_tokens_seen": 41153125, + "step": 1928, + "time_per_iteration": 2.7246270179748535 + }, + { + "auxiliary_loss_clip": 0.0121647, + "auxiliary_loss_mlp": 0.01029702, + "balance_loss_clip": 1.05801642, + "balance_loss_mlp": 1.02054048, + "epoch": 0.23194853604280646, + "flos": 19828939155840.0, + "grad_norm": 2.0072921564387833, + "language_loss": 0.77879345, + "learning_rate": 3.5873782335257985e-06, + "loss": 0.80125523, + "num_input_tokens_seen": 41171290, + "step": 1929, + "time_per_iteration": 2.62139630317688 + }, + { + "auxiliary_loss_clip": 0.01234125, + "auxiliary_loss_mlp": 0.0103149, + "balance_loss_clip": 1.05641735, + "balance_loss_mlp": 1.02206683, + "epoch": 0.23206877893344555, + "flos": 15305020830720.0, + "grad_norm": 2.1400079529176774, + "language_loss": 0.78307146, + "learning_rate": 3.5869042463094744e-06, + "loss": 0.8057276, + "num_input_tokens_seen": 41189005, + "step": 1930, + "time_per_iteration": 2.720277786254883 + }, + { + "auxiliary_loss_clip": 0.01221749, + "auxiliary_loss_mlp": 0.01033203, + "balance_loss_clip": 1.04945779, + "balance_loss_mlp": 1.02361906, + "epoch": 0.23218902182408466, + "flos": 22711488572160.0, + "grad_norm": 1.6835056505804298, + "language_loss": 0.76804256, + "learning_rate": 3.586430018361038e-06, + "loss": 0.79059207, + "num_input_tokens_seen": 41208775, + "step": 1931, + "time_per_iteration": 2.800182819366455 + }, + { + "auxiliary_loss_clip": 0.01208407, + "auxiliary_loss_mlp": 0.01033865, + "balance_loss_clip": 1.05279875, + "balance_loss_mlp": 1.02429295, + "epoch": 0.23230926471472374, + "flos": 22710734386560.0, + "grad_norm": 2.086126473552163, + "language_loss": 0.76317799, + "learning_rate": 3.5859555497524283e-06, + "loss": 0.78560072, + "num_input_tokens_seen": 41226010, + "step": 1932, + "time_per_iteration": 2.687718391418457 + }, + { + "auxiliary_loss_clip": 0.01218894, + "auxiliary_loss_mlp": 0.01028198, + "balance_loss_clip": 1.05942273, + "balance_loss_mlp": 1.01811886, + "epoch": 0.23242950760536282, + "flos": 20375499479040.0, + "grad_norm": 2.1285749636350917, + "language_loss": 0.92153287, + "learning_rate": 3.5854808405556237e-06, + "loss": 0.9440037, + "num_input_tokens_seen": 41245245, + "step": 1933, + "time_per_iteration": 2.6144299507141113 + }, + { + "auxiliary_loss_clip": 0.01238517, + "auxiliary_loss_mlp": 0.01032154, + "balance_loss_clip": 1.05688047, + "balance_loss_mlp": 1.02338004, + "epoch": 0.23254975049600193, + "flos": 16908324301440.0, + "grad_norm": 9.37978518661215, + "language_loss": 0.75285721, + "learning_rate": 3.5850058908426355e-06, + "loss": 0.77556396, + "num_input_tokens_seen": 41263795, + "step": 1934, + "time_per_iteration": 2.82173228263855 + }, + { + "auxiliary_loss_clip": 0.01227911, + "auxiliary_loss_mlp": 0.01035991, + "balance_loss_clip": 1.05514324, + "balance_loss_mlp": 1.02645981, + "epoch": 0.23266999338664102, + "flos": 23294821443840.0, + "grad_norm": 2.0090016641861905, + "language_loss": 0.85310245, + "learning_rate": 3.584530700685514e-06, + "loss": 0.87574148, + "num_input_tokens_seen": 41284055, + "step": 1935, + "time_per_iteration": 2.8882012367248535 + }, + { + "auxiliary_loss_clip": 0.01223889, + "auxiliary_loss_mlp": 0.0103638, + "balance_loss_clip": 1.05839908, + "balance_loss_mlp": 1.02765393, + "epoch": 0.2327902362772801, + "flos": 19569987031680.0, + "grad_norm": 2.0208207978664348, + "language_loss": 0.88385373, + "learning_rate": 3.5840552701563448e-06, + "loss": 0.90645641, + "num_input_tokens_seen": 41300255, + "step": 1936, + "time_per_iteration": 2.630730152130127 + }, + { + "auxiliary_loss_clip": 0.01209903, + "auxiliary_loss_mlp": 0.01030588, + "balance_loss_clip": 1.06083, + "balance_loss_mlp": 1.02161729, + "epoch": 0.2329104791679192, + "flos": 16727514215040.0, + "grad_norm": 2.0683482493320673, + "language_loss": 0.81886458, + "learning_rate": 3.5835795993272513e-06, + "loss": 0.84126955, + "num_input_tokens_seen": 41318540, + "step": 1937, + "time_per_iteration": 2.6837525367736816 + }, + { + "auxiliary_loss_clip": 0.01260588, + "auxiliary_loss_mlp": 0.01033765, + "balance_loss_clip": 1.0463717, + "balance_loss_mlp": 1.02452064, + "epoch": 0.2330307220585583, + "flos": 22163743100160.0, + "grad_norm": 2.248330473511724, + "language_loss": 0.71048361, + "learning_rate": 3.583103688270391e-06, + "loss": 0.73342711, + "num_input_tokens_seen": 41338320, + "step": 1938, + "time_per_iteration": 3.8402984142303467 + }, + { + "auxiliary_loss_clip": 0.01207754, + "auxiliary_loss_mlp": 0.01040041, + "balance_loss_clip": 1.05211401, + "balance_loss_mlp": 1.03030181, + "epoch": 0.23315096494919738, + "flos": 19317319787520.0, + "grad_norm": 2.3409370736836963, + "language_loss": 0.89183885, + "learning_rate": 3.58262753705796e-06, + "loss": 0.91431677, + "num_input_tokens_seen": 41353210, + "step": 1939, + "time_per_iteration": 3.753824472427368 + }, + { + "auxiliary_loss_clip": 0.01121567, + "auxiliary_loss_mlp": 0.01018232, + "balance_loss_clip": 1.02284634, + "balance_loss_mlp": 1.01616943, + "epoch": 0.23327120783983646, + "flos": 53031048946560.0, + "grad_norm": 0.762764076692688, + "language_loss": 0.55461085, + "learning_rate": 3.5821511457621902e-06, + "loss": 0.57600886, + "num_input_tokens_seen": 41410510, + "step": 1940, + "time_per_iteration": 3.9876582622528076 + }, + { + "auxiliary_loss_clip": 0.01219375, + "auxiliary_loss_mlp": 0.01039088, + "balance_loss_clip": 1.05507088, + "balance_loss_mlp": 1.02880037, + "epoch": 0.23339145073047557, + "flos": 17126984344320.0, + "grad_norm": 2.618426422751523, + "language_loss": 0.81396234, + "learning_rate": 3.5816745144553497e-06, + "loss": 0.83654696, + "num_input_tokens_seen": 41425830, + "step": 1941, + "time_per_iteration": 2.6406302452087402 + }, + { + "auxiliary_loss_clip": 0.01235474, + "auxiliary_loss_mlp": 0.01030093, + "balance_loss_clip": 1.05332887, + "balance_loss_mlp": 1.02130783, + "epoch": 0.23351169362111465, + "flos": 13078918419840.0, + "grad_norm": 2.6287876859766914, + "language_loss": 0.75203669, + "learning_rate": 3.5811976432097424e-06, + "loss": 0.7746923, + "num_input_tokens_seen": 41443500, + "step": 1942, + "time_per_iteration": 2.775904417037964 + }, + { + "auxiliary_loss_clip": 0.01221719, + "auxiliary_loss_mlp": 0.01380707, + "balance_loss_clip": 1.06179118, + "balance_loss_mlp": 1.00010276, + "epoch": 0.23363193651175373, + "flos": 15851257931520.0, + "grad_norm": 1.8977508708747963, + "language_loss": 0.84218597, + "learning_rate": 3.58072053209771e-06, + "loss": 0.8682102, + "num_input_tokens_seen": 41460055, + "step": 1943, + "time_per_iteration": 2.659554958343506 + }, + { + "auxiliary_loss_clip": 0.01218757, + "auxiliary_loss_mlp": 0.01032063, + "balance_loss_clip": 1.05352104, + "balance_loss_mlp": 1.0217576, + "epoch": 0.23375217940239285, + "flos": 21025769345280.0, + "grad_norm": 2.187992604324128, + "language_loss": 0.79170728, + "learning_rate": 3.5802431811916296e-06, + "loss": 0.81421554, + "num_input_tokens_seen": 41476665, + "step": 1944, + "time_per_iteration": 3.6186792850494385 + }, + { + "auxiliary_loss_clip": 0.01225014, + "auxiliary_loss_mlp": 0.0103776, + "balance_loss_clip": 1.05759656, + "balance_loss_mlp": 1.02804995, + "epoch": 0.23387242229303193, + "flos": 20594698225920.0, + "grad_norm": 1.736702301206519, + "language_loss": 0.8064031, + "learning_rate": 3.579765590563916e-06, + "loss": 0.82903075, + "num_input_tokens_seen": 41496065, + "step": 1945, + "time_per_iteration": 2.702885389328003 + }, + { + "auxiliary_loss_clip": 0.01205863, + "auxiliary_loss_mlp": 0.01038039, + "balance_loss_clip": 1.05764365, + "balance_loss_mlp": 1.02844882, + "epoch": 0.233992665183671, + "flos": 24279491952000.0, + "grad_norm": 2.124324600953569, + "language_loss": 0.81304449, + "learning_rate": 3.579287760287017e-06, + "loss": 0.83548355, + "num_input_tokens_seen": 41516815, + "step": 1946, + "time_per_iteration": 2.679915428161621 + }, + { + "auxiliary_loss_clip": 0.01216763, + "auxiliary_loss_mlp": 0.01035372, + "balance_loss_clip": 1.0593915, + "balance_loss_mlp": 1.02656806, + "epoch": 0.2341129080743101, + "flos": 30154621121280.0, + "grad_norm": 1.860884515267993, + "language_loss": 0.72412926, + "learning_rate": 3.578809690433421e-06, + "loss": 0.74665058, + "num_input_tokens_seen": 41538525, + "step": 1947, + "time_per_iteration": 2.727325201034546 + }, + { + "auxiliary_loss_clip": 0.01212711, + "auxiliary_loss_mlp": 0.01036898, + "balance_loss_clip": 1.06128788, + "balance_loss_mlp": 1.02754068, + "epoch": 0.2342331509649492, + "flos": 22784135829120.0, + "grad_norm": 2.053746653125953, + "language_loss": 0.81237328, + "learning_rate": 3.578331381075651e-06, + "loss": 0.83486933, + "num_input_tokens_seen": 41559025, + "step": 1948, + "time_per_iteration": 2.6160616874694824 + }, + { + "auxiliary_loss_clip": 0.01217939, + "auxiliary_loss_mlp": 0.01034418, + "balance_loss_clip": 1.05682087, + "balance_loss_mlp": 1.02523255, + "epoch": 0.2343533938555883, + "flos": 23623152687360.0, + "grad_norm": 2.0518781941780855, + "language_loss": 0.69403708, + "learning_rate": 3.5778528322862646e-06, + "loss": 0.71656066, + "num_input_tokens_seen": 41577845, + "step": 1949, + "time_per_iteration": 2.6733686923980713 + }, + { + "auxiliary_loss_clip": 0.0122089, + "auxiliary_loss_mlp": 0.01037307, + "balance_loss_clip": 1.05911422, + "balance_loss_mlp": 1.02846777, + "epoch": 0.23447363674622737, + "flos": 24570332375040.0, + "grad_norm": 1.8235232496764548, + "language_loss": 0.86812252, + "learning_rate": 3.5773740441378585e-06, + "loss": 0.89070451, + "num_input_tokens_seen": 41598600, + "step": 1950, + "time_per_iteration": 2.6721551418304443 + }, + { + "auxiliary_loss_clip": 0.01216016, + "auxiliary_loss_mlp": 0.01030269, + "balance_loss_clip": 1.05845737, + "balance_loss_mlp": 1.02141154, + "epoch": 0.23459387963686648, + "flos": 53140322119680.0, + "grad_norm": 1.7708270301568652, + "language_loss": 0.73663801, + "learning_rate": 3.5768950167030633e-06, + "loss": 0.75910085, + "num_input_tokens_seen": 41623300, + "step": 1951, + "time_per_iteration": 2.9126315116882324 + }, + { + "auxiliary_loss_clip": 0.01211225, + "auxiliary_loss_mlp": 0.01034744, + "balance_loss_clip": 1.05328071, + "balance_loss_mlp": 1.02551079, + "epoch": 0.23471412252750556, + "flos": 23951412103680.0, + "grad_norm": 1.767350866413924, + "language_loss": 0.78334582, + "learning_rate": 3.576415750054548e-06, + "loss": 0.8058055, + "num_input_tokens_seen": 41643420, + "step": 1952, + "time_per_iteration": 2.7255215644836426 + }, + { + "auxiliary_loss_clip": 0.01215833, + "auxiliary_loss_mlp": 0.01036421, + "balance_loss_clip": 1.05388606, + "balance_loss_mlp": 1.02719462, + "epoch": 0.23483436541814465, + "flos": 15706573948800.0, + "grad_norm": 2.0750311697819623, + "language_loss": 0.85994303, + "learning_rate": 3.5759362442650172e-06, + "loss": 0.8824656, + "num_input_tokens_seen": 41660170, + "step": 1953, + "time_per_iteration": 2.8425018787384033 + }, + { + "auxiliary_loss_clip": 0.01217896, + "auxiliary_loss_mlp": 0.01032521, + "balance_loss_clip": 1.05986834, + "balance_loss_mlp": 1.02310371, + "epoch": 0.23495460830878373, + "flos": 24936262179840.0, + "grad_norm": 2.412095937891718, + "language_loss": 0.8528502, + "learning_rate": 3.5754564994072113e-06, + "loss": 0.87535435, + "num_input_tokens_seen": 41679010, + "step": 1954, + "time_per_iteration": 2.686802864074707 + }, + { + "auxiliary_loss_clip": 0.01221343, + "auxiliary_loss_mlp": 0.01033525, + "balance_loss_clip": 1.05446661, + "balance_loss_mlp": 1.02337992, + "epoch": 0.23507485119942284, + "flos": 30482665056000.0, + "grad_norm": 2.7182846138454235, + "language_loss": 0.59891117, + "learning_rate": 3.5749765155539067e-06, + "loss": 0.62145984, + "num_input_tokens_seen": 41699495, + "step": 1955, + "time_per_iteration": 2.7298879623413086 + }, + { + "auxiliary_loss_clip": 0.01228469, + "auxiliary_loss_mlp": 0.01035859, + "balance_loss_clip": 1.0538274, + "balance_loss_mlp": 1.02552307, + "epoch": 0.23519509409006192, + "flos": 18329129746560.0, + "grad_norm": 2.1614666707822625, + "language_loss": 0.92298687, + "learning_rate": 3.574496292777917e-06, + "loss": 0.94563019, + "num_input_tokens_seen": 41717705, + "step": 1956, + "time_per_iteration": 2.795107126235962 + }, + { + "auxiliary_loss_clip": 0.01235945, + "auxiliary_loss_mlp": 0.01043609, + "balance_loss_clip": 1.05922198, + "balance_loss_mlp": 1.03387511, + "epoch": 0.235315336980701, + "flos": 29643217234560.0, + "grad_norm": 2.4662293048346955, + "language_loss": 0.71425998, + "learning_rate": 3.574015831152092e-06, + "loss": 0.73705554, + "num_input_tokens_seen": 41738120, + "step": 1957, + "time_per_iteration": 2.681091785430908 + }, + { + "auxiliary_loss_clip": 0.01213527, + "auxiliary_loss_mlp": 0.01034634, + "balance_loss_clip": 1.05476367, + "balance_loss_mlp": 1.02507997, + "epoch": 0.23543557987134012, + "flos": 18551704371840.0, + "grad_norm": 2.4879956302053587, + "language_loss": 0.83251619, + "learning_rate": 3.573535130749316e-06, + "loss": 0.85499775, + "num_input_tokens_seen": 41756070, + "step": 1958, + "time_per_iteration": 2.699803590774536 + }, + { + "auxiliary_loss_clip": 0.01217571, + "auxiliary_loss_mlp": 0.010347, + "balance_loss_clip": 1.05792451, + "balance_loss_mlp": 1.02513361, + "epoch": 0.2355558227619792, + "flos": 24679033908480.0, + "grad_norm": 1.952972566930701, + "language_loss": 0.73574376, + "learning_rate": 3.5730541916425127e-06, + "loss": 0.75826657, + "num_input_tokens_seen": 41777550, + "step": 1959, + "time_per_iteration": 2.6800248622894287 + }, + { + "auxiliary_loss_clip": 0.01236372, + "auxiliary_loss_mlp": 0.01039942, + "balance_loss_clip": 1.05607867, + "balance_loss_mlp": 1.02977908, + "epoch": 0.23567606565261828, + "flos": 21944795748480.0, + "grad_norm": 3.2267692851668697, + "language_loss": 0.86487567, + "learning_rate": 3.572573013904639e-06, + "loss": 0.88763881, + "num_input_tokens_seen": 41797460, + "step": 1960, + "time_per_iteration": 2.7340621948242188 + }, + { + "auxiliary_loss_clip": 0.01212534, + "auxiliary_loss_mlp": 0.01030368, + "balance_loss_clip": 1.06245458, + "balance_loss_mlp": 1.02162409, + "epoch": 0.2357963085432574, + "flos": 13589352639360.0, + "grad_norm": 2.097000591701735, + "language_loss": 0.9182713, + "learning_rate": 3.572091597608689e-06, + "loss": 0.94070029, + "num_input_tokens_seen": 41815585, + "step": 1961, + "time_per_iteration": 2.6020517349243164 + }, + { + "auxiliary_loss_clip": 0.01233922, + "auxiliary_loss_mlp": 0.01032236, + "balance_loss_clip": 1.05959511, + "balance_loss_mlp": 1.02251434, + "epoch": 0.23591655143389648, + "flos": 22088689632000.0, + "grad_norm": 2.89472100883696, + "language_loss": 0.73182023, + "learning_rate": 3.571609942827694e-06, + "loss": 0.75448179, + "num_input_tokens_seen": 41834700, + "step": 1962, + "time_per_iteration": 2.6959986686706543 + }, + { + "auxiliary_loss_clip": 0.01225372, + "auxiliary_loss_mlp": 0.01033703, + "balance_loss_clip": 1.05669212, + "balance_loss_mlp": 1.02435708, + "epoch": 0.23603679432453556, + "flos": 17017349057280.0, + "grad_norm": 1.8256701890161056, + "language_loss": 0.8844772, + "learning_rate": 3.57112804963472e-06, + "loss": 0.90706801, + "num_input_tokens_seen": 41852915, + "step": 1963, + "time_per_iteration": 2.7158803939819336 + }, + { + "auxiliary_loss_clip": 0.01230442, + "auxiliary_loss_mlp": 0.01033006, + "balance_loss_clip": 1.05838037, + "balance_loss_mlp": 1.02478027, + "epoch": 0.23615703721517464, + "flos": 19171307001600.0, + "grad_norm": 1.7697671576238765, + "language_loss": 0.76904082, + "learning_rate": 3.57064591810287e-06, + "loss": 0.79167533, + "num_input_tokens_seen": 41870415, + "step": 1964, + "time_per_iteration": 3.5800583362579346 + }, + { + "auxiliary_loss_clip": 0.01212387, + "auxiliary_loss_mlp": 0.01380123, + "balance_loss_clip": 1.06384122, + "balance_loss_mlp": 1.00006366, + "epoch": 0.23627728010581375, + "flos": 19098803399040.0, + "grad_norm": 2.3572571444417845, + "language_loss": 0.80456543, + "learning_rate": 3.570163548305284e-06, + "loss": 0.83049059, + "num_input_tokens_seen": 41889345, + "step": 1965, + "time_per_iteration": 3.5344860553741455 + }, + { + "auxiliary_loss_clip": 0.01228021, + "auxiliary_loss_mlp": 0.01031911, + "balance_loss_clip": 1.05907881, + "balance_loss_mlp": 1.02138472, + "epoch": 0.23639752299645284, + "flos": 14282213057280.0, + "grad_norm": 2.163470023834258, + "language_loss": 0.6968388, + "learning_rate": 3.569680940315135e-06, + "loss": 0.71943808, + "num_input_tokens_seen": 41905745, + "step": 1966, + "time_per_iteration": 2.6874806880950928 + }, + { + "auxiliary_loss_clip": 0.01240272, + "auxiliary_loss_mlp": 0.01034558, + "balance_loss_clip": 1.05519927, + "balance_loss_mlp": 1.02485478, + "epoch": 0.23651776588709192, + "flos": 22893411980160.0, + "grad_norm": 2.0030262165746677, + "language_loss": 0.81981659, + "learning_rate": 3.5691980942056356e-06, + "loss": 0.84256488, + "num_input_tokens_seen": 41925115, + "step": 1967, + "time_per_iteration": 3.6187524795532227 + }, + { + "auxiliary_loss_clip": 0.01221197, + "auxiliary_loss_mlp": 0.01031844, + "balance_loss_clip": 1.0578804, + "balance_loss_mlp": 1.02287889, + "epoch": 0.23663800877773103, + "flos": 18624531196800.0, + "grad_norm": 1.6218710303676336, + "language_loss": 0.79451829, + "learning_rate": 3.5687150100500332e-06, + "loss": 0.81704873, + "num_input_tokens_seen": 41944815, + "step": 1968, + "time_per_iteration": 2.697390556335449 + }, + { + "auxiliary_loss_clip": 0.01221419, + "auxiliary_loss_mlp": 0.01039877, + "balance_loss_clip": 1.05865192, + "balance_loss_mlp": 1.02991676, + "epoch": 0.2367582516683701, + "flos": 25555828896000.0, + "grad_norm": 1.631982822856336, + "language_loss": 0.74579334, + "learning_rate": 3.568231687921611e-06, + "loss": 0.76840639, + "num_input_tokens_seen": 41964990, + "step": 1969, + "time_per_iteration": 2.6474392414093018 + }, + { + "auxiliary_loss_clip": 0.01210177, + "auxiliary_loss_mlp": 0.01028701, + "balance_loss_clip": 1.06185699, + "balance_loss_mlp": 1.02009451, + "epoch": 0.2368784945590092, + "flos": 23295072839040.0, + "grad_norm": 1.5196269667380542, + "language_loss": 0.8032701, + "learning_rate": 3.5677481278936883e-06, + "loss": 0.82565892, + "num_input_tokens_seen": 41984570, + "step": 1970, + "time_per_iteration": 3.550915479660034 + }, + { + "auxiliary_loss_clip": 0.01129277, + "auxiliary_loss_mlp": 0.01003783, + "balance_loss_clip": 1.0293715, + "balance_loss_mlp": 1.00168538, + "epoch": 0.23699873744964828, + "flos": 69859291875840.0, + "grad_norm": 0.8503064334274163, + "language_loss": 0.57821238, + "learning_rate": 3.5672643300396214e-06, + "loss": 0.59954298, + "num_input_tokens_seen": 42053715, + "step": 1971, + "time_per_iteration": 3.2962558269500732 + }, + { + "auxiliary_loss_clip": 0.01233411, + "auxiliary_loss_mlp": 0.01031317, + "balance_loss_clip": 1.05422795, + "balance_loss_mlp": 1.02325833, + "epoch": 0.2371189803402874, + "flos": 21835052720640.0, + "grad_norm": 2.6142686081562134, + "language_loss": 0.67589545, + "learning_rate": 3.566780294432802e-06, + "loss": 0.69854271, + "num_input_tokens_seen": 42070890, + "step": 1972, + "time_per_iteration": 2.7265422344207764 + }, + { + "auxiliary_loss_clip": 0.01211445, + "auxiliary_loss_mlp": 0.01030095, + "balance_loss_clip": 1.06295609, + "balance_loss_mlp": 1.02082598, + "epoch": 0.23723922323092647, + "flos": 21908490076800.0, + "grad_norm": 2.354916966452423, + "language_loss": 0.74605286, + "learning_rate": 3.566296021146657e-06, + "loss": 0.76846826, + "num_input_tokens_seen": 42090270, + "step": 1973, + "time_per_iteration": 2.658784866333008 + }, + { + "auxiliary_loss_clip": 0.01216677, + "auxiliary_loss_mlp": 0.0103685, + "balance_loss_clip": 1.06372297, + "balance_loss_mlp": 1.02780235, + "epoch": 0.23735946612156555, + "flos": 32708803380480.0, + "grad_norm": 1.5394274346019958, + "language_loss": 0.7318188, + "learning_rate": 3.565811510254652e-06, + "loss": 0.75435412, + "num_input_tokens_seen": 42111150, + "step": 1974, + "time_per_iteration": 2.7205119132995605 + }, + { + "auxiliary_loss_clip": 0.0111274, + "auxiliary_loss_mlp": 0.00999839, + "balance_loss_clip": 1.02564061, + "balance_loss_mlp": 0.99811071, + "epoch": 0.23747970901220466, + "flos": 70546944821760.0, + "grad_norm": 0.8388312067058108, + "language_loss": 0.58243763, + "learning_rate": 3.5653267618302845e-06, + "loss": 0.60356343, + "num_input_tokens_seen": 42178730, + "step": 1975, + "time_per_iteration": 3.271505117416382 + }, + { + "auxiliary_loss_clip": 0.01210995, + "auxiliary_loss_mlp": 0.01032608, + "balance_loss_clip": 1.06112456, + "balance_loss_mlp": 1.0226723, + "epoch": 0.23759995190284375, + "flos": 20849807594880.0, + "grad_norm": 1.9277854374899888, + "language_loss": 0.85528892, + "learning_rate": 3.564841775947093e-06, + "loss": 0.87772501, + "num_input_tokens_seen": 42199620, + "step": 1976, + "time_per_iteration": 2.665072441101074 + }, + { + "auxiliary_loss_clip": 0.01232247, + "auxiliary_loss_mlp": 0.01030366, + "balance_loss_clip": 1.05474901, + "balance_loss_mlp": 1.02065063, + "epoch": 0.23772019479348283, + "flos": 32921645420160.0, + "grad_norm": 2.146333465093228, + "language_loss": 0.76450872, + "learning_rate": 3.5643565526786475e-06, + "loss": 0.78713495, + "num_input_tokens_seen": 42219560, + "step": 1977, + "time_per_iteration": 2.891266107559204 + }, + { + "auxiliary_loss_clip": 0.01212745, + "auxiliary_loss_mlp": 0.01032066, + "balance_loss_clip": 1.06201351, + "balance_loss_mlp": 1.02286315, + "epoch": 0.2378404376841219, + "flos": 32342765834880.0, + "grad_norm": 2.004209285426471, + "language_loss": 0.77171481, + "learning_rate": 3.5638710920985574e-06, + "loss": 0.79416287, + "num_input_tokens_seen": 42241020, + "step": 1978, + "time_per_iteration": 2.7727720737457275 + }, + { + "auxiliary_loss_clip": 0.01224505, + "auxiliary_loss_mlp": 0.01380479, + "balance_loss_clip": 1.05823815, + "balance_loss_mlp": 1.00009239, + "epoch": 0.23796068057476102, + "flos": 22997624313600.0, + "grad_norm": 2.0147208469444995, + "language_loss": 0.82013327, + "learning_rate": 3.5633853942804655e-06, + "loss": 0.84618312, + "num_input_tokens_seen": 42259345, + "step": 1979, + "time_per_iteration": 2.740449905395508 + }, + { + "auxiliary_loss_clip": 0.01230636, + "auxiliary_loss_mlp": 0.01032026, + "balance_loss_clip": 1.05107069, + "balance_loss_mlp": 1.02241778, + "epoch": 0.2380809234654001, + "flos": 13480938414720.0, + "grad_norm": 2.137309528839119, + "language_loss": 0.7650423, + "learning_rate": 3.5628994592980527e-06, + "loss": 0.78766888, + "num_input_tokens_seen": 42277250, + "step": 1980, + "time_per_iteration": 2.859048366546631 + }, + { + "auxiliary_loss_clip": 0.01210353, + "auxiliary_loss_mlp": 0.01027639, + "balance_loss_clip": 1.05950689, + "balance_loss_mlp": 1.01809609, + "epoch": 0.2382011663560392, + "flos": 16871803148160.0, + "grad_norm": 1.817250359464653, + "language_loss": 0.70298743, + "learning_rate": 3.562413287225034e-06, + "loss": 0.72536731, + "num_input_tokens_seen": 42295360, + "step": 1981, + "time_per_iteration": 2.580620050430298 + }, + { + "auxiliary_loss_clip": 0.01216314, + "auxiliary_loss_mlp": 0.01032668, + "balance_loss_clip": 1.06196654, + "balance_loss_mlp": 1.023018, + "epoch": 0.2383214092466783, + "flos": 18441135331200.0, + "grad_norm": 2.5304622704375315, + "language_loss": 0.88813204, + "learning_rate": 3.5619268781351623e-06, + "loss": 0.91062188, + "num_input_tokens_seen": 42313430, + "step": 1982, + "time_per_iteration": 2.6803364753723145 + }, + { + "auxiliary_loss_clip": 0.0122309, + "auxiliary_loss_mlp": 0.01030273, + "balance_loss_clip": 1.05890036, + "balance_loss_mlp": 1.02128494, + "epoch": 0.23844165213731738, + "flos": 19755717281280.0, + "grad_norm": 2.6683561989828726, + "language_loss": 0.76919401, + "learning_rate": 3.5614402321022256e-06, + "loss": 0.7917276, + "num_input_tokens_seen": 42331260, + "step": 1983, + "time_per_iteration": 2.7972049713134766 + }, + { + "auxiliary_loss_clip": 0.01232361, + "auxiliary_loss_mlp": 0.01029133, + "balance_loss_clip": 1.05098999, + "balance_loss_mlp": 1.01956654, + "epoch": 0.23856189502795647, + "flos": 23367360960000.0, + "grad_norm": 2.0108533783547538, + "language_loss": 0.86941266, + "learning_rate": 3.5609533492000463e-06, + "loss": 0.89202762, + "num_input_tokens_seen": 42350150, + "step": 1984, + "time_per_iteration": 2.9146928787231445 + }, + { + "auxiliary_loss_clip": 0.01229218, + "auxiliary_loss_mlp": 0.01036596, + "balance_loss_clip": 1.06374574, + "balance_loss_mlp": 1.02679753, + "epoch": 0.23868213791859555, + "flos": 23475056912640.0, + "grad_norm": 1.9288205347473244, + "language_loss": 0.78295588, + "learning_rate": 3.560466229502485e-06, + "loss": 0.80561399, + "num_input_tokens_seen": 42369495, + "step": 1985, + "time_per_iteration": 2.7262465953826904 + }, + { + "auxiliary_loss_clip": 0.01225727, + "auxiliary_loss_mlp": 0.01380317, + "balance_loss_clip": 1.06174016, + "balance_loss_mlp": 1.00006413, + "epoch": 0.23880238080923466, + "flos": 16617340224000.0, + "grad_norm": 2.293428901682506, + "language_loss": 0.89727855, + "learning_rate": 3.5599788730834384e-06, + "loss": 0.92333901, + "num_input_tokens_seen": 42387455, + "step": 1986, + "time_per_iteration": 2.652833938598633 + }, + { + "auxiliary_loss_clip": 0.01226508, + "auxiliary_loss_mlp": 0.01036393, + "balance_loss_clip": 1.06100488, + "balance_loss_mlp": 1.02651095, + "epoch": 0.23892262369987374, + "flos": 17348409734400.0, + "grad_norm": 2.351709692306571, + "language_loss": 0.78841263, + "learning_rate": 3.559491280016836e-06, + "loss": 0.81104165, + "num_input_tokens_seen": 42405400, + "step": 1987, + "time_per_iteration": 2.6997182369232178 + }, + { + "auxiliary_loss_clip": 0.01229881, + "auxiliary_loss_mlp": 0.0104009, + "balance_loss_clip": 1.0596596, + "balance_loss_mlp": 1.02979672, + "epoch": 0.23904286659051283, + "flos": 22309899540480.0, + "grad_norm": 2.1178479127651615, + "language_loss": 0.70990014, + "learning_rate": 3.5590034503766465e-06, + "loss": 0.73259985, + "num_input_tokens_seen": 42425065, + "step": 1988, + "time_per_iteration": 2.6304073333740234 + }, + { + "auxiliary_loss_clip": 0.01213309, + "auxiliary_loss_mlp": 0.01040151, + "balance_loss_clip": 1.06269956, + "balance_loss_mlp": 1.03054893, + "epoch": 0.23916310948115194, + "flos": 21178246579200.0, + "grad_norm": 2.333008846984031, + "language_loss": 0.81350845, + "learning_rate": 3.558515384236874e-06, + "loss": 0.83604306, + "num_input_tokens_seen": 42442495, + "step": 1989, + "time_per_iteration": 2.721611499786377 + }, + { + "auxiliary_loss_clip": 0.01226192, + "auxiliary_loss_mlp": 0.01380701, + "balance_loss_clip": 1.05723453, + "balance_loss_mlp": 1.00006938, + "epoch": 0.23928335237179102, + "flos": 14137349506560.0, + "grad_norm": 2.0895312029862008, + "language_loss": 0.84009588, + "learning_rate": 3.558027081671556e-06, + "loss": 0.8661648, + "num_input_tokens_seen": 42459480, + "step": 1990, + "time_per_iteration": 3.5959737300872803 + }, + { + "auxiliary_loss_clip": 0.01218425, + "auxiliary_loss_mlp": 0.01038942, + "balance_loss_clip": 1.05448604, + "balance_loss_mlp": 1.02866578, + "epoch": 0.2394035952624301, + "flos": 23769596436480.0, + "grad_norm": 1.9261198176035144, + "language_loss": 0.68779159, + "learning_rate": 3.557538542754769e-06, + "loss": 0.7103653, + "num_input_tokens_seen": 42479175, + "step": 1991, + "time_per_iteration": 3.625875949859619 + }, + { + "auxiliary_loss_clip": 0.01215019, + "auxiliary_loss_mlp": 0.0103802, + "balance_loss_clip": 1.06418037, + "balance_loss_mlp": 1.02850699, + "epoch": 0.2395238381530692, + "flos": 24206198250240.0, + "grad_norm": 1.9065419897291456, + "language_loss": 0.66809142, + "learning_rate": 3.557049767560623e-06, + "loss": 0.69062179, + "num_input_tokens_seen": 42498090, + "step": 1992, + "time_per_iteration": 3.528292417526245 + }, + { + "auxiliary_loss_clip": 0.01250977, + "auxiliary_loss_mlp": 0.01038087, + "balance_loss_clip": 1.06085372, + "balance_loss_mlp": 1.02891386, + "epoch": 0.2396440810437083, + "flos": 25295763450240.0, + "grad_norm": 1.9198472408742144, + "language_loss": 0.86001039, + "learning_rate": 3.5565607561632655e-06, + "loss": 0.88290107, + "num_input_tokens_seen": 42516930, + "step": 1993, + "time_per_iteration": 2.758660078048706 + }, + { + "auxiliary_loss_clip": 0.01224416, + "auxiliary_loss_mlp": 0.01032039, + "balance_loss_clip": 1.05742395, + "balance_loss_mlp": 1.02208519, + "epoch": 0.23976432393434738, + "flos": 28543093436160.0, + "grad_norm": 2.48266430094434, + "language_loss": 0.79708743, + "learning_rate": 3.5560715086368787e-06, + "loss": 0.81965196, + "num_input_tokens_seen": 42534800, + "step": 1994, + "time_per_iteration": 2.728670358657837 + }, + { + "auxiliary_loss_clip": 0.01223164, + "auxiliary_loss_mlp": 0.01035191, + "balance_loss_clip": 1.05797935, + "balance_loss_mlp": 1.02487314, + "epoch": 0.23988456682498646, + "flos": 19494358945920.0, + "grad_norm": 1.8377072682862738, + "language_loss": 0.82208133, + "learning_rate": 3.5555820250556816e-06, + "loss": 0.84466493, + "num_input_tokens_seen": 42552000, + "step": 1995, + "time_per_iteration": 2.615694284439087 + }, + { + "auxiliary_loss_clip": 0.0123518, + "auxiliary_loss_mlp": 0.01036397, + "balance_loss_clip": 1.06254148, + "balance_loss_mlp": 1.02673531, + "epoch": 0.24000480971562557, + "flos": 20266331068800.0, + "grad_norm": 2.2455780008025763, + "language_loss": 0.6914404, + "learning_rate": 3.5550923054939278e-06, + "loss": 0.71415621, + "num_input_tokens_seen": 42571455, + "step": 1996, + "time_per_iteration": 3.7206201553344727 + }, + { + "auxiliary_loss_clip": 0.01234604, + "auxiliary_loss_mlp": 0.01037261, + "balance_loss_clip": 1.05225992, + "balance_loss_mlp": 1.02730727, + "epoch": 0.24012505260626466, + "flos": 25443176866560.0, + "grad_norm": 1.9629151289158275, + "language_loss": 0.74834079, + "learning_rate": 3.5546023500259083e-06, + "loss": 0.77105945, + "num_input_tokens_seen": 42592550, + "step": 1997, + "time_per_iteration": 3.0455634593963623 + }, + { + "auxiliary_loss_clip": 0.01248324, + "auxiliary_loss_mlp": 0.01029368, + "balance_loss_clip": 1.05449486, + "balance_loss_mlp": 1.02013576, + "epoch": 0.24024529549690374, + "flos": 15553342529280.0, + "grad_norm": 1.7798403290631324, + "language_loss": 0.80241597, + "learning_rate": 3.5541121587259477e-06, + "loss": 0.82519287, + "num_input_tokens_seen": 42610385, + "step": 1998, + "time_per_iteration": 2.817549228668213 + }, + { + "auxiliary_loss_clip": 0.01119546, + "auxiliary_loss_mlp": 0.01010709, + "balance_loss_clip": 1.03210664, + "balance_loss_mlp": 1.00844407, + "epoch": 0.24036553838754285, + "flos": 57122351867520.0, + "grad_norm": 0.8561136053053718, + "language_loss": 0.57908118, + "learning_rate": 3.553621731668408e-06, + "loss": 0.60038376, + "num_input_tokens_seen": 42673595, + "step": 1999, + "time_per_iteration": 3.262378215789795 + }, + { + "auxiliary_loss_clip": 0.01212691, + "auxiliary_loss_mlp": 0.01028045, + "balance_loss_clip": 1.05748081, + "balance_loss_mlp": 1.01887155, + "epoch": 0.24048578127818193, + "flos": 24969946158720.0, + "grad_norm": 2.0672784282187737, + "language_loss": 0.83237886, + "learning_rate": 3.553131068927688e-06, + "loss": 0.85478622, + "num_input_tokens_seen": 42692000, + "step": 2000, + "time_per_iteration": 2.7770841121673584 + }, + { + "auxiliary_loss_clip": 0.01237661, + "auxiliary_loss_mlp": 0.01028762, + "balance_loss_clip": 1.05942309, + "balance_loss_mlp": 1.01938009, + "epoch": 0.24060602416882101, + "flos": 23330947547520.0, + "grad_norm": 1.6710593159468126, + "language_loss": 0.79992861, + "learning_rate": 3.552640170578219e-06, + "loss": 0.82259285, + "num_input_tokens_seen": 42712250, + "step": 2001, + "time_per_iteration": 2.744896411895752 + }, + { + "auxiliary_loss_clip": 0.01225745, + "auxiliary_loss_mlp": 0.0103462, + "balance_loss_clip": 1.05627871, + "balance_loss_mlp": 1.02489853, + "epoch": 0.2407262670594601, + "flos": 14173260128640.0, + "grad_norm": 2.1434548554299147, + "language_loss": 0.77903414, + "learning_rate": 3.5521490366944703e-06, + "loss": 0.80163777, + "num_input_tokens_seen": 42729900, + "step": 2002, + "time_per_iteration": 2.670137882232666 + }, + { + "auxiliary_loss_clip": 0.01237105, + "auxiliary_loss_mlp": 0.01034416, + "balance_loss_clip": 1.05906403, + "balance_loss_mlp": 1.02445006, + "epoch": 0.2408465099500992, + "flos": 13663113217920.0, + "grad_norm": 2.168396588296961, + "language_loss": 0.80213141, + "learning_rate": 3.5516576673509474e-06, + "loss": 0.82484663, + "num_input_tokens_seen": 42747900, + "step": 2003, + "time_per_iteration": 2.6895358562469482 + }, + { + "auxiliary_loss_clip": 0.01216869, + "auxiliary_loss_mlp": 0.01035402, + "balance_loss_clip": 1.06448793, + "balance_loss_mlp": 1.02546632, + "epoch": 0.2409667528407383, + "flos": 31248029076480.0, + "grad_norm": 1.695805482744119, + "language_loss": 0.86252689, + "learning_rate": 3.5511660626221896e-06, + "loss": 0.88504958, + "num_input_tokens_seen": 42768540, + "step": 2004, + "time_per_iteration": 2.781017541885376 + }, + { + "auxiliary_loss_clip": 0.01226319, + "auxiliary_loss_mlp": 0.01380602, + "balance_loss_clip": 1.05989051, + "balance_loss_mlp": 1.000103, + "epoch": 0.24108699573137737, + "flos": 22199941031040.0, + "grad_norm": 2.7970015721661077, + "language_loss": 0.89238048, + "learning_rate": 3.5506742225827744e-06, + "loss": 0.9184497, + "num_input_tokens_seen": 42785395, + "step": 2005, + "time_per_iteration": 2.6846470832824707 + }, + { + "auxiliary_loss_clip": 0.01238141, + "auxiliary_loss_mlp": 0.01037884, + "balance_loss_clip": 1.0598371, + "balance_loss_mlp": 1.02796543, + "epoch": 0.24120723862201648, + "flos": 26103035664000.0, + "grad_norm": 2.191002222068962, + "language_loss": 0.90024829, + "learning_rate": 3.5501821473073116e-06, + "loss": 0.9230085, + "num_input_tokens_seen": 42801980, + "step": 2006, + "time_per_iteration": 2.8114535808563232 + }, + { + "auxiliary_loss_clip": 0.01230465, + "auxiliary_loss_mlp": 0.01034684, + "balance_loss_clip": 1.05699253, + "balance_loss_mlp": 1.0248667, + "epoch": 0.24132748151265557, + "flos": 18624926246400.0, + "grad_norm": 2.0657233708876546, + "language_loss": 0.86696666, + "learning_rate": 3.54968983687045e-06, + "loss": 0.88961816, + "num_input_tokens_seen": 42818850, + "step": 2007, + "time_per_iteration": 2.7794418334960938 + }, + { + "auxiliary_loss_clip": 0.01229235, + "auxiliary_loss_mlp": 0.01038317, + "balance_loss_clip": 1.05770659, + "balance_loss_mlp": 1.02802324, + "epoch": 0.24144772440329465, + "flos": 15267673664640.0, + "grad_norm": 4.287532904838505, + "language_loss": 0.8953976, + "learning_rate": 3.549197291346872e-06, + "loss": 0.91807306, + "num_input_tokens_seen": 42835375, + "step": 2008, + "time_per_iteration": 2.826183557510376 + }, + { + "auxiliary_loss_clip": 0.01222348, + "auxiliary_loss_mlp": 0.01035191, + "balance_loss_clip": 1.06046057, + "balance_loss_mlp": 1.02508807, + "epoch": 0.24156796729393373, + "flos": 24024274842240.0, + "grad_norm": 2.101311223973695, + "language_loss": 0.79327649, + "learning_rate": 3.548704510811297e-06, + "loss": 0.81585187, + "num_input_tokens_seen": 42854570, + "step": 2009, + "time_per_iteration": 2.7403812408447266 + }, + { + "auxiliary_loss_clip": 0.01245793, + "auxiliary_loss_mlp": 0.01034782, + "balance_loss_clip": 1.05415606, + "balance_loss_mlp": 1.0250423, + "epoch": 0.24168821018457284, + "flos": 26286790665600.0, + "grad_norm": 3.48109067590751, + "language_loss": 0.74886036, + "learning_rate": 3.5482114953384787e-06, + "loss": 0.77166617, + "num_input_tokens_seen": 42873800, + "step": 2010, + "time_per_iteration": 2.8605992794036865 + }, + { + "auxiliary_loss_clip": 0.01223247, + "auxiliary_loss_mlp": 0.01031633, + "balance_loss_clip": 1.05993831, + "balance_loss_mlp": 1.02263856, + "epoch": 0.24180845307521193, + "flos": 18223193560320.0, + "grad_norm": 1.8937378842739292, + "language_loss": 0.84650028, + "learning_rate": 3.5477182450032077e-06, + "loss": 0.86904907, + "num_input_tokens_seen": 42892400, + "step": 2011, + "time_per_iteration": 2.7875893115997314 + }, + { + "auxiliary_loss_clip": 0.0121892, + "auxiliary_loss_mlp": 0.01044616, + "balance_loss_clip": 1.06033731, + "balance_loss_mlp": 1.03512144, + "epoch": 0.241928695965851, + "flos": 20449260057600.0, + "grad_norm": 2.4664147963730905, + "language_loss": 0.83125722, + "learning_rate": 3.5472247598803097e-06, + "loss": 0.85389251, + "num_input_tokens_seen": 42911745, + "step": 2012, + "time_per_iteration": 2.7106575965881348 + }, + { + "auxiliary_loss_clip": 0.01215754, + "auxiliary_loss_mlp": 0.01033947, + "balance_loss_clip": 1.06288457, + "balance_loss_mlp": 1.02295566, + "epoch": 0.24204893885649012, + "flos": 25556475340800.0, + "grad_norm": 2.125528250228635, + "language_loss": 0.85821205, + "learning_rate": 3.546731040044645e-06, + "loss": 0.88070905, + "num_input_tokens_seen": 42926915, + "step": 2013, + "time_per_iteration": 2.684868097305298 + }, + { + "auxiliary_loss_clip": 0.01213189, + "auxiliary_loss_mlp": 0.01027272, + "balance_loss_clip": 1.06345844, + "balance_loss_mlp": 1.01838541, + "epoch": 0.2421691817471292, + "flos": 30660207004800.0, + "grad_norm": 1.7901222947136735, + "language_loss": 0.75262296, + "learning_rate": 3.546237085571112e-06, + "loss": 0.77502757, + "num_input_tokens_seen": 42945350, + "step": 2014, + "time_per_iteration": 2.6804592609405518 + }, + { + "auxiliary_loss_clip": 0.0122326, + "auxiliary_loss_mlp": 0.01037128, + "balance_loss_clip": 1.0616101, + "balance_loss_mlp": 1.02712619, + "epoch": 0.24228942463776829, + "flos": 21945011230080.0, + "grad_norm": 1.8991873765981548, + "language_loss": 0.7254039, + "learning_rate": 3.5457428965346425e-06, + "loss": 0.74800777, + "num_input_tokens_seen": 42964290, + "step": 2015, + "time_per_iteration": 2.658132314682007 + }, + { + "auxiliary_loss_clip": 0.01249396, + "auxiliary_loss_mlp": 0.01031454, + "balance_loss_clip": 1.05200362, + "balance_loss_mlp": 1.0224179, + "epoch": 0.2424096675284074, + "flos": 33984493879680.0, + "grad_norm": 1.7508851768913711, + "language_loss": 0.74963117, + "learning_rate": 3.545248473010205e-06, + "loss": 0.77243966, + "num_input_tokens_seen": 42987095, + "step": 2016, + "time_per_iteration": 4.027482748031616 + }, + { + "auxiliary_loss_clip": 0.01214895, + "auxiliary_loss_mlp": 0.01380902, + "balance_loss_clip": 1.06192231, + "balance_loss_mlp": 1.00011158, + "epoch": 0.24252991041904648, + "flos": 21653416621440.0, + "grad_norm": 1.7882014809858182, + "language_loss": 0.87480783, + "learning_rate": 3.544753815072802e-06, + "loss": 0.90076584, + "num_input_tokens_seen": 43005750, + "step": 2017, + "time_per_iteration": 3.734201669692993 + }, + { + "auxiliary_loss_clip": 0.01251574, + "auxiliary_loss_mlp": 0.0103609, + "balance_loss_clip": 1.04676247, + "balance_loss_mlp": 1.02615964, + "epoch": 0.24265015330968556, + "flos": 21870065502720.0, + "grad_norm": 1.858207190342427, + "language_loss": 0.88157237, + "learning_rate": 3.544258922797474e-06, + "loss": 0.90444899, + "num_input_tokens_seen": 43023870, + "step": 2018, + "time_per_iteration": 3.7980294227600098 + }, + { + "auxiliary_loss_clip": 0.01210998, + "auxiliary_loss_mlp": 0.01032699, + "balance_loss_clip": 1.06131685, + "balance_loss_mlp": 1.02362764, + "epoch": 0.24277039620032465, + "flos": 25628260671360.0, + "grad_norm": 1.7346316405057196, + "language_loss": 0.78268087, + "learning_rate": 3.543763796259295e-06, + "loss": 0.80511785, + "num_input_tokens_seen": 43043825, + "step": 2019, + "time_per_iteration": 2.936100959777832 + }, + { + "auxiliary_loss_clip": 0.01219696, + "auxiliary_loss_mlp": 0.01031255, + "balance_loss_clip": 1.05934346, + "balance_loss_mlp": 1.0214622, + "epoch": 0.24289063909096376, + "flos": 26286575184000.0, + "grad_norm": 1.8369128144211242, + "language_loss": 0.90837026, + "learning_rate": 3.5432684355333754e-06, + "loss": 0.93087977, + "num_input_tokens_seen": 43062480, + "step": 2020, + "time_per_iteration": 2.709251880645752 + }, + { + "auxiliary_loss_clip": 0.01219618, + "auxiliary_loss_mlp": 0.0103254, + "balance_loss_clip": 1.05981731, + "balance_loss_mlp": 1.02315283, + "epoch": 0.24301088198160284, + "flos": 25075056332160.0, + "grad_norm": 1.8423011213128584, + "language_loss": 0.76866627, + "learning_rate": 3.5427728406948613e-06, + "loss": 0.79118782, + "num_input_tokens_seen": 43081595, + "step": 2021, + "time_per_iteration": 2.6898670196533203 + }, + { + "auxiliary_loss_clip": 0.01129732, + "auxiliary_loss_mlp": 0.01006185, + "balance_loss_clip": 1.02736783, + "balance_loss_mlp": 1.00418222, + "epoch": 0.24313112487224192, + "flos": 69900948673920.0, + "grad_norm": 0.7745373753242537, + "language_loss": 0.57904786, + "learning_rate": 3.542277011818934e-06, + "loss": 0.60040706, + "num_input_tokens_seen": 43145430, + "step": 2022, + "time_per_iteration": 4.732370853424072 + }, + { + "auxiliary_loss_clip": 0.01227424, + "auxiliary_loss_mlp": 0.01036454, + "balance_loss_clip": 1.06007385, + "balance_loss_mlp": 1.02714968, + "epoch": 0.24325136776288103, + "flos": 40662334235520.0, + "grad_norm": 2.14413584841978, + "language_loss": 0.74103069, + "learning_rate": 3.5417809489808104e-06, + "loss": 0.76366949, + "num_input_tokens_seen": 43167040, + "step": 2023, + "time_per_iteration": 2.9479169845581055 + }, + { + "auxiliary_loss_clip": 0.01221463, + "auxiliary_loss_mlp": 0.01034917, + "balance_loss_clip": 1.06046176, + "balance_loss_mlp": 1.02558279, + "epoch": 0.24337161065352012, + "flos": 25046400257280.0, + "grad_norm": 1.7671063561276381, + "language_loss": 0.72254807, + "learning_rate": 3.5412846522557422e-06, + "loss": 0.74511182, + "num_input_tokens_seen": 43187930, + "step": 2024, + "time_per_iteration": 2.8081274032592773 + }, + { + "auxiliary_loss_clip": 0.01211311, + "auxiliary_loss_mlp": 0.01038368, + "balance_loss_clip": 1.0618372, + "balance_loss_mlp": 1.02805042, + "epoch": 0.2434918535441592, + "flos": 18661160090880.0, + "grad_norm": 2.25644409096004, + "language_loss": 0.7409718, + "learning_rate": 3.540788121719018e-06, + "loss": 0.76346868, + "num_input_tokens_seen": 43206350, + "step": 2025, + "time_per_iteration": 2.6049070358276367 + }, + { + "auxiliary_loss_clip": 0.01230394, + "auxiliary_loss_mlp": 0.01033112, + "balance_loss_clip": 1.05923581, + "balance_loss_mlp": 1.02418923, + "epoch": 0.24361209643479828, + "flos": 23915142345600.0, + "grad_norm": 2.072010927290113, + "language_loss": 0.81900847, + "learning_rate": 3.5402913574459604e-06, + "loss": 0.84164357, + "num_input_tokens_seen": 43226255, + "step": 2026, + "time_per_iteration": 2.731055736541748 + }, + { + "auxiliary_loss_clip": 0.01241573, + "auxiliary_loss_mlp": 0.0103137, + "balance_loss_clip": 1.0485692, + "balance_loss_mlp": 1.02209568, + "epoch": 0.2437323393254374, + "flos": 28657505232000.0, + "grad_norm": 1.6987814109184582, + "language_loss": 0.85835624, + "learning_rate": 3.5397943595119297e-06, + "loss": 0.88108563, + "num_input_tokens_seen": 43247675, + "step": 2027, + "time_per_iteration": 2.9259192943573 + }, + { + "auxiliary_loss_clip": 0.01224196, + "auxiliary_loss_mlp": 0.01035692, + "balance_loss_clip": 1.05983663, + "balance_loss_mlp": 1.0256784, + "epoch": 0.24385258221607647, + "flos": 23550325862400.0, + "grad_norm": 3.047386757245122, + "language_loss": 0.77385378, + "learning_rate": 3.5392971279923177e-06, + "loss": 0.79645264, + "num_input_tokens_seen": 43265895, + "step": 2028, + "time_per_iteration": 2.923021078109741 + }, + { + "auxiliary_loss_clip": 0.01228053, + "auxiliary_loss_mlp": 0.01037911, + "balance_loss_clip": 1.05362427, + "balance_loss_mlp": 1.02776074, + "epoch": 0.24397282510671556, + "flos": 25336091445120.0, + "grad_norm": 1.9766565308290003, + "language_loss": 0.82924151, + "learning_rate": 3.5387996629625557e-06, + "loss": 0.85190117, + "num_input_tokens_seen": 43283485, + "step": 2029, + "time_per_iteration": 2.7792038917541504 + }, + { + "auxiliary_loss_clip": 0.01101407, + "auxiliary_loss_mlp": 0.01003694, + "balance_loss_clip": 1.02221203, + "balance_loss_mlp": 1.00170279, + "epoch": 0.24409306799735467, + "flos": 65187421430400.0, + "grad_norm": 0.807590172358764, + "language_loss": 0.54984045, + "learning_rate": 3.5383019644981083e-06, + "loss": 0.57089138, + "num_input_tokens_seen": 43347180, + "step": 2030, + "time_per_iteration": 3.264082431793213 + }, + { + "auxiliary_loss_clip": 0.0122578, + "auxiliary_loss_mlp": 0.01029262, + "balance_loss_clip": 1.05791628, + "balance_loss_mlp": 1.01927233, + "epoch": 0.24421331088799375, + "flos": 19537093152000.0, + "grad_norm": 2.2293157518348155, + "language_loss": 0.73154575, + "learning_rate": 3.5378040326744763e-06, + "loss": 0.75409621, + "num_input_tokens_seen": 43366665, + "step": 2031, + "time_per_iteration": 2.6741816997528076 + }, + { + "auxiliary_loss_clip": 0.01238256, + "auxiliary_loss_mlp": 0.01025278, + "balance_loss_clip": 1.06045985, + "balance_loss_mlp": 1.01586032, + "epoch": 0.24433355377863283, + "flos": 21068575378560.0, + "grad_norm": 2.132296940135181, + "language_loss": 0.85516703, + "learning_rate": 3.5373058675671946e-06, + "loss": 0.87780237, + "num_input_tokens_seen": 43384670, + "step": 2032, + "time_per_iteration": 2.6791348457336426 + }, + { + "auxiliary_loss_clip": 0.0123366, + "auxiliary_loss_mlp": 0.01031082, + "balance_loss_clip": 1.05359197, + "balance_loss_mlp": 1.02135432, + "epoch": 0.24445379666927192, + "flos": 22637189289600.0, + "grad_norm": 1.8664954769675217, + "language_loss": 0.72390699, + "learning_rate": 3.536807469251836e-06, + "loss": 0.74655437, + "num_input_tokens_seen": 43403825, + "step": 2033, + "time_per_iteration": 2.744872570037842 + }, + { + "auxiliary_loss_clip": 0.01239893, + "auxiliary_loss_mlp": 0.01033455, + "balance_loss_clip": 1.05559325, + "balance_loss_mlp": 1.02399564, + "epoch": 0.24457403955991103, + "flos": 21251612108160.0, + "grad_norm": 2.0668055624918984, + "language_loss": 0.82956976, + "learning_rate": 3.5363088378040055e-06, + "loss": 0.85230321, + "num_input_tokens_seen": 43422715, + "step": 2034, + "time_per_iteration": 2.6906561851501465 + }, + { + "auxiliary_loss_clip": 0.01099402, + "auxiliary_loss_mlp": 0.01375598, + "balance_loss_clip": 1.02042818, + "balance_loss_mlp": 0.9998371, + "epoch": 0.2446942824505501, + "flos": 66997820764800.0, + "grad_norm": 0.7783830919643595, + "language_loss": 0.64370143, + "learning_rate": 3.5358099732993463e-06, + "loss": 0.66845143, + "num_input_tokens_seen": 43481825, + "step": 2035, + "time_per_iteration": 3.115248203277588 + }, + { + "auxiliary_loss_clip": 0.01231643, + "auxiliary_loss_mlp": 0.01030222, + "balance_loss_clip": 1.0574069, + "balance_loss_mlp": 1.02112007, + "epoch": 0.2448145253411892, + "flos": 20411122792320.0, + "grad_norm": 2.5429922958455258, + "language_loss": 0.89275479, + "learning_rate": 3.535310875813535e-06, + "loss": 0.91537344, + "num_input_tokens_seen": 43500220, + "step": 2036, + "time_per_iteration": 2.728522300720215 + }, + { + "auxiliary_loss_clip": 0.01218942, + "auxiliary_loss_mlp": 0.01030249, + "balance_loss_clip": 1.05957353, + "balance_loss_mlp": 1.02053356, + "epoch": 0.2449347682318283, + "flos": 28804739080320.0, + "grad_norm": 1.7445730857012194, + "language_loss": 0.81697673, + "learning_rate": 3.5348115454222843e-06, + "loss": 0.8394686, + "num_input_tokens_seen": 43522805, + "step": 2037, + "time_per_iteration": 2.7349536418914795 + }, + { + "auxiliary_loss_clip": 0.01219606, + "auxiliary_loss_mlp": 0.01032765, + "balance_loss_clip": 1.05417371, + "balance_loss_mlp": 1.02377057, + "epoch": 0.2450550111224674, + "flos": 22528990546560.0, + "grad_norm": 1.77850681421698, + "language_loss": 0.85738379, + "learning_rate": 3.5343119822013425e-06, + "loss": 0.87990755, + "num_input_tokens_seen": 43541915, + "step": 2038, + "time_per_iteration": 2.746840715408325 + }, + { + "auxiliary_loss_clip": 0.01223715, + "auxiliary_loss_mlp": 0.01039017, + "balance_loss_clip": 1.05988014, + "balance_loss_mlp": 1.02897906, + "epoch": 0.24517525401310647, + "flos": 21759137326080.0, + "grad_norm": 9.859016094664346, + "language_loss": 0.77613795, + "learning_rate": 3.533812186226493e-06, + "loss": 0.7987653, + "num_input_tokens_seen": 43562625, + "step": 2039, + "time_per_iteration": 2.629817485809326 + }, + { + "auxiliary_loss_clip": 0.01207332, + "auxiliary_loss_mlp": 0.01030967, + "balance_loss_clip": 1.05914807, + "balance_loss_mlp": 1.02218151, + "epoch": 0.24529549690374555, + "flos": 25043311687680.0, + "grad_norm": 1.921088503109026, + "language_loss": 0.75752687, + "learning_rate": 3.5333121575735545e-06, + "loss": 0.77990985, + "num_input_tokens_seen": 43582265, + "step": 2040, + "time_per_iteration": 2.7183899879455566 + }, + { + "auxiliary_loss_clip": 0.01228077, + "auxiliary_loss_mlp": 0.01033828, + "balance_loss_clip": 1.0590378, + "balance_loss_mlp": 1.02537024, + "epoch": 0.24541573979438466, + "flos": 32123638915200.0, + "grad_norm": 1.8673500839837094, + "language_loss": 0.75950623, + "learning_rate": 3.532811896318381e-06, + "loss": 0.78212529, + "num_input_tokens_seen": 43604335, + "step": 2041, + "time_per_iteration": 2.76538348197937 + }, + { + "auxiliary_loss_clip": 0.01239157, + "auxiliary_loss_mlp": 0.01033736, + "balance_loss_clip": 1.05652046, + "balance_loss_mlp": 1.0238533, + "epoch": 0.24553598268502375, + "flos": 31357556622720.0, + "grad_norm": 2.994946941901402, + "language_loss": 0.81945181, + "learning_rate": 3.5323114025368615e-06, + "loss": 0.84218073, + "num_input_tokens_seen": 43619400, + "step": 2042, + "time_per_iteration": 3.6924400329589844 + }, + { + "auxiliary_loss_clip": 0.01210269, + "auxiliary_loss_mlp": 0.01037954, + "balance_loss_clip": 1.05445611, + "balance_loss_mlp": 1.02870965, + "epoch": 0.24565622557566283, + "flos": 14027462824320.0, + "grad_norm": 2.0548334998064495, + "language_loss": 0.82092798, + "learning_rate": 3.53181067630492e-06, + "loss": 0.84341019, + "num_input_tokens_seen": 43636870, + "step": 2043, + "time_per_iteration": 3.5574631690979004 + }, + { + "auxiliary_loss_clip": 0.01216236, + "auxiliary_loss_mlp": 0.01034981, + "balance_loss_clip": 1.0542841, + "balance_loss_mlp": 1.02499747, + "epoch": 0.24577646846630194, + "flos": 16581465515520.0, + "grad_norm": 2.703574482438782, + "language_loss": 0.75811613, + "learning_rate": 3.5313097176985175e-06, + "loss": 0.78062832, + "num_input_tokens_seen": 43655180, + "step": 2044, + "time_per_iteration": 3.5540263652801514 + }, + { + "auxiliary_loss_clip": 0.01220093, + "auxiliary_loss_mlp": 0.01030454, + "balance_loss_clip": 1.06000245, + "balance_loss_mlp": 1.0209651, + "epoch": 0.24589671135694102, + "flos": 18807424272000.0, + "grad_norm": 1.844643571232828, + "language_loss": 0.81407285, + "learning_rate": 3.5308085267936482e-06, + "loss": 0.83657831, + "num_input_tokens_seen": 43672895, + "step": 2045, + "time_per_iteration": 2.5885426998138428 + }, + { + "auxiliary_loss_clip": 0.01257957, + "auxiliary_loss_mlp": 0.01379709, + "balance_loss_clip": 1.05668676, + "balance_loss_mlp": 1.00009608, + "epoch": 0.2460169542475801, + "flos": 19938538529280.0, + "grad_norm": 1.6631793014145086, + "language_loss": 0.89960706, + "learning_rate": 3.530307103666342e-06, + "loss": 0.92598373, + "num_input_tokens_seen": 43691975, + "step": 2046, + "time_per_iteration": 2.7467620372772217 + }, + { + "auxiliary_loss_clip": 0.01240227, + "auxiliary_loss_mlp": 0.01037165, + "balance_loss_clip": 1.05695558, + "balance_loss_mlp": 1.02695513, + "epoch": 0.24613719713821922, + "flos": 24171221381760.0, + "grad_norm": 1.949944028510636, + "language_loss": 0.80404675, + "learning_rate": 3.5298054483926658e-06, + "loss": 0.82682073, + "num_input_tokens_seen": 43712670, + "step": 2047, + "time_per_iteration": 2.7142245769500732 + }, + { + "auxiliary_loss_clip": 0.01224322, + "auxiliary_loss_mlp": 0.01033212, + "balance_loss_clip": 1.05868495, + "balance_loss_mlp": 1.02356815, + "epoch": 0.2462574400288583, + "flos": 30221055325440.0, + "grad_norm": 2.883779632020731, + "language_loss": 0.82961524, + "learning_rate": 3.5293035610487187e-06, + "loss": 0.85219049, + "num_input_tokens_seen": 43732035, + "step": 2048, + "time_per_iteration": 3.632497549057007 + }, + { + "auxiliary_loss_clip": 0.01111619, + "auxiliary_loss_mlp": 0.01008291, + "balance_loss_clip": 1.01427555, + "balance_loss_mlp": 1.00628817, + "epoch": 0.24637768291949738, + "flos": 68943030819840.0, + "grad_norm": 0.7294651415359925, + "language_loss": 0.61951756, + "learning_rate": 3.5288014417106374e-06, + "loss": 0.64071667, + "num_input_tokens_seen": 43798055, + "step": 2049, + "time_per_iteration": 3.2497286796569824 + }, + { + "auxiliary_loss_clip": 0.01232837, + "auxiliary_loss_mlp": 0.01035182, + "balance_loss_clip": 1.0561018, + "balance_loss_mlp": 1.02511454, + "epoch": 0.24649792581013646, + "flos": 34383999922560.0, + "grad_norm": 1.927642426177276, + "language_loss": 0.75321186, + "learning_rate": 3.528299090454593e-06, + "loss": 0.77589208, + "num_input_tokens_seen": 43818590, + "step": 2050, + "time_per_iteration": 2.8483943939208984 + }, + { + "auxiliary_loss_clip": 0.01218853, + "auxiliary_loss_mlp": 0.01032133, + "balance_loss_clip": 1.05661178, + "balance_loss_mlp": 1.02291179, + "epoch": 0.24661816870077558, + "flos": 19680448331520.0, + "grad_norm": 2.271796053286622, + "language_loss": 0.82540429, + "learning_rate": 3.527796507356792e-06, + "loss": 0.8479141, + "num_input_tokens_seen": 43832480, + "step": 2051, + "time_per_iteration": 2.669682025909424 + }, + { + "auxiliary_loss_clip": 0.01224236, + "auxiliary_loss_mlp": 0.01032131, + "balance_loss_clip": 1.06038272, + "balance_loss_mlp": 1.0226959, + "epoch": 0.24673841159141466, + "flos": 20002279213440.0, + "grad_norm": 3.074510859881675, + "language_loss": 0.9020102, + "learning_rate": 3.527293692493475e-06, + "loss": 0.92457384, + "num_input_tokens_seen": 43848345, + "step": 2052, + "time_per_iteration": 2.6318836212158203 + }, + { + "auxiliary_loss_clip": 0.01218705, + "auxiliary_loss_mlp": 0.01039711, + "balance_loss_clip": 1.05718255, + "balance_loss_mlp": 1.02951837, + "epoch": 0.24685865448205374, + "flos": 21646593037440.0, + "grad_norm": 2.1573914909039456, + "language_loss": 0.72980338, + "learning_rate": 3.52679064594092e-06, + "loss": 0.75238752, + "num_input_tokens_seen": 43865685, + "step": 2053, + "time_per_iteration": 2.7400968074798584 + }, + { + "auxiliary_loss_clip": 0.01224243, + "auxiliary_loss_mlp": 0.01034573, + "balance_loss_clip": 1.04557729, + "balance_loss_mlp": 1.0251615, + "epoch": 0.24697889737269285, + "flos": 17960470508160.0, + "grad_norm": 2.453698078200559, + "language_loss": 0.74846709, + "learning_rate": 3.5262873677754375e-06, + "loss": 0.77105522, + "num_input_tokens_seen": 43883690, + "step": 2054, + "time_per_iteration": 3.04030442237854 + }, + { + "auxiliary_loss_clip": 0.01210014, + "auxiliary_loss_mlp": 0.01030057, + "balance_loss_clip": 1.06071162, + "balance_loss_mlp": 1.02081192, + "epoch": 0.24709914026333193, + "flos": 27344611221120.0, + "grad_norm": 1.703991849851044, + "language_loss": 0.80250013, + "learning_rate": 3.5257838580733745e-06, + "loss": 0.82490087, + "num_input_tokens_seen": 43903295, + "step": 2055, + "time_per_iteration": 2.712836265563965 + }, + { + "auxiliary_loss_clip": 0.0122349, + "auxiliary_loss_mlp": 0.01029203, + "balance_loss_clip": 1.05998707, + "balance_loss_mlp": 1.02005935, + "epoch": 0.24721938315397102, + "flos": 19275519335040.0, + "grad_norm": 2.3049524780301716, + "language_loss": 0.87080294, + "learning_rate": 3.5252801169111138e-06, + "loss": 0.89332986, + "num_input_tokens_seen": 43920960, + "step": 2056, + "time_per_iteration": 2.615565538406372 + }, + { + "auxiliary_loss_clip": 0.01224746, + "auxiliary_loss_mlp": 0.01031152, + "balance_loss_clip": 1.05878258, + "balance_loss_mlp": 1.0220567, + "epoch": 0.2473396260446101, + "flos": 23185796688000.0, + "grad_norm": 1.8583298488158333, + "language_loss": 0.79984415, + "learning_rate": 3.524776144365072e-06, + "loss": 0.82240307, + "num_input_tokens_seen": 43939415, + "step": 2057, + "time_per_iteration": 2.6894819736480713 + }, + { + "auxiliary_loss_clip": 0.01218242, + "auxiliary_loss_mlp": 0.01031601, + "balance_loss_clip": 1.05750084, + "balance_loss_mlp": 1.0220046, + "epoch": 0.2474598689352492, + "flos": 21142443697920.0, + "grad_norm": 1.635224657409408, + "language_loss": 0.79136765, + "learning_rate": 3.5242719405117016e-06, + "loss": 0.81386608, + "num_input_tokens_seen": 43959220, + "step": 2058, + "time_per_iteration": 2.7062013149261475 + }, + { + "auxiliary_loss_clip": 0.01228658, + "auxiliary_loss_mlp": 0.01380217, + "balance_loss_clip": 1.05917811, + "balance_loss_mlp": 1.00012231, + "epoch": 0.2475801118258883, + "flos": 21648352803840.0, + "grad_norm": 2.514840162354125, + "language_loss": 0.75136423, + "learning_rate": 3.5237675054274893e-06, + "loss": 0.77745295, + "num_input_tokens_seen": 43978420, + "step": 2059, + "time_per_iteration": 2.698370933532715 + }, + { + "auxiliary_loss_clip": 0.01220271, + "auxiliary_loss_mlp": 0.01030708, + "balance_loss_clip": 1.06014228, + "balance_loss_mlp": 1.02119553, + "epoch": 0.24770035471652738, + "flos": 22674500542080.0, + "grad_norm": 7.3655694562514675, + "language_loss": 0.80395854, + "learning_rate": 3.5232628391889584e-06, + "loss": 0.82646823, + "num_input_tokens_seen": 43996710, + "step": 2060, + "time_per_iteration": 2.630218744277954 + }, + { + "auxiliary_loss_clip": 0.0123939, + "auxiliary_loss_mlp": 0.01027487, + "balance_loss_clip": 1.05472362, + "balance_loss_mlp": 1.01851666, + "epoch": 0.2478205976071665, + "flos": 22163814927360.0, + "grad_norm": 2.1998158716787417, + "language_loss": 0.64622945, + "learning_rate": 3.522757941872666e-06, + "loss": 0.66889822, + "num_input_tokens_seen": 44014865, + "step": 2061, + "time_per_iteration": 2.7618143558502197 + }, + { + "auxiliary_loss_clip": 0.01215003, + "auxiliary_loss_mlp": 0.01380039, + "balance_loss_clip": 1.06495523, + "balance_loss_mlp": 1.000103, + "epoch": 0.24794084049780557, + "flos": 24973106555520.0, + "grad_norm": 1.5265644155811233, + "language_loss": 0.82403606, + "learning_rate": 3.5222528135552042e-06, + "loss": 0.84998649, + "num_input_tokens_seen": 44036325, + "step": 2062, + "time_per_iteration": 2.6908695697784424 + }, + { + "auxiliary_loss_clip": 0.01217713, + "auxiliary_loss_mlp": 0.01032549, + "balance_loss_clip": 1.06174314, + "balance_loss_mlp": 1.02390003, + "epoch": 0.24806108338844465, + "flos": 18296379521280.0, + "grad_norm": 1.7057297395284086, + "language_loss": 0.80424774, + "learning_rate": 3.521747454313201e-06, + "loss": 0.82675034, + "num_input_tokens_seen": 44055005, + "step": 2063, + "time_per_iteration": 2.717841386795044 + }, + { + "auxiliary_loss_clip": 0.01221344, + "auxiliary_loss_mlp": 0.01038151, + "balance_loss_clip": 1.0499537, + "balance_loss_mlp": 1.02896631, + "epoch": 0.24818132627908374, + "flos": 19282163351040.0, + "grad_norm": 3.9226026747136564, + "language_loss": 0.66656649, + "learning_rate": 3.521241864223319e-06, + "loss": 0.68916142, + "num_input_tokens_seen": 44073965, + "step": 2064, + "time_per_iteration": 2.745756149291992 + }, + { + "auxiliary_loss_clip": 0.01125121, + "auxiliary_loss_mlp": 0.01005854, + "balance_loss_clip": 1.02165794, + "balance_loss_mlp": 1.00377941, + "epoch": 0.24830156916972285, + "flos": 70285837881600.0, + "grad_norm": 0.7888047600010081, + "language_loss": 0.61965692, + "learning_rate": 3.5207360433622552e-06, + "loss": 0.64096665, + "num_input_tokens_seen": 44135965, + "step": 2065, + "time_per_iteration": 3.2293834686279297 + }, + { + "auxiliary_loss_clip": 0.01220574, + "auxiliary_loss_mlp": 0.01033182, + "balance_loss_clip": 1.05883706, + "balance_loss_mlp": 1.02413988, + "epoch": 0.24842181206036193, + "flos": 40409128287360.0, + "grad_norm": 1.6927779934345957, + "language_loss": 0.74607641, + "learning_rate": 3.5202299918067437e-06, + "loss": 0.76861393, + "num_input_tokens_seen": 44159560, + "step": 2066, + "time_per_iteration": 2.8857107162475586 + }, + { + "auxiliary_loss_clip": 0.01217459, + "auxiliary_loss_mlp": 0.01034718, + "balance_loss_clip": 1.05918694, + "balance_loss_mlp": 1.02577734, + "epoch": 0.248542054951001, + "flos": 20082432412800.0, + "grad_norm": 2.3372431267064173, + "language_loss": 0.69162488, + "learning_rate": 3.519723709633551e-06, + "loss": 0.71414667, + "num_input_tokens_seen": 44178320, + "step": 2067, + "time_per_iteration": 2.634671688079834 + }, + { + "auxiliary_loss_clip": 0.01218433, + "auxiliary_loss_mlp": 0.01031124, + "balance_loss_clip": 1.05722928, + "balance_loss_mlp": 1.02214718, + "epoch": 0.24866229784164012, + "flos": 23513948363520.0, + "grad_norm": 2.1715670530012794, + "language_loss": 0.83472335, + "learning_rate": 3.519217196919479e-06, + "loss": 0.85721898, + "num_input_tokens_seen": 44197305, + "step": 2068, + "time_per_iteration": 3.62564754486084 + }, + { + "auxiliary_loss_clip": 0.01228122, + "auxiliary_loss_mlp": 0.01038584, + "balance_loss_clip": 1.05841291, + "balance_loss_mlp": 1.0289104, + "epoch": 0.2487825407322792, + "flos": 19865101173120.0, + "grad_norm": 2.0598769834573427, + "language_loss": 0.73250651, + "learning_rate": 3.518710453741367e-06, + "loss": 0.75517356, + "num_input_tokens_seen": 44216505, + "step": 2069, + "time_per_iteration": 3.6784462928771973 + }, + { + "auxiliary_loss_clip": 0.01214742, + "auxiliary_loss_mlp": 0.01380056, + "balance_loss_clip": 1.05457616, + "balance_loss_mlp": 1.00002146, + "epoch": 0.2489027836229183, + "flos": 22017622573440.0, + "grad_norm": 2.364489353865733, + "language_loss": 0.67932814, + "learning_rate": 3.518203480176086e-06, + "loss": 0.70527613, + "num_input_tokens_seen": 44235435, + "step": 2070, + "time_per_iteration": 3.6321465969085693 + }, + { + "auxiliary_loss_clip": 0.01253221, + "auxiliary_loss_mlp": 0.01035724, + "balance_loss_clip": 1.04659605, + "balance_loss_mlp": 1.02676511, + "epoch": 0.2490230265135574, + "flos": 23294354567040.0, + "grad_norm": 2.229371264789957, + "language_loss": 0.80861366, + "learning_rate": 3.517696276300545e-06, + "loss": 0.83150303, + "num_input_tokens_seen": 44256975, + "step": 2071, + "time_per_iteration": 2.822768449783325 + }, + { + "auxiliary_loss_clip": 0.01220416, + "auxiliary_loss_mlp": 0.01049764, + "balance_loss_clip": 1.06263769, + "balance_loss_mlp": 1.0402329, + "epoch": 0.24914326940419648, + "flos": 19826784339840.0, + "grad_norm": 2.7230443716224344, + "language_loss": 0.68796396, + "learning_rate": 3.517188842191685e-06, + "loss": 0.71066576, + "num_input_tokens_seen": 44275125, + "step": 2072, + "time_per_iteration": 2.6622917652130127 + }, + { + "auxiliary_loss_clip": 0.01213655, + "auxiliary_loss_mlp": 0.01031373, + "balance_loss_clip": 1.05752826, + "balance_loss_mlp": 1.02190232, + "epoch": 0.24926351229483557, + "flos": 20229271211520.0, + "grad_norm": 1.8315238934436402, + "language_loss": 0.73818892, + "learning_rate": 3.5166811779264837e-06, + "loss": 0.76063919, + "num_input_tokens_seen": 44295445, + "step": 2073, + "time_per_iteration": 2.6737101078033447 + }, + { + "auxiliary_loss_clip": 0.01208305, + "auxiliary_loss_mlp": 0.0103302, + "balance_loss_clip": 1.05991793, + "balance_loss_mlp": 1.02421081, + "epoch": 0.24938375518547465, + "flos": 23294570048640.0, + "grad_norm": 1.9670142333872231, + "language_loss": 0.77948451, + "learning_rate": 3.5161732835819545e-06, + "loss": 0.80189776, + "num_input_tokens_seen": 44314755, + "step": 2074, + "time_per_iteration": 3.5835156440734863 + }, + { + "auxiliary_loss_clip": 0.01207488, + "auxiliary_loss_mlp": 0.01033764, + "balance_loss_clip": 1.06022596, + "balance_loss_mlp": 1.02500188, + "epoch": 0.24950399807611376, + "flos": 17311673099520.0, + "grad_norm": 1.7648449369674784, + "language_loss": 0.83231044, + "learning_rate": 3.515665159235143e-06, + "loss": 0.85472298, + "num_input_tokens_seen": 44333640, + "step": 2075, + "time_per_iteration": 2.5851874351501465 + }, + { + "auxiliary_loss_clip": 0.0122034, + "auxiliary_loss_mlp": 0.01030392, + "balance_loss_clip": 1.05442023, + "balance_loss_mlp": 1.02181494, + "epoch": 0.24962424096675284, + "flos": 19024863252480.0, + "grad_norm": 1.5184575385305128, + "language_loss": 0.75047755, + "learning_rate": 3.5151568049631318e-06, + "loss": 0.77298486, + "num_input_tokens_seen": 44352355, + "step": 2076, + "time_per_iteration": 2.6964640617370605 + }, + { + "auxiliary_loss_clip": 0.0121097, + "auxiliary_loss_mlp": 0.01029829, + "balance_loss_clip": 1.06164312, + "balance_loss_mlp": 1.02051306, + "epoch": 0.24974448385739192, + "flos": 33398790710400.0, + "grad_norm": 1.7392806422407396, + "language_loss": 0.80573785, + "learning_rate": 3.5146482208430385e-06, + "loss": 0.82814586, + "num_input_tokens_seen": 44374185, + "step": 2077, + "time_per_iteration": 2.736570358276367 + }, + { + "auxiliary_loss_clip": 0.01233889, + "auxiliary_loss_mlp": 0.01039802, + "balance_loss_clip": 1.04779637, + "balance_loss_mlp": 1.02943683, + "epoch": 0.24986472674803104, + "flos": 30007279532160.0, + "grad_norm": 2.470464380283846, + "language_loss": 0.67676413, + "learning_rate": 3.514139406952014e-06, + "loss": 0.69950104, + "num_input_tokens_seen": 44396210, + "step": 2078, + "time_per_iteration": 2.810161590576172 + }, + { + "auxiliary_loss_clip": 0.01217235, + "auxiliary_loss_mlp": 0.01037921, + "balance_loss_clip": 1.05978, + "balance_loss_mlp": 1.02920675, + "epoch": 0.24998496963867012, + "flos": 26613074833920.0, + "grad_norm": 2.0214597872551576, + "language_loss": 0.83327508, + "learning_rate": 3.5136303633672454e-06, + "loss": 0.85582662, + "num_input_tokens_seen": 44416340, + "step": 2079, + "time_per_iteration": 2.681514024734497 + }, + { + "auxiliary_loss_clip": 0.01243268, + "auxiliary_loss_mlp": 0.01379794, + "balance_loss_clip": 1.05772448, + "balance_loss_mlp": 1.00010872, + "epoch": 0.25010521252930923, + "flos": 23553989049600.0, + "grad_norm": 1.7559197650254197, + "language_loss": 0.74613923, + "learning_rate": 3.5131210901659544e-06, + "loss": 0.77236986, + "num_input_tokens_seen": 44438095, + "step": 2080, + "time_per_iteration": 2.741938352584839 + }, + { + "auxiliary_loss_clip": 0.01227729, + "auxiliary_loss_mlp": 0.01032596, + "balance_loss_clip": 1.05410075, + "balance_loss_mlp": 1.0239116, + "epoch": 0.2502254554199483, + "flos": 23441193365760.0, + "grad_norm": 2.48608459474283, + "language_loss": 0.81882632, + "learning_rate": 3.5126115874253967e-06, + "loss": 0.84142959, + "num_input_tokens_seen": 44457650, + "step": 2081, + "time_per_iteration": 2.8057162761688232 + }, + { + "auxiliary_loss_clip": 0.01236003, + "auxiliary_loss_mlp": 0.01029844, + "balance_loss_clip": 1.05716193, + "balance_loss_mlp": 1.02020001, + "epoch": 0.2503456983105874, + "flos": 28761681651840.0, + "grad_norm": 2.1495778581183322, + "language_loss": 0.80722177, + "learning_rate": 3.5121018552228644e-06, + "loss": 0.82988024, + "num_input_tokens_seen": 44476155, + "step": 2082, + "time_per_iteration": 2.740056037902832 + }, + { + "auxiliary_loss_clip": 0.01234569, + "auxiliary_loss_mlp": 0.01036796, + "balance_loss_clip": 1.05784702, + "balance_loss_mlp": 1.02782536, + "epoch": 0.2504659412012265, + "flos": 18770256673920.0, + "grad_norm": 2.5232629337923353, + "language_loss": 0.76118112, + "learning_rate": 3.5115918936356827e-06, + "loss": 0.78389472, + "num_input_tokens_seen": 44492910, + "step": 2083, + "time_per_iteration": 2.855879545211792 + }, + { + "auxiliary_loss_clip": 0.01207565, + "auxiliary_loss_mlp": 0.01037492, + "balance_loss_clip": 1.05066431, + "balance_loss_mlp": 1.02820587, + "epoch": 0.25058618409186556, + "flos": 16873383346560.0, + "grad_norm": 1.8660111825496413, + "language_loss": 0.78465223, + "learning_rate": 3.5110817027412123e-06, + "loss": 0.8071028, + "num_input_tokens_seen": 44512000, + "step": 2084, + "time_per_iteration": 2.7938971519470215 + }, + { + "auxiliary_loss_clip": 0.01226071, + "auxiliary_loss_mlp": 0.01034324, + "balance_loss_clip": 1.05309761, + "balance_loss_mlp": 1.02500212, + "epoch": 0.25070642698250467, + "flos": 24425540651520.0, + "grad_norm": 1.915342264529471, + "language_loss": 0.68506426, + "learning_rate": 3.5105712826168493e-06, + "loss": 0.70766819, + "num_input_tokens_seen": 44531650, + "step": 2085, + "time_per_iteration": 2.7941017150878906 + }, + { + "auxiliary_loss_clip": 0.01215073, + "auxiliary_loss_mlp": 0.01379712, + "balance_loss_clip": 1.05858159, + "balance_loss_mlp": 1.00003386, + "epoch": 0.2508266698731437, + "flos": 20260944028800.0, + "grad_norm": 1.9602359210215499, + "language_loss": 0.70658189, + "learning_rate": 3.5100606333400235e-06, + "loss": 0.7325297, + "num_input_tokens_seen": 44548785, + "step": 2086, + "time_per_iteration": 2.7271203994750977 + }, + { + "auxiliary_loss_clip": 0.01235457, + "auxiliary_loss_mlp": 0.01033891, + "balance_loss_clip": 1.05766237, + "balance_loss_mlp": 1.02449167, + "epoch": 0.25094691276378284, + "flos": 19245318975360.0, + "grad_norm": 2.8836181823197147, + "language_loss": 0.77509308, + "learning_rate": 3.5095497549882006e-06, + "loss": 0.79778659, + "num_input_tokens_seen": 44567230, + "step": 2087, + "time_per_iteration": 2.736361265182495 + }, + { + "auxiliary_loss_clip": 0.01220028, + "auxiliary_loss_mlp": 0.01038664, + "balance_loss_clip": 1.06118858, + "balance_loss_mlp": 1.02966404, + "epoch": 0.25106715565442195, + "flos": 26943237671040.0, + "grad_norm": 2.1334266070353523, + "language_loss": 0.71949273, + "learning_rate": 3.50903864763888e-06, + "loss": 0.74207968, + "num_input_tokens_seen": 44588020, + "step": 2088, + "time_per_iteration": 2.7247262001037598 + }, + { + "auxiliary_loss_clip": 0.01223302, + "auxiliary_loss_mlp": 0.01031387, + "balance_loss_clip": 1.06078577, + "balance_loss_mlp": 1.02180886, + "epoch": 0.251187398545061, + "flos": 48359570572800.0, + "grad_norm": 2.0685188628059925, + "language_loss": 0.75891906, + "learning_rate": 3.5085273113695965e-06, + "loss": 0.78146589, + "num_input_tokens_seen": 44612590, + "step": 2089, + "time_per_iteration": 2.8553106784820557 + }, + { + "auxiliary_loss_clip": 0.01210456, + "auxiliary_loss_mlp": 0.01034499, + "balance_loss_clip": 1.06161308, + "balance_loss_mlp": 1.02511764, + "epoch": 0.2513076414357001, + "flos": 27016100409600.0, + "grad_norm": 1.7806143073837712, + "language_loss": 0.78674042, + "learning_rate": 3.508015746257919e-06, + "loss": 0.80919003, + "num_input_tokens_seen": 44631630, + "step": 2090, + "time_per_iteration": 2.6844940185546875 + }, + { + "auxiliary_loss_clip": 0.01238792, + "auxiliary_loss_mlp": 0.0103524, + "balance_loss_clip": 1.05692005, + "balance_loss_mlp": 1.02595997, + "epoch": 0.2514278843263392, + "flos": 19463619882240.0, + "grad_norm": 2.1036162392830375, + "language_loss": 0.83427, + "learning_rate": 3.5075039523814518e-06, + "loss": 0.85701025, + "num_input_tokens_seen": 44650820, + "step": 2091, + "time_per_iteration": 2.6807057857513428 + }, + { + "auxiliary_loss_clip": 0.01224537, + "auxiliary_loss_mlp": 0.0103418, + "balance_loss_clip": 1.05950904, + "balance_loss_mlp": 1.02448809, + "epoch": 0.2515481272169783, + "flos": 16866092885760.0, + "grad_norm": 2.144878397307424, + "language_loss": 0.81909293, + "learning_rate": 3.506991929817834e-06, + "loss": 0.84168005, + "num_input_tokens_seen": 44667540, + "step": 2092, + "time_per_iteration": 2.650557279586792 + }, + { + "auxiliary_loss_clip": 0.01208765, + "auxiliary_loss_mlp": 0.0102497, + "balance_loss_clip": 1.06438863, + "balance_loss_mlp": 1.0163393, + "epoch": 0.2516683701076174, + "flos": 23732464752000.0, + "grad_norm": 2.6616176606872903, + "language_loss": 0.82509005, + "learning_rate": 3.506479678644738e-06, + "loss": 0.84742737, + "num_input_tokens_seen": 44687935, + "step": 2093, + "time_per_iteration": 2.6269052028656006 + }, + { + "auxiliary_loss_clip": 0.01231535, + "auxiliary_loss_mlp": 0.01034234, + "balance_loss_clip": 1.05395508, + "balance_loss_mlp": 1.02503657, + "epoch": 0.2517886129982565, + "flos": 27635954434560.0, + "grad_norm": 114.89622509254856, + "language_loss": 0.73651421, + "learning_rate": 3.505967198939873e-06, + "loss": 0.75917184, + "num_input_tokens_seen": 44704975, + "step": 2094, + "time_per_iteration": 3.6834490299224854 + }, + { + "auxiliary_loss_clip": 0.01221061, + "auxiliary_loss_mlp": 0.01032691, + "balance_loss_clip": 1.0533576, + "balance_loss_mlp": 1.02413166, + "epoch": 0.25190885588889556, + "flos": 38104596529920.0, + "grad_norm": 1.9947318985298663, + "language_loss": 0.78201473, + "learning_rate": 3.5054544907809813e-06, + "loss": 0.80455232, + "num_input_tokens_seen": 44725475, + "step": 2095, + "time_per_iteration": 3.831298828125 + }, + { + "auxiliary_loss_clip": 0.01223513, + "auxiliary_loss_mlp": 0.01380035, + "balance_loss_clip": 1.05863488, + "balance_loss_mlp": 1.00007319, + "epoch": 0.25202909877953467, + "flos": 22269894768000.0, + "grad_norm": 2.410135688629215, + "language_loss": 0.8083927, + "learning_rate": 3.50494155424584e-06, + "loss": 0.83442819, + "num_input_tokens_seen": 44744380, + "step": 2096, + "time_per_iteration": 3.700204849243164 + }, + { + "auxiliary_loss_clip": 0.01220419, + "auxiliary_loss_mlp": 0.01030075, + "balance_loss_clip": 1.0599792, + "balance_loss_mlp": 1.02134323, + "epoch": 0.2521493416701738, + "flos": 21761759018880.0, + "grad_norm": 1.7944974620139096, + "language_loss": 0.83422399, + "learning_rate": 3.504428389412262e-06, + "loss": 0.85672891, + "num_input_tokens_seen": 44765190, + "step": 2097, + "time_per_iteration": 2.727177143096924 + }, + { + "auxiliary_loss_clip": 0.01213768, + "auxiliary_loss_mlp": 0.01028821, + "balance_loss_clip": 1.05768108, + "balance_loss_mlp": 1.0193615, + "epoch": 0.25226958456081283, + "flos": 27746738956800.0, + "grad_norm": 4.7033508974582805, + "language_loss": 0.73435926, + "learning_rate": 3.5039149963580927e-06, + "loss": 0.75678515, + "num_input_tokens_seen": 44785210, + "step": 2098, + "time_per_iteration": 2.8204612731933594 + }, + { + "auxiliary_loss_clip": 0.01221191, + "auxiliary_loss_mlp": 0.01031567, + "balance_loss_clip": 1.05927122, + "balance_loss_mlp": 1.02257288, + "epoch": 0.25238982745145194, + "flos": 30732171903360.0, + "grad_norm": 3.079995695486305, + "language_loss": 0.70115268, + "learning_rate": 3.503401375161215e-06, + "loss": 0.72368026, + "num_input_tokens_seen": 44804955, + "step": 2099, + "time_per_iteration": 2.7657387256622314 + }, + { + "auxiliary_loss_clip": 0.01205731, + "auxiliary_loss_mlp": 0.01029951, + "balance_loss_clip": 1.05919552, + "balance_loss_mlp": 1.02071881, + "epoch": 0.252510070342091, + "flos": 20266331068800.0, + "grad_norm": 3.388325926503128, + "language_loss": 0.83663917, + "learning_rate": 3.502887525899544e-06, + "loss": 0.85899603, + "num_input_tokens_seen": 44823935, + "step": 2100, + "time_per_iteration": 3.5313668251037598 + }, + { + "auxiliary_loss_clip": 0.01224801, + "auxiliary_loss_mlp": 0.01034936, + "balance_loss_clip": 1.0566901, + "balance_loss_mlp": 1.02553654, + "epoch": 0.2526303132327301, + "flos": 22747399194240.0, + "grad_norm": 2.1197750163295477, + "language_loss": 0.82893789, + "learning_rate": 3.50237344865103e-06, + "loss": 0.85153526, + "num_input_tokens_seen": 44844935, + "step": 2101, + "time_per_iteration": 2.7069246768951416 + }, + { + "auxiliary_loss_clip": 0.0120978, + "auxiliary_loss_mlp": 0.01032638, + "balance_loss_clip": 1.06180167, + "balance_loss_mlp": 1.02400124, + "epoch": 0.2527505561233692, + "flos": 30263466309120.0, + "grad_norm": 2.5647290340647184, + "language_loss": 0.7608279, + "learning_rate": 3.501859143493658e-06, + "loss": 0.78325212, + "num_input_tokens_seen": 44865565, + "step": 2102, + "time_per_iteration": 2.680898666381836 + }, + { + "auxiliary_loss_clip": 0.01107054, + "auxiliary_loss_mlp": 0.01000394, + "balance_loss_clip": 1.02854872, + "balance_loss_mlp": 0.99842721, + "epoch": 0.2528707990140083, + "flos": 58492917164160.0, + "grad_norm": 0.933474489071914, + "language_loss": 0.60532445, + "learning_rate": 3.5013446105054488e-06, + "loss": 0.62639892, + "num_input_tokens_seen": 44918485, + "step": 2103, + "time_per_iteration": 2.924302577972412 + }, + { + "auxiliary_loss_clip": 0.01216513, + "auxiliary_loss_mlp": 0.01036666, + "balance_loss_clip": 1.05246449, + "balance_loss_mlp": 1.02761173, + "epoch": 0.2529910419046474, + "flos": 24645134448000.0, + "grad_norm": 1.7498182130074202, + "language_loss": 0.75139898, + "learning_rate": 3.5008298497644555e-06, + "loss": 0.77393079, + "num_input_tokens_seen": 44937530, + "step": 2104, + "time_per_iteration": 2.713329315185547 + }, + { + "auxiliary_loss_clip": 0.0123791, + "auxiliary_loss_mlp": 0.01032872, + "balance_loss_clip": 1.0577507, + "balance_loss_mlp": 1.0235076, + "epoch": 0.2531112847952865, + "flos": 23842135952640.0, + "grad_norm": 1.9115499471751514, + "language_loss": 0.87851214, + "learning_rate": 3.500314861348767e-06, + "loss": 0.9012199, + "num_input_tokens_seen": 44958165, + "step": 2105, + "time_per_iteration": 2.7490594387054443 + }, + { + "auxiliary_loss_clip": 0.01225381, + "auxiliary_loss_mlp": 0.01036376, + "balance_loss_clip": 1.05747163, + "balance_loss_mlp": 1.02652907, + "epoch": 0.25323152768592555, + "flos": 16143822207360.0, + "grad_norm": 1.8308964842969995, + "language_loss": 0.77109367, + "learning_rate": 3.499799645336507e-06, + "loss": 0.79371119, + "num_input_tokens_seen": 44975060, + "step": 2106, + "time_per_iteration": 2.8674490451812744 + }, + { + "auxiliary_loss_clip": 0.01220329, + "auxiliary_loss_mlp": 0.01028459, + "balance_loss_clip": 1.06219506, + "balance_loss_mlp": 1.02040088, + "epoch": 0.25335177057656466, + "flos": 28405161210240.0, + "grad_norm": 1.8499969309369615, + "language_loss": 0.86861753, + "learning_rate": 3.4992842018058336e-06, + "loss": 0.89110541, + "num_input_tokens_seen": 44997960, + "step": 2107, + "time_per_iteration": 2.716810464859009 + }, + { + "auxiliary_loss_clip": 0.01241413, + "auxiliary_loss_mlp": 0.01030797, + "balance_loss_clip": 1.05664182, + "balance_loss_mlp": 1.02142119, + "epoch": 0.25347201346720377, + "flos": 18799666934400.0, + "grad_norm": 2.7610290948353597, + "language_loss": 0.88217378, + "learning_rate": 3.4987685308349384e-06, + "loss": 0.9048959, + "num_input_tokens_seen": 45015690, + "step": 2108, + "time_per_iteration": 2.648028612136841 + }, + { + "auxiliary_loss_clip": 0.01227929, + "auxiliary_loss_mlp": 0.01032424, + "balance_loss_clip": 1.05044746, + "balance_loss_mlp": 1.02344131, + "epoch": 0.2535922563578428, + "flos": 15815490963840.0, + "grad_norm": 2.211712648109795, + "language_loss": 0.61665857, + "learning_rate": 3.4982526325020497e-06, + "loss": 0.63926208, + "num_input_tokens_seen": 45032660, + "step": 2109, + "time_per_iteration": 2.7571210861206055 + }, + { + "auxiliary_loss_clip": 0.01226524, + "auxiliary_loss_mlp": 0.01034545, + "balance_loss_clip": 1.05646062, + "balance_loss_mlp": 1.02519321, + "epoch": 0.25371249924848194, + "flos": 16318922031360.0, + "grad_norm": 2.118249265403673, + "language_loss": 0.81725073, + "learning_rate": 3.4977365068854273e-06, + "loss": 0.83986139, + "num_input_tokens_seen": 45048280, + "step": 2110, + "time_per_iteration": 2.612539529800415 + }, + { + "auxiliary_loss_clip": 0.01214988, + "auxiliary_loss_mlp": 0.01030595, + "balance_loss_clip": 1.05496693, + "balance_loss_mlp": 1.02124858, + "epoch": 0.25383274213912105, + "flos": 21761615364480.0, + "grad_norm": 1.784678956235311, + "language_loss": 0.7356075, + "learning_rate": 3.4972201540633676e-06, + "loss": 0.75806338, + "num_input_tokens_seen": 45067635, + "step": 2111, + "time_per_iteration": 2.731476306915283 + }, + { + "auxiliary_loss_clip": 0.01216818, + "auxiliary_loss_mlp": 0.01033315, + "balance_loss_clip": 1.05631125, + "balance_loss_mlp": 1.02442765, + "epoch": 0.2539529850297601, + "flos": 21396870708480.0, + "grad_norm": 1.8555367411313184, + "language_loss": 0.85012782, + "learning_rate": 3.4967035741142008e-06, + "loss": 0.87262917, + "num_input_tokens_seen": 45086455, + "step": 2112, + "time_per_iteration": 2.874671459197998 + }, + { + "auxiliary_loss_clip": 0.01217284, + "auxiliary_loss_mlp": 0.01034287, + "balance_loss_clip": 1.06089306, + "balance_loss_mlp": 1.02584112, + "epoch": 0.2540732279203992, + "flos": 25228467319680.0, + "grad_norm": 1.8438296706737982, + "language_loss": 0.82112944, + "learning_rate": 3.4961867671162917e-06, + "loss": 0.8436451, + "num_input_tokens_seen": 45106385, + "step": 2113, + "time_per_iteration": 2.673252820968628 + }, + { + "auxiliary_loss_clip": 0.01209725, + "auxiliary_loss_mlp": 0.01036758, + "balance_loss_clip": 1.05956268, + "balance_loss_mlp": 1.02741241, + "epoch": 0.2541934708110383, + "flos": 19427386037760.0, + "grad_norm": 2.8518102257277684, + "language_loss": 0.77218843, + "learning_rate": 3.4956697331480402e-06, + "loss": 0.79465324, + "num_input_tokens_seen": 45124955, + "step": 2114, + "time_per_iteration": 2.6098625659942627 + }, + { + "auxiliary_loss_clip": 0.01236301, + "auxiliary_loss_mlp": 0.01034088, + "balance_loss_clip": 1.05417717, + "balance_loss_mlp": 1.02444994, + "epoch": 0.2543137137016774, + "flos": 23949436855680.0, + "grad_norm": 1.5315839582750712, + "language_loss": 0.79944849, + "learning_rate": 3.495152472287879e-06, + "loss": 0.82215232, + "num_input_tokens_seen": 45145665, + "step": 2115, + "time_per_iteration": 2.804514169692993 + }, + { + "auxiliary_loss_clip": 0.0123808, + "auxiliary_loss_mlp": 0.01036418, + "balance_loss_clip": 1.05869269, + "balance_loss_mlp": 1.02735233, + "epoch": 0.2544339565923165, + "flos": 25593283802880.0, + "grad_norm": 1.543667602558463, + "language_loss": 0.74030709, + "learning_rate": 3.4946349846142766e-06, + "loss": 0.76305217, + "num_input_tokens_seen": 45164805, + "step": 2116, + "time_per_iteration": 2.806182384490967 + }, + { + "auxiliary_loss_clip": 0.01207448, + "auxiliary_loss_mlp": 0.0103061, + "balance_loss_clip": 1.0602262, + "balance_loss_mlp": 1.02150846, + "epoch": 0.25455419948295555, + "flos": 21689470897920.0, + "grad_norm": 3.192980015718693, + "language_loss": 0.75612187, + "learning_rate": 3.4941172702057353e-06, + "loss": 0.77850246, + "num_input_tokens_seen": 45184865, + "step": 2117, + "time_per_iteration": 2.784372568130493 + }, + { + "auxiliary_loss_clip": 0.01228397, + "auxiliary_loss_mlp": 0.01030518, + "balance_loss_clip": 1.06001627, + "balance_loss_mlp": 1.02185726, + "epoch": 0.25467444237359466, + "flos": 26250341339520.0, + "grad_norm": 1.823697049294683, + "language_loss": 0.80296338, + "learning_rate": 3.4935993291407924e-06, + "loss": 0.82555258, + "num_input_tokens_seen": 45203690, + "step": 2118, + "time_per_iteration": 2.7931201457977295 + }, + { + "auxiliary_loss_clip": 0.01222622, + "auxiliary_loss_mlp": 0.01037413, + "balance_loss_clip": 1.05819547, + "balance_loss_mlp": 1.02846014, + "epoch": 0.25479468526423377, + "flos": 26979686997120.0, + "grad_norm": 2.3098919261683464, + "language_loss": 0.71061528, + "learning_rate": 3.4930811614980183e-06, + "loss": 0.73321563, + "num_input_tokens_seen": 45225385, + "step": 2119, + "time_per_iteration": 2.783573865890503 + }, + { + "auxiliary_loss_clip": 0.01212104, + "auxiliary_loss_mlp": 0.01034823, + "balance_loss_clip": 1.0591085, + "balance_loss_mlp": 1.02581692, + "epoch": 0.2549149281548728, + "flos": 23475811098240.0, + "grad_norm": 1.6374231468039468, + "language_loss": 0.79476827, + "learning_rate": 3.4925627673560198e-06, + "loss": 0.81723762, + "num_input_tokens_seen": 45246045, + "step": 2120, + "time_per_iteration": 3.578296422958374 + }, + { + "auxiliary_loss_clip": 0.01233405, + "auxiliary_loss_mlp": 0.0104319, + "balance_loss_clip": 1.05631137, + "balance_loss_mlp": 1.03363514, + "epoch": 0.25503517104551193, + "flos": 25812302981760.0, + "grad_norm": 1.871256454013379, + "language_loss": 0.88211864, + "learning_rate": 3.4920441467934357e-06, + "loss": 0.90488464, + "num_input_tokens_seen": 45266560, + "step": 2121, + "time_per_iteration": 3.689966917037964 + }, + { + "auxiliary_loss_clip": 0.01222869, + "auxiliary_loss_mlp": 0.01032, + "balance_loss_clip": 1.05446374, + "balance_loss_mlp": 1.02318478, + "epoch": 0.25515541393615104, + "flos": 26645106787200.0, + "grad_norm": 2.1486186407143895, + "language_loss": 0.82530546, + "learning_rate": 3.491525299888941e-06, + "loss": 0.84785414, + "num_input_tokens_seen": 45285405, + "step": 2122, + "time_per_iteration": 3.7079710960388184 + }, + { + "auxiliary_loss_clip": 0.01119278, + "auxiliary_loss_mlp": 0.01375479, + "balance_loss_clip": 1.0271827, + "balance_loss_mlp": 0.99963123, + "epoch": 0.2552756568267901, + "flos": 65955945847680.0, + "grad_norm": 0.8808916388320076, + "language_loss": 0.62709188, + "learning_rate": 3.491006226721244e-06, + "loss": 0.65203947, + "num_input_tokens_seen": 45349615, + "step": 2123, + "time_per_iteration": 3.2771036624908447 + }, + { + "auxiliary_loss_clip": 0.01228692, + "auxiliary_loss_mlp": 0.0137991, + "balance_loss_clip": 1.06030893, + "balance_loss_mlp": 1.00008237, + "epoch": 0.2553958997174292, + "flos": 17931096161280.0, + "grad_norm": 3.1444915060547594, + "language_loss": 0.78076643, + "learning_rate": 3.4904869273690882e-06, + "loss": 0.80685246, + "num_input_tokens_seen": 45367505, + "step": 2124, + "time_per_iteration": 2.7324275970458984 + }, + { + "auxiliary_loss_clip": 0.01219941, + "auxiliary_loss_mlp": 0.01036658, + "balance_loss_clip": 1.05872488, + "balance_loss_mlp": 1.02759826, + "epoch": 0.2555161426080683, + "flos": 23367791923200.0, + "grad_norm": 2.402258955213244, + "language_loss": 0.88924193, + "learning_rate": 3.489967401911251e-06, + "loss": 0.91180789, + "num_input_tokens_seen": 45386805, + "step": 2125, + "time_per_iteration": 2.680601119995117 + }, + { + "auxiliary_loss_clip": 0.0121205, + "auxiliary_loss_mlp": 0.01044231, + "balance_loss_clip": 1.06138468, + "balance_loss_mlp": 1.03451502, + "epoch": 0.2556363854987074, + "flos": 40625130723840.0, + "grad_norm": 1.740031615147917, + "language_loss": 0.69532555, + "learning_rate": 3.4894476504265428e-06, + "loss": 0.71788836, + "num_input_tokens_seen": 45411045, + "step": 2126, + "time_per_iteration": 3.7130954265594482 + }, + { + "auxiliary_loss_clip": 0.01112838, + "auxiliary_loss_mlp": 0.01001962, + "balance_loss_clip": 1.02489567, + "balance_loss_mlp": 0.99989951, + "epoch": 0.2557566283893465, + "flos": 68019443389440.0, + "grad_norm": 0.8092878302569183, + "language_loss": 0.54483747, + "learning_rate": 3.4889276729938104e-06, + "loss": 0.5659855, + "num_input_tokens_seen": 45469575, + "step": 2127, + "time_per_iteration": 3.1482627391815186 + }, + { + "auxiliary_loss_clip": 0.01221912, + "auxiliary_loss_mlp": 0.01034205, + "balance_loss_clip": 1.05808854, + "balance_loss_mlp": 1.02485263, + "epoch": 0.2558768712799856, + "flos": 22635645004800.0, + "grad_norm": 2.0470469400447815, + "language_loss": 0.80747694, + "learning_rate": 3.488407469691934e-06, + "loss": 0.83003807, + "num_input_tokens_seen": 45490270, + "step": 2128, + "time_per_iteration": 2.688539743423462 + }, + { + "auxiliary_loss_clip": 0.01221187, + "auxiliary_loss_mlp": 0.01033424, + "balance_loss_clip": 1.05544889, + "balance_loss_mlp": 1.02391148, + "epoch": 0.25599711417062465, + "flos": 26396354125440.0, + "grad_norm": 1.9418583047561533, + "language_loss": 0.80594158, + "learning_rate": 3.487887040599828e-06, + "loss": 0.82848763, + "num_input_tokens_seen": 45510070, + "step": 2129, + "time_per_iteration": 2.7621874809265137 + }, + { + "auxiliary_loss_clip": 0.01211709, + "auxiliary_loss_mlp": 0.01043199, + "balance_loss_clip": 1.06279731, + "balance_loss_mlp": 1.03391242, + "epoch": 0.25611735706126376, + "flos": 22852042490880.0, + "grad_norm": 2.2936117753457066, + "language_loss": 0.76699805, + "learning_rate": 3.4873663857964407e-06, + "loss": 0.78954709, + "num_input_tokens_seen": 45527285, + "step": 2130, + "time_per_iteration": 2.592083215713501 + }, + { + "auxiliary_loss_clip": 0.01244794, + "auxiliary_loss_mlp": 0.01040488, + "balance_loss_clip": 1.05807328, + "balance_loss_mlp": 1.03114247, + "epoch": 0.2562375999519028, + "flos": 23367863750400.0, + "grad_norm": 1.7341865038768025, + "language_loss": 0.66456866, + "learning_rate": 3.4868455053607556e-06, + "loss": 0.68742144, + "num_input_tokens_seen": 45546900, + "step": 2131, + "time_per_iteration": 2.7649693489074707 + }, + { + "auxiliary_loss_clip": 0.01223006, + "auxiliary_loss_mlp": 0.01035302, + "balance_loss_clip": 1.06043577, + "balance_loss_mlp": 1.02599788, + "epoch": 0.2563578428425419, + "flos": 22856962654080.0, + "grad_norm": 1.8480817165581211, + "language_loss": 0.71173704, + "learning_rate": 3.486324399371789e-06, + "loss": 0.73432016, + "num_input_tokens_seen": 45566200, + "step": 2132, + "time_per_iteration": 2.6388165950775146 + }, + { + "auxiliary_loss_clip": 0.01227831, + "auxiliary_loss_mlp": 0.01041688, + "balance_loss_clip": 1.05629659, + "balance_loss_mlp": 1.03324223, + "epoch": 0.25647808573318104, + "flos": 21653883498240.0, + "grad_norm": 1.9522416182345808, + "language_loss": 0.7835381, + "learning_rate": 3.485803067908593e-06, + "loss": 0.80623329, + "num_input_tokens_seen": 45585710, + "step": 2133, + "time_per_iteration": 2.7399585247039795 + }, + { + "auxiliary_loss_clip": 0.01219587, + "auxiliary_loss_mlp": 0.01037125, + "balance_loss_clip": 1.04526687, + "balance_loss_mlp": 1.0277909, + "epoch": 0.2565983286238201, + "flos": 33730569659520.0, + "grad_norm": 1.9270040038557141, + "language_loss": 0.79641438, + "learning_rate": 3.485281511050253e-06, + "loss": 0.81898153, + "num_input_tokens_seen": 45607845, + "step": 2134, + "time_per_iteration": 2.89172625541687 + }, + { + "auxiliary_loss_clip": 0.0122013, + "auxiliary_loss_mlp": 0.01034994, + "balance_loss_clip": 1.05841517, + "balance_loss_mlp": 1.0259583, + "epoch": 0.2567185715144592, + "flos": 16216002587520.0, + "grad_norm": 2.930242363723049, + "language_loss": 0.89752817, + "learning_rate": 3.484759728875889e-06, + "loss": 0.92007947, + "num_input_tokens_seen": 45623210, + "step": 2135, + "time_per_iteration": 2.607203483581543 + }, + { + "auxiliary_loss_clip": 0.01227418, + "auxiliary_loss_mlp": 0.01036468, + "balance_loss_clip": 1.0522505, + "balance_loss_mlp": 1.02771771, + "epoch": 0.2568388144050983, + "flos": 17458475984640.0, + "grad_norm": 1.8028112973528916, + "language_loss": 0.80703604, + "learning_rate": 3.4842377214646543e-06, + "loss": 0.82967496, + "num_input_tokens_seen": 45641505, + "step": 2136, + "time_per_iteration": 2.726004123687744 + }, + { + "auxiliary_loss_clip": 0.01207086, + "auxiliary_loss_mlp": 0.01031052, + "balance_loss_clip": 1.06074202, + "balance_loss_mlp": 1.02248085, + "epoch": 0.25695905729573737, + "flos": 20887442069760.0, + "grad_norm": 1.6863281047630239, + "language_loss": 0.66934741, + "learning_rate": 3.483715488895737e-06, + "loss": 0.69172883, + "num_input_tokens_seen": 45661835, + "step": 2137, + "time_per_iteration": 2.5689237117767334 + }, + { + "auxiliary_loss_clip": 0.01240311, + "auxiliary_loss_mlp": 0.01037754, + "balance_loss_clip": 1.05197382, + "balance_loss_mlp": 1.02878916, + "epoch": 0.2570793001863765, + "flos": 24717278914560.0, + "grad_norm": 1.8641157754687216, + "language_loss": 0.77892792, + "learning_rate": 3.48319303124836e-06, + "loss": 0.80170864, + "num_input_tokens_seen": 45682215, + "step": 2138, + "time_per_iteration": 2.8665881156921387 + }, + { + "auxiliary_loss_clip": 0.01221644, + "auxiliary_loss_mlp": 0.01033221, + "balance_loss_clip": 1.05923033, + "balance_loss_mlp": 1.02429843, + "epoch": 0.2571995430770156, + "flos": 26906896085760.0, + "grad_norm": 2.1351567957111603, + "language_loss": 0.67117822, + "learning_rate": 3.4826703486017798e-06, + "loss": 0.69372684, + "num_input_tokens_seen": 45701840, + "step": 2139, + "time_per_iteration": 2.7191741466522217 + }, + { + "auxiliary_loss_clip": 0.01214284, + "auxiliary_loss_mlp": 0.0103468, + "balance_loss_clip": 1.06078792, + "balance_loss_mlp": 1.02564371, + "epoch": 0.25731978596765465, + "flos": 19792561656960.0, + "grad_norm": 1.9564542335770276, + "language_loss": 0.76971006, + "learning_rate": 3.4821474410352867e-06, + "loss": 0.79219973, + "num_input_tokens_seen": 45720500, + "step": 2140, + "time_per_iteration": 2.719269275665283 + }, + { + "auxiliary_loss_clip": 0.01143657, + "auxiliary_loss_mlp": 0.01000314, + "balance_loss_clip": 1.03176236, + "balance_loss_mlp": 0.99812108, + "epoch": 0.25744002885829376, + "flos": 70564970471040.0, + "grad_norm": 0.9074357102165657, + "language_loss": 0.62675869, + "learning_rate": 3.481624308628205e-06, + "loss": 0.64819843, + "num_input_tokens_seen": 45781870, + "step": 2141, + "time_per_iteration": 3.3988194465637207 + }, + { + "auxiliary_loss_clip": 0.01225219, + "auxiliary_loss_mlp": 0.01031601, + "balance_loss_clip": 1.05910492, + "balance_loss_mlp": 1.02217746, + "epoch": 0.25756027174893287, + "flos": 18038181582720.0, + "grad_norm": 3.4434310164846167, + "language_loss": 1.00298738, + "learning_rate": 3.481100951459893e-06, + "loss": 1.02555561, + "num_input_tokens_seen": 45794890, + "step": 2142, + "time_per_iteration": 2.6483821868896484 + }, + { + "auxiliary_loss_clip": 0.01212493, + "auxiliary_loss_mlp": 0.01031252, + "balance_loss_clip": 1.05797815, + "balance_loss_mlp": 1.02225733, + "epoch": 0.2576805146395719, + "flos": 22674069578880.0, + "grad_norm": 1.6274432379922987, + "language_loss": 0.78719813, + "learning_rate": 3.4805773696097453e-06, + "loss": 0.80963558, + "num_input_tokens_seen": 45815780, + "step": 2143, + "time_per_iteration": 2.7020912170410156 + }, + { + "auxiliary_loss_clip": 0.01217297, + "auxiliary_loss_mlp": 0.01035876, + "balance_loss_clip": 1.05647206, + "balance_loss_mlp": 1.02682805, + "epoch": 0.25780075753021103, + "flos": 16472225278080.0, + "grad_norm": 2.0396082468840517, + "language_loss": 0.87725496, + "learning_rate": 3.4800535631571874e-06, + "loss": 0.89978671, + "num_input_tokens_seen": 45831310, + "step": 2144, + "time_per_iteration": 2.6723506450653076 + }, + { + "auxiliary_loss_clip": 0.01231009, + "auxiliary_loss_mlp": 0.01033343, + "balance_loss_clip": 1.05895507, + "balance_loss_mlp": 1.02418804, + "epoch": 0.25792100042085014, + "flos": 22820297846400.0, + "grad_norm": 2.0351084815111604, + "language_loss": 0.75930166, + "learning_rate": 3.4795295321816804e-06, + "loss": 0.78194523, + "num_input_tokens_seen": 45850135, + "step": 2145, + "time_per_iteration": 2.861455202102661 + }, + { + "auxiliary_loss_clip": 0.01212217, + "auxiliary_loss_mlp": 0.01032385, + "balance_loss_clip": 1.05632901, + "balance_loss_mlp": 1.02327752, + "epoch": 0.2580412433114892, + "flos": 18697286194560.0, + "grad_norm": 2.2048764887066525, + "language_loss": 0.91163772, + "learning_rate": 3.47900527676272e-06, + "loss": 0.9340837, + "num_input_tokens_seen": 45868470, + "step": 2146, + "time_per_iteration": 3.568800449371338 + }, + { + "auxiliary_loss_clip": 0.01208226, + "auxiliary_loss_mlp": 0.01040789, + "balance_loss_clip": 1.06175339, + "balance_loss_mlp": 1.0317347, + "epoch": 0.2581614862021283, + "flos": 14283146810880.0, + "grad_norm": 2.4628166734724966, + "language_loss": 0.88580066, + "learning_rate": 3.478480796979835e-06, + "loss": 0.90829074, + "num_input_tokens_seen": 45886355, + "step": 2147, + "time_per_iteration": 3.5869698524475098 + }, + { + "auxiliary_loss_clip": 0.01218204, + "auxiliary_loss_mlp": 0.01032618, + "balance_loss_clip": 1.05638266, + "balance_loss_mlp": 1.02344441, + "epoch": 0.25828172909276736, + "flos": 29498281856640.0, + "grad_norm": 1.5507516987900578, + "language_loss": 0.77818406, + "learning_rate": 3.4779560929125894e-06, + "loss": 0.8006922, + "num_input_tokens_seen": 45907900, + "step": 2148, + "time_per_iteration": 3.6074435710906982 + }, + { + "auxiliary_loss_clip": 0.01131597, + "auxiliary_loss_mlp": 0.0101165, + "balance_loss_clip": 1.02177429, + "balance_loss_mlp": 1.00952816, + "epoch": 0.2584019719834065, + "flos": 67114387376640.0, + "grad_norm": 0.7735430887039086, + "language_loss": 0.56907368, + "learning_rate": 3.4774311646405783e-06, + "loss": 0.59050608, + "num_input_tokens_seen": 45977805, + "step": 2149, + "time_per_iteration": 3.4195263385772705 + }, + { + "auxiliary_loss_clip": 0.01215307, + "auxiliary_loss_mlp": 0.01032451, + "balance_loss_clip": 1.05029535, + "balance_loss_mlp": 1.02364743, + "epoch": 0.2585222148740456, + "flos": 22893555634560.0, + "grad_norm": 2.3640260540185425, + "language_loss": 0.83601123, + "learning_rate": 3.476906012243435e-06, + "loss": 0.8584888, + "num_input_tokens_seen": 45996715, + "step": 2150, + "time_per_iteration": 2.793482780456543 + }, + { + "auxiliary_loss_clip": 0.01202928, + "auxiliary_loss_mlp": 0.01030433, + "balance_loss_clip": 1.05678165, + "balance_loss_mlp": 1.02150464, + "epoch": 0.25864245776468464, + "flos": 28909202808960.0, + "grad_norm": 1.6738088538608928, + "language_loss": 0.81102443, + "learning_rate": 3.476380635800824e-06, + "loss": 0.83335805, + "num_input_tokens_seen": 46017915, + "step": 2151, + "time_per_iteration": 2.748281717300415 + }, + { + "auxiliary_loss_clip": 0.01221194, + "auxiliary_loss_mlp": 0.01032879, + "balance_loss_clip": 1.05758333, + "balance_loss_mlp": 1.0232408, + "epoch": 0.25876270065532375, + "flos": 14793185980800.0, + "grad_norm": 1.9287776948149993, + "language_loss": 0.86116803, + "learning_rate": 3.475855035392444e-06, + "loss": 0.88370878, + "num_input_tokens_seen": 46033235, + "step": 2152, + "time_per_iteration": 3.7423088550567627 + }, + { + "auxiliary_loss_clip": 0.01242707, + "auxiliary_loss_mlp": 0.01031268, + "balance_loss_clip": 1.04953766, + "balance_loss_mlp": 1.02249467, + "epoch": 0.25888294354596286, + "flos": 60467821810560.0, + "grad_norm": 2.4198376250014153, + "language_loss": 0.71380925, + "learning_rate": 3.475329211098029e-06, + "loss": 0.73654896, + "num_input_tokens_seen": 46056390, + "step": 2153, + "time_per_iteration": 3.284705638885498 + }, + { + "auxiliary_loss_clip": 0.01245531, + "auxiliary_loss_mlp": 0.01032112, + "balance_loss_clip": 1.05553126, + "balance_loss_mlp": 1.02242637, + "epoch": 0.2590031864366019, + "flos": 27851166771840.0, + "grad_norm": 1.5961107591696126, + "language_loss": 0.82426864, + "learning_rate": 3.4748031629973453e-06, + "loss": 0.84704506, + "num_input_tokens_seen": 46077120, + "step": 2154, + "time_per_iteration": 2.793630838394165 + }, + { + "auxiliary_loss_clip": 0.01138774, + "auxiliary_loss_mlp": 0.01008325, + "balance_loss_clip": 1.01876187, + "balance_loss_mlp": 1.00614381, + "epoch": 0.25912342932724103, + "flos": 62422444206720.0, + "grad_norm": 0.9117589813085403, + "language_loss": 0.56667352, + "learning_rate": 3.4742768911701944e-06, + "loss": 0.58814454, + "num_input_tokens_seen": 46139815, + "step": 2155, + "time_per_iteration": 3.4253079891204834 + }, + { + "auxiliary_loss_clip": 0.01225272, + "auxiliary_loss_mlp": 0.01036662, + "balance_loss_clip": 1.06317258, + "balance_loss_mlp": 1.02713108, + "epoch": 0.25924367221788014, + "flos": 12378839368320.0, + "grad_norm": 2.8191502074097015, + "language_loss": 0.70694005, + "learning_rate": 3.4737503956964113e-06, + "loss": 0.72955942, + "num_input_tokens_seen": 46152120, + "step": 2156, + "time_per_iteration": 2.7269585132598877 + }, + { + "auxiliary_loss_clip": 0.01214615, + "auxiliary_loss_mlp": 0.01030888, + "balance_loss_clip": 1.05301404, + "balance_loss_mlp": 1.01980197, + "epoch": 0.2593639151085192, + "flos": 14575208296320.0, + "grad_norm": 2.063478122054419, + "language_loss": 0.67224717, + "learning_rate": 3.473223676655865e-06, + "loss": 0.69470221, + "num_input_tokens_seen": 46170120, + "step": 2157, + "time_per_iteration": 2.6292102336883545 + }, + { + "auxiliary_loss_clip": 0.01215841, + "auxiliary_loss_mlp": 0.01038533, + "balance_loss_clip": 1.05334294, + "balance_loss_mlp": 1.02770853, + "epoch": 0.2594841579991583, + "flos": 15230937029760.0, + "grad_norm": 1.8839944524495196, + "language_loss": 0.80066514, + "learning_rate": 3.472696734128459e-06, + "loss": 0.82320893, + "num_input_tokens_seen": 46187985, + "step": 2158, + "time_per_iteration": 2.685307025909424 + }, + { + "auxiliary_loss_clip": 0.01217532, + "auxiliary_loss_mlp": 0.01032012, + "balance_loss_clip": 1.06012547, + "balance_loss_mlp": 1.02202785, + "epoch": 0.2596044008897974, + "flos": 23623583650560.0, + "grad_norm": 1.6115928058186417, + "language_loss": 0.75736755, + "learning_rate": 3.4721695681941286e-06, + "loss": 0.77986294, + "num_input_tokens_seen": 46207025, + "step": 2159, + "time_per_iteration": 2.729637861251831 + }, + { + "auxiliary_loss_clip": 0.01219813, + "auxiliary_loss_mlp": 0.0138045, + "balance_loss_clip": 1.05420852, + "balance_loss_mlp": 1.00029063, + "epoch": 0.25972464378043647, + "flos": 13772281628160.0, + "grad_norm": 2.077876990088672, + "language_loss": 0.82587576, + "learning_rate": 3.471642178932845e-06, + "loss": 0.85187829, + "num_input_tokens_seen": 46225670, + "step": 2160, + "time_per_iteration": 2.6980783939361572 + }, + { + "auxiliary_loss_clip": 0.01223413, + "auxiliary_loss_mlp": 0.01032601, + "balance_loss_clip": 1.05294061, + "balance_loss_mlp": 1.02335072, + "epoch": 0.2598448866710756, + "flos": 19573578391680.0, + "grad_norm": 1.9198768288589718, + "language_loss": 0.89319217, + "learning_rate": 3.471114566424613e-06, + "loss": 0.91575229, + "num_input_tokens_seen": 46244130, + "step": 2161, + "time_per_iteration": 2.684412717819214 + }, + { + "auxiliary_loss_clip": 0.01224037, + "auxiliary_loss_mlp": 0.01028857, + "balance_loss_clip": 1.05800557, + "balance_loss_mlp": 1.01939762, + "epoch": 0.25996512956171464, + "flos": 21653237053440.0, + "grad_norm": 1.879518072407555, + "language_loss": 0.75898999, + "learning_rate": 3.4705867307494715e-06, + "loss": 0.78151894, + "num_input_tokens_seen": 46263200, + "step": 2162, + "time_per_iteration": 2.654383420944214 + }, + { + "auxiliary_loss_clip": 0.01218485, + "auxiliary_loss_mlp": 0.01036841, + "balance_loss_clip": 1.05777097, + "balance_loss_mlp": 1.02714324, + "epoch": 0.26008537245235375, + "flos": 18223480869120.0, + "grad_norm": 2.0588020410860532, + "language_loss": 0.84151947, + "learning_rate": 3.470058671987492e-06, + "loss": 0.86407274, + "num_input_tokens_seen": 46281465, + "step": 2163, + "time_per_iteration": 2.6107332706451416 + }, + { + "auxiliary_loss_clip": 0.01221569, + "auxiliary_loss_mlp": 0.01042555, + "balance_loss_clip": 1.05882633, + "balance_loss_mlp": 1.03173709, + "epoch": 0.26020561534299286, + "flos": 24645385843200.0, + "grad_norm": 1.9654048654504916, + "language_loss": 0.84543318, + "learning_rate": 3.4695303902187805e-06, + "loss": 0.86807442, + "num_input_tokens_seen": 46301020, + "step": 2164, + "time_per_iteration": 2.6695451736450195 + }, + { + "auxiliary_loss_clip": 0.0122235, + "auxiliary_loss_mlp": 0.01032069, + "balance_loss_clip": 1.05239534, + "balance_loss_mlp": 1.02240109, + "epoch": 0.2603258582336319, + "flos": 25773662926080.0, + "grad_norm": 2.502587255415566, + "language_loss": 0.78623116, + "learning_rate": 3.469001885523478e-06, + "loss": 0.80877537, + "num_input_tokens_seen": 46321740, + "step": 2165, + "time_per_iteration": 2.796175479888916 + }, + { + "auxiliary_loss_clip": 0.01209021, + "auxiliary_loss_mlp": 0.01030898, + "balance_loss_clip": 1.05984306, + "balance_loss_mlp": 1.02105093, + "epoch": 0.260446101124271, + "flos": 28766314506240.0, + "grad_norm": 1.6572090550511516, + "language_loss": 0.8079071, + "learning_rate": 3.4684731579817568e-06, + "loss": 0.83030629, + "num_input_tokens_seen": 46342730, + "step": 2166, + "time_per_iteration": 2.6582467555999756 + }, + { + "auxiliary_loss_clip": 0.01244961, + "auxiliary_loss_mlp": 0.01031209, + "balance_loss_clip": 1.05293095, + "balance_loss_mlp": 1.02215517, + "epoch": 0.26056634401491013, + "flos": 25666757072640.0, + "grad_norm": 2.1528915410694904, + "language_loss": 0.7708956, + "learning_rate": 3.4679442076738247e-06, + "loss": 0.7936573, + "num_input_tokens_seen": 46362445, + "step": 2167, + "time_per_iteration": 2.8891611099243164 + }, + { + "auxiliary_loss_clip": 0.01210321, + "auxiliary_loss_mlp": 0.0104065, + "balance_loss_clip": 1.06010938, + "balance_loss_mlp": 1.03095186, + "epoch": 0.2606865869055492, + "flos": 27052765217280.0, + "grad_norm": 1.8175369931256642, + "language_loss": 0.83359325, + "learning_rate": 3.4674150346799245e-06, + "loss": 0.85610294, + "num_input_tokens_seen": 46382145, + "step": 2168, + "time_per_iteration": 2.6521947383880615 + }, + { + "auxiliary_loss_clip": 0.01224507, + "auxiliary_loss_mlp": 0.01042271, + "balance_loss_clip": 1.05706906, + "balance_loss_mlp": 1.03253198, + "epoch": 0.2608068297961883, + "flos": 17712615686400.0, + "grad_norm": 2.112556089454543, + "language_loss": 0.80072427, + "learning_rate": 3.4668856390803295e-06, + "loss": 0.82339203, + "num_input_tokens_seen": 46400025, + "step": 2169, + "time_per_iteration": 2.7007148265838623 + }, + { + "auxiliary_loss_clip": 0.01198295, + "auxiliary_loss_mlp": 0.01032515, + "balance_loss_clip": 1.05282497, + "balance_loss_mlp": 1.02298427, + "epoch": 0.2609270726868274, + "flos": 18551632544640.0, + "grad_norm": 2.1047394334265648, + "language_loss": 0.89974499, + "learning_rate": 3.4663560209553495e-06, + "loss": 0.9220531, + "num_input_tokens_seen": 46418090, + "step": 2170, + "time_per_iteration": 2.584711790084839 + }, + { + "auxiliary_loss_clip": 0.0121034, + "auxiliary_loss_mlp": 0.01033391, + "balance_loss_clip": 1.0529989, + "balance_loss_mlp": 1.02385426, + "epoch": 0.26104731557746647, + "flos": 21835699165440.0, + "grad_norm": 2.0314482942690804, + "language_loss": 0.79382288, + "learning_rate": 3.4658261803853267e-06, + "loss": 0.81626022, + "num_input_tokens_seen": 46436015, + "step": 2171, + "time_per_iteration": 2.738187551498413 + }, + { + "auxiliary_loss_clip": 0.01222859, + "auxiliary_loss_mlp": 0.01032071, + "balance_loss_clip": 1.05892599, + "balance_loss_mlp": 1.02299345, + "epoch": 0.2611675584681056, + "flos": 21689650465920.0, + "grad_norm": 2.276285441052986, + "language_loss": 0.80886102, + "learning_rate": 3.4652961174506383e-06, + "loss": 0.83141035, + "num_input_tokens_seen": 46455885, + "step": 2172, + "time_per_iteration": 3.675459384918213 + }, + { + "auxiliary_loss_clip": 0.01107158, + "auxiliary_loss_mlp": 0.01016155, + "balance_loss_clip": 1.02210784, + "balance_loss_mlp": 1.01417589, + "epoch": 0.2612878013587447, + "flos": 71862101389440.0, + "grad_norm": 0.9840752790909838, + "language_loss": 0.58175421, + "learning_rate": 3.464765832231694e-06, + "loss": 0.60298729, + "num_input_tokens_seen": 46510050, + "step": 2173, + "time_per_iteration": 4.15064811706543 + }, + { + "auxiliary_loss_clip": 0.01218538, + "auxiliary_loss_mlp": 0.01033969, + "balance_loss_clip": 1.06047869, + "balance_loss_mlp": 1.02415216, + "epoch": 0.26140804424938374, + "flos": 20227511445120.0, + "grad_norm": 2.37522165481168, + "language_loss": 0.70966601, + "learning_rate": 3.4642353248089373e-06, + "loss": 0.73219109, + "num_input_tokens_seen": 46528810, + "step": 2174, + "time_per_iteration": 3.6938436031341553 + }, + { + "auxiliary_loss_clip": 0.0121844, + "auxiliary_loss_mlp": 0.0102942, + "balance_loss_clip": 1.05660796, + "balance_loss_mlp": 1.01982403, + "epoch": 0.26152828714002285, + "flos": 25557085872000.0, + "grad_norm": 1.651570458092742, + "language_loss": 0.80570805, + "learning_rate": 3.463704595262846e-06, + "loss": 0.82818663, + "num_input_tokens_seen": 46549690, + "step": 2175, + "time_per_iteration": 2.920823335647583 + }, + { + "auxiliary_loss_clip": 0.01229912, + "auxiliary_loss_mlp": 0.01035577, + "balance_loss_clip": 1.05502355, + "balance_loss_mlp": 1.02614164, + "epoch": 0.26164853003066196, + "flos": 25446516831360.0, + "grad_norm": 1.7886439830626768, + "language_loss": 0.70679724, + "learning_rate": 3.463173643673931e-06, + "loss": 0.72945213, + "num_input_tokens_seen": 46572215, + "step": 2176, + "time_per_iteration": 2.7191970348358154 + }, + { + "auxiliary_loss_clip": 0.01114574, + "auxiliary_loss_mlp": 0.01002755, + "balance_loss_clip": 1.02381086, + "balance_loss_mlp": 1.00064504, + "epoch": 0.261768772921301, + "flos": 53944580568960.0, + "grad_norm": 0.9146904947225053, + "language_loss": 0.63533998, + "learning_rate": 3.4626424701227387e-06, + "loss": 0.65651321, + "num_input_tokens_seen": 46627275, + "step": 2177, + "time_per_iteration": 3.1703293323516846 + }, + { + "auxiliary_loss_clip": 0.01104093, + "auxiliary_loss_mlp": 0.01003534, + "balance_loss_clip": 1.02548099, + "balance_loss_mlp": 1.00155473, + "epoch": 0.26188901581194013, + "flos": 70687606481280.0, + "grad_norm": 0.822668406739809, + "language_loss": 0.55777979, + "learning_rate": 3.4621110746898452e-06, + "loss": 0.57885599, + "num_input_tokens_seen": 46695135, + "step": 2178, + "time_per_iteration": 3.289511203765869 + }, + { + "auxiliary_loss_clip": 0.01218447, + "auxiliary_loss_mlp": 0.01028679, + "balance_loss_clip": 1.05827844, + "balance_loss_mlp": 1.01954722, + "epoch": 0.2620092587025792, + "flos": 21069580959360.0, + "grad_norm": 3.2240525677607614, + "language_loss": 0.7458837, + "learning_rate": 3.4615794574558654e-06, + "loss": 0.76835501, + "num_input_tokens_seen": 46714145, + "step": 2179, + "time_per_iteration": 3.6061549186706543 + }, + { + "auxiliary_loss_clip": 0.01224776, + "auxiliary_loss_mlp": 0.0103865, + "balance_loss_clip": 1.05570459, + "balance_loss_mlp": 1.02954245, + "epoch": 0.2621295015932183, + "flos": 18369601395840.0, + "grad_norm": 2.2861791854339506, + "language_loss": 0.83927566, + "learning_rate": 3.4610476185014436e-06, + "loss": 0.86190987, + "num_input_tokens_seen": 46731405, + "step": 2180, + "time_per_iteration": 2.6947364807128906 + }, + { + "auxiliary_loss_clip": 0.01208679, + "auxiliary_loss_mlp": 0.0103299, + "balance_loss_clip": 1.05946851, + "balance_loss_mlp": 1.02232647, + "epoch": 0.2622497444838574, + "flos": 23659997063040.0, + "grad_norm": 1.7169384953770814, + "language_loss": 0.79377961, + "learning_rate": 3.4605155579072597e-06, + "loss": 0.81619632, + "num_input_tokens_seen": 46751260, + "step": 2181, + "time_per_iteration": 2.6372616291046143 + }, + { + "auxiliary_loss_clip": 0.01232058, + "auxiliary_loss_mlp": 0.01032776, + "balance_loss_clip": 1.05069363, + "balance_loss_mlp": 1.02341771, + "epoch": 0.26236998737449646, + "flos": 22123810154880.0, + "grad_norm": 1.940760603383293, + "language_loss": 0.71541196, + "learning_rate": 3.459983275754027e-06, + "loss": 0.73806036, + "num_input_tokens_seen": 46770155, + "step": 2182, + "time_per_iteration": 2.9069955348968506 + }, + { + "auxiliary_loss_clip": 0.01209634, + "auxiliary_loss_mlp": 0.01030977, + "balance_loss_clip": 1.0601896, + "balance_loss_mlp": 1.02221501, + "epoch": 0.26249023026513557, + "flos": 17895185539200.0, + "grad_norm": 2.214306641495156, + "language_loss": 0.79647022, + "learning_rate": 3.4594507721224918e-06, + "loss": 0.81887627, + "num_input_tokens_seen": 46788805, + "step": 2183, + "time_per_iteration": 2.5900607109069824 + }, + { + "auxiliary_loss_clip": 0.01224329, + "auxiliary_loss_mlp": 0.01033214, + "balance_loss_clip": 1.05568314, + "balance_loss_mlp": 1.0247736, + "epoch": 0.2626104731557747, + "flos": 18332936588160.0, + "grad_norm": 1.9697836149293575, + "language_loss": 0.82054555, + "learning_rate": 3.4589180470934353e-06, + "loss": 0.84312105, + "num_input_tokens_seen": 46808670, + "step": 2184, + "time_per_iteration": 2.7852089405059814 + }, + { + "auxiliary_loss_clip": 0.01220797, + "auxiliary_loss_mlp": 0.01032575, + "balance_loss_clip": 1.05583739, + "balance_loss_mlp": 1.02283525, + "epoch": 0.26273071604641374, + "flos": 19317714837120.0, + "grad_norm": 2.7768479267388444, + "language_loss": 0.76913261, + "learning_rate": 3.4583851007476713e-06, + "loss": 0.79166627, + "num_input_tokens_seen": 46827140, + "step": 2185, + "time_per_iteration": 2.6660053730010986 + }, + { + "auxiliary_loss_clip": 0.01232776, + "auxiliary_loss_mlp": 0.01028458, + "balance_loss_clip": 1.05404699, + "balance_loss_mlp": 1.01897514, + "epoch": 0.26285095893705285, + "flos": 18327477720960.0, + "grad_norm": 2.2004648010039753, + "language_loss": 0.68424332, + "learning_rate": 3.4578519331660464e-06, + "loss": 0.70685565, + "num_input_tokens_seen": 46844135, + "step": 2186, + "time_per_iteration": 2.6474878787994385 + }, + { + "auxiliary_loss_clip": 0.01211849, + "auxiliary_loss_mlp": 0.01038173, + "balance_loss_clip": 1.05851877, + "balance_loss_mlp": 1.02947116, + "epoch": 0.26297120182769196, + "flos": 20193827466240.0, + "grad_norm": 2.0728316903144703, + "language_loss": 0.82033056, + "learning_rate": 3.4573185444294426e-06, + "loss": 0.84283078, + "num_input_tokens_seen": 46862500, + "step": 2187, + "time_per_iteration": 2.714618444442749 + }, + { + "auxiliary_loss_clip": 0.01222099, + "auxiliary_loss_mlp": 0.01380888, + "balance_loss_clip": 1.05555284, + "balance_loss_mlp": 1.00034916, + "epoch": 0.263091444718331, + "flos": 22418421505920.0, + "grad_norm": 1.7051026739117001, + "language_loss": 0.78804195, + "learning_rate": 3.456784934618774e-06, + "loss": 0.81407177, + "num_input_tokens_seen": 46883665, + "step": 2188, + "time_per_iteration": 2.6744048595428467 + }, + { + "auxiliary_loss_clip": 0.01222134, + "auxiliary_loss_mlp": 0.0103131, + "balance_loss_clip": 1.05545974, + "balance_loss_mlp": 1.02188683, + "epoch": 0.2632116876089701, + "flos": 19024827338880.0, + "grad_norm": 1.8950154477754424, + "language_loss": 0.79905772, + "learning_rate": 3.4562511038149897e-06, + "loss": 0.82159221, + "num_input_tokens_seen": 46899160, + "step": 2189, + "time_per_iteration": 2.686706304550171 + }, + { + "auxiliary_loss_clip": 0.01133349, + "auxiliary_loss_mlp": 0.01025834, + "balance_loss_clip": 1.01374555, + "balance_loss_mlp": 1.02366436, + "epoch": 0.26333193049960923, + "flos": 67308054531840.0, + "grad_norm": 0.8664077272640088, + "language_loss": 0.57725883, + "learning_rate": 3.4557170520990705e-06, + "loss": 0.59885061, + "num_input_tokens_seen": 46959835, + "step": 2190, + "time_per_iteration": 3.291184663772583 + }, + { + "auxiliary_loss_clip": 0.01211303, + "auxiliary_loss_mlp": 0.01029162, + "balance_loss_clip": 1.05774665, + "balance_loss_mlp": 1.01998854, + "epoch": 0.2634521733902483, + "flos": 25048806468480.0, + "grad_norm": 1.4752950161457492, + "language_loss": 0.86464643, + "learning_rate": 3.4551827795520324e-06, + "loss": 0.88705099, + "num_input_tokens_seen": 46982720, + "step": 2191, + "time_per_iteration": 2.7689669132232666 + }, + { + "auxiliary_loss_clip": 0.01215701, + "auxiliary_loss_mlp": 0.01027578, + "balance_loss_clip": 1.05709803, + "balance_loss_mlp": 1.01854205, + "epoch": 0.2635724162808874, + "flos": 20594985534720.0, + "grad_norm": 1.651864980237364, + "language_loss": 0.84919953, + "learning_rate": 3.4546482862549226e-06, + "loss": 0.87163234, + "num_input_tokens_seen": 47003035, + "step": 2192, + "time_per_iteration": 2.6281447410583496 + }, + { + "auxiliary_loss_clip": 0.01224498, + "auxiliary_loss_mlp": 0.01033561, + "balance_loss_clip": 1.05269694, + "balance_loss_mlp": 1.02419138, + "epoch": 0.2636926591715265, + "flos": 19244636616960.0, + "grad_norm": 3.0715770658252923, + "language_loss": 0.78716815, + "learning_rate": 3.4541135722888253e-06, + "loss": 0.80974871, + "num_input_tokens_seen": 47019625, + "step": 2193, + "time_per_iteration": 2.7919211387634277 + }, + { + "auxiliary_loss_clip": 0.01209385, + "auxiliary_loss_mlp": 0.01029083, + "balance_loss_clip": 1.06006503, + "balance_loss_mlp": 1.01989806, + "epoch": 0.26381290206216557, + "flos": 28804882734720.0, + "grad_norm": 1.7399282832062217, + "language_loss": 0.804896, + "learning_rate": 3.453578637734854e-06, + "loss": 0.82728076, + "num_input_tokens_seen": 47040815, + "step": 2194, + "time_per_iteration": 2.6959660053253174 + }, + { + "auxiliary_loss_clip": 0.01211026, + "auxiliary_loss_mlp": 0.01040484, + "balance_loss_clip": 1.06294394, + "balance_loss_mlp": 1.03100133, + "epoch": 0.2639331449528047, + "flos": 25008909436800.0, + "grad_norm": 1.630562100958971, + "language_loss": 0.78735113, + "learning_rate": 3.4530434826741605e-06, + "loss": 0.80986625, + "num_input_tokens_seen": 47061755, + "step": 2195, + "time_per_iteration": 2.6638059616088867 + }, + { + "auxiliary_loss_clip": 0.01217943, + "auxiliary_loss_mlp": 0.01031667, + "balance_loss_clip": 1.05563736, + "balance_loss_mlp": 1.02268457, + "epoch": 0.26405338784344373, + "flos": 46535775465600.0, + "grad_norm": 1.5526169543963748, + "language_loss": 0.68758756, + "learning_rate": 3.452508107187926e-06, + "loss": 0.71008366, + "num_input_tokens_seen": 47085130, + "step": 2196, + "time_per_iteration": 2.912713050842285 + }, + { + "auxiliary_loss_clip": 0.01248515, + "auxiliary_loss_mlp": 0.01038417, + "balance_loss_clip": 1.05010247, + "balance_loss_mlp": 1.02854681, + "epoch": 0.26417363073408284, + "flos": 21179467641600.0, + "grad_norm": 1.7466587451577231, + "language_loss": 0.77068424, + "learning_rate": 3.451972511357366e-06, + "loss": 0.79355359, + "num_input_tokens_seen": 47104675, + "step": 2197, + "time_per_iteration": 2.7853245735168457 + }, + { + "auxiliary_loss_clip": 0.01211425, + "auxiliary_loss_mlp": 0.01033553, + "balance_loss_clip": 1.05775142, + "balance_loss_mlp": 1.02514863, + "epoch": 0.26429387362472195, + "flos": 22674751937280.0, + "grad_norm": 1.811539468310212, + "language_loss": 0.85124999, + "learning_rate": 3.45143669526373e-06, + "loss": 0.87369972, + "num_input_tokens_seen": 47124435, + "step": 2198, + "time_per_iteration": 3.5709547996520996 + }, + { + "auxiliary_loss_clip": 0.01131026, + "auxiliary_loss_mlp": 0.01009056, + "balance_loss_clip": 1.02721357, + "balance_loss_mlp": 1.0071249, + "epoch": 0.264414116515361, + "flos": 67180534272000.0, + "grad_norm": 0.7777959835442676, + "language_loss": 0.63257444, + "learning_rate": 3.450900658988302e-06, + "loss": 0.65397525, + "num_input_tokens_seen": 47185985, + "step": 2199, + "time_per_iteration": 4.017361879348755 + }, + { + "auxiliary_loss_clip": 0.01211535, + "auxiliary_loss_mlp": 0.01033904, + "balance_loss_clip": 1.05652571, + "balance_loss_mlp": 1.02410471, + "epoch": 0.2645343594060001, + "flos": 25664709997440.0, + "grad_norm": 2.084564260327516, + "language_loss": 0.77859712, + "learning_rate": 3.450364402612397e-06, + "loss": 0.8010515, + "num_input_tokens_seen": 47203140, + "step": 2200, + "time_per_iteration": 3.603665828704834 + }, + { + "auxiliary_loss_clip": 0.01219898, + "auxiliary_loss_mlp": 0.01037554, + "balance_loss_clip": 1.05553865, + "balance_loss_mlp": 1.02739096, + "epoch": 0.26465460229663923, + "flos": 22491822948480.0, + "grad_norm": 2.1551511208209586, + "language_loss": 0.83966374, + "learning_rate": 3.449827926217366e-06, + "loss": 0.86223829, + "num_input_tokens_seen": 47222575, + "step": 2201, + "time_per_iteration": 2.8372690677642822 + }, + { + "auxiliary_loss_clip": 0.01225623, + "auxiliary_loss_mlp": 0.01033248, + "balance_loss_clip": 1.05334866, + "balance_loss_mlp": 1.02378845, + "epoch": 0.2647748451872783, + "flos": 29388036038400.0, + "grad_norm": 1.8648576279445714, + "language_loss": 0.80534261, + "learning_rate": 3.449291229884591e-06, + "loss": 0.82793128, + "num_input_tokens_seen": 47243815, + "step": 2202, + "time_per_iteration": 2.7660768032073975 + }, + { + "auxiliary_loss_clip": 0.01236824, + "auxiliary_loss_mlp": 0.01027505, + "balance_loss_clip": 1.0560925, + "balance_loss_mlp": 1.01841509, + "epoch": 0.2648950880779174, + "flos": 26797799502720.0, + "grad_norm": 1.7204800136598697, + "language_loss": 0.86534262, + "learning_rate": 3.4487543136954887e-06, + "loss": 0.88798594, + "num_input_tokens_seen": 47263435, + "step": 2203, + "time_per_iteration": 2.789651393890381 + }, + { + "auxiliary_loss_clip": 0.01230437, + "auxiliary_loss_mlp": 0.01033747, + "balance_loss_clip": 1.05494356, + "balance_loss_mlp": 1.02512205, + "epoch": 0.2650153309685565, + "flos": 28841008838400.0, + "grad_norm": 1.718640614249469, + "language_loss": 0.91146791, + "learning_rate": 3.448217177731509e-06, + "loss": 0.93410975, + "num_input_tokens_seen": 47283920, + "step": 2204, + "time_per_iteration": 2.887577533721924 + }, + { + "auxiliary_loss_clip": 0.01218165, + "auxiliary_loss_mlp": 0.01032925, + "balance_loss_clip": 1.05775189, + "balance_loss_mlp": 1.02380586, + "epoch": 0.26513557385919556, + "flos": 20303247271680.0, + "grad_norm": 1.9593457535676184, + "language_loss": 0.77985072, + "learning_rate": 3.4476798220741348e-06, + "loss": 0.80236161, + "num_input_tokens_seen": 47302800, + "step": 2205, + "time_per_iteration": 3.57889986038208 + }, + { + "auxiliary_loss_clip": 0.01209353, + "auxiliary_loss_mlp": 0.01038043, + "balance_loss_clip": 1.06321228, + "balance_loss_mlp": 1.02946591, + "epoch": 0.26525581674983467, + "flos": 17676274101120.0, + "grad_norm": 2.3107993555692254, + "language_loss": 0.78677309, + "learning_rate": 3.4471422468048826e-06, + "loss": 0.80924714, + "num_input_tokens_seen": 47321525, + "step": 2206, + "time_per_iteration": 2.6441757678985596 + }, + { + "auxiliary_loss_clip": 0.01205617, + "auxiliary_loss_mlp": 0.01032574, + "balance_loss_clip": 1.05888343, + "balance_loss_mlp": 1.02318609, + "epoch": 0.2653760596404738, + "flos": 26833746038400.0, + "grad_norm": 2.173148603802907, + "language_loss": 0.72300112, + "learning_rate": 3.4466044520053022e-06, + "loss": 0.74538302, + "num_input_tokens_seen": 47340530, + "step": 2207, + "time_per_iteration": 2.640467405319214 + }, + { + "auxiliary_loss_clip": 0.01205239, + "auxiliary_loss_mlp": 0.01034849, + "balance_loss_clip": 1.05267239, + "balance_loss_mlp": 1.02531815, + "epoch": 0.26549630253111284, + "flos": 22782160581120.0, + "grad_norm": 1.9297562939679014, + "language_loss": 0.5995369, + "learning_rate": 3.446066437756977e-06, + "loss": 0.62193769, + "num_input_tokens_seen": 47359735, + "step": 2208, + "time_per_iteration": 2.7569785118103027 + }, + { + "auxiliary_loss_clip": 0.01220664, + "auxiliary_loss_mlp": 0.0102889, + "balance_loss_clip": 1.05686986, + "balance_loss_mlp": 1.01994348, + "epoch": 0.26561654542175195, + "flos": 23550002640000.0, + "grad_norm": 2.808159602193546, + "language_loss": 0.75476497, + "learning_rate": 3.4455282041415224e-06, + "loss": 0.77726054, + "num_input_tokens_seen": 47378945, + "step": 2209, + "time_per_iteration": 2.7264621257781982 + }, + { + "auxiliary_loss_clip": 0.01235071, + "auxiliary_loss_mlp": 0.01039806, + "balance_loss_clip": 1.05671883, + "balance_loss_mlp": 1.03010249, + "epoch": 0.265736788312391, + "flos": 26906680604160.0, + "grad_norm": 2.114032761275534, + "language_loss": 0.8703264, + "learning_rate": 3.4449897512405894e-06, + "loss": 0.89307523, + "num_input_tokens_seen": 47398095, + "step": 2210, + "time_per_iteration": 2.755340814590454 + }, + { + "auxiliary_loss_clip": 0.01238268, + "auxiliary_loss_mlp": 0.01380383, + "balance_loss_clip": 1.04725623, + "balance_loss_mlp": 1.00027621, + "epoch": 0.2658570312030301, + "flos": 23477139901440.0, + "grad_norm": 2.022354431196128, + "language_loss": 0.75557667, + "learning_rate": 3.444451079135859e-06, + "loss": 0.78176314, + "num_input_tokens_seen": 47417605, + "step": 2211, + "time_per_iteration": 2.8340351581573486 + }, + { + "auxiliary_loss_clip": 0.01222418, + "auxiliary_loss_mlp": 0.01380462, + "balance_loss_clip": 1.04820371, + "balance_loss_mlp": 1.00023055, + "epoch": 0.2659772740936692, + "flos": 21866402315520.0, + "grad_norm": 1.8309581700566324, + "language_loss": 0.74086559, + "learning_rate": 3.4439121879090493e-06, + "loss": 0.76689446, + "num_input_tokens_seen": 47435385, + "step": 2212, + "time_per_iteration": 2.7328402996063232 + }, + { + "auxiliary_loss_clip": 0.01224757, + "auxiliary_loss_mlp": 0.01033694, + "balance_loss_clip": 1.05641758, + "balance_loss_mlp": 1.02433562, + "epoch": 0.2660975169843083, + "flos": 19793100360960.0, + "grad_norm": 1.9465613920858156, + "language_loss": 0.83115602, + "learning_rate": 3.4433730776419082e-06, + "loss": 0.85374057, + "num_input_tokens_seen": 47454310, + "step": 2213, + "time_per_iteration": 2.836148738861084 + }, + { + "auxiliary_loss_clip": 0.01216577, + "auxiliary_loss_mlp": 0.01380633, + "balance_loss_clip": 1.05672204, + "balance_loss_mlp": 1.0004214, + "epoch": 0.2662177598749474, + "flos": 29018981750400.0, + "grad_norm": 3.1177407336931093, + "language_loss": 0.80465829, + "learning_rate": 3.4428337484162183e-06, + "loss": 0.83063036, + "num_input_tokens_seen": 47475120, + "step": 2214, + "time_per_iteration": 2.7113139629364014 + }, + { + "auxiliary_loss_clip": 0.01217971, + "auxiliary_loss_mlp": 0.01027149, + "balance_loss_clip": 1.05442607, + "balance_loss_mlp": 1.01748705, + "epoch": 0.2663380027655865, + "flos": 21762549118080.0, + "grad_norm": 2.2210508036272203, + "language_loss": 0.8459549, + "learning_rate": 3.442294200313797e-06, + "loss": 0.86840606, + "num_input_tokens_seen": 47493150, + "step": 2215, + "time_per_iteration": 2.748279094696045 + }, + { + "auxiliary_loss_clip": 0.0110071, + "auxiliary_loss_mlp": 0.01003372, + "balance_loss_clip": 1.02418447, + "balance_loss_mlp": 1.00157213, + "epoch": 0.26645824565622556, + "flos": 66980333819520.0, + "grad_norm": 0.765188641017228, + "language_loss": 0.52677399, + "learning_rate": 3.4417544334164916e-06, + "loss": 0.54781479, + "num_input_tokens_seen": 47557295, + "step": 2216, + "time_per_iteration": 3.229485034942627 + }, + { + "auxiliary_loss_clip": 0.01229581, + "auxiliary_loss_mlp": 0.0102833, + "balance_loss_clip": 1.05453408, + "balance_loss_mlp": 1.01943052, + "epoch": 0.26657848854686467, + "flos": 25264198373760.0, + "grad_norm": 1.8829617004853132, + "language_loss": 0.77360475, + "learning_rate": 3.4412144478061854e-06, + "loss": 0.79618382, + "num_input_tokens_seen": 47579705, + "step": 2217, + "time_per_iteration": 2.764317274093628 + }, + { + "auxiliary_loss_clip": 0.0126294, + "auxiliary_loss_mlp": 0.01034763, + "balance_loss_clip": 1.04971194, + "balance_loss_mlp": 1.02512455, + "epoch": 0.2666987314375038, + "flos": 23696769611520.0, + "grad_norm": 1.8079416582060677, + "language_loss": 0.75140035, + "learning_rate": 3.4406742435647925e-06, + "loss": 0.77437747, + "num_input_tokens_seen": 47599770, + "step": 2218, + "time_per_iteration": 3.052056312561035 + }, + { + "auxiliary_loss_clip": 0.01213833, + "auxiliary_loss_mlp": 0.01031219, + "balance_loss_clip": 1.06007683, + "balance_loss_mlp": 1.02239192, + "epoch": 0.26681897432814283, + "flos": 27048958375680.0, + "grad_norm": 2.148802480498236, + "language_loss": 0.79126745, + "learning_rate": 3.440133820774263e-06, + "loss": 0.81371802, + "num_input_tokens_seen": 47619580, + "step": 2219, + "time_per_iteration": 2.9522221088409424 + }, + { + "auxiliary_loss_clip": 0.01228616, + "auxiliary_loss_mlp": 0.01038177, + "balance_loss_clip": 1.05927444, + "balance_loss_mlp": 1.02779961, + "epoch": 0.26693921721878194, + "flos": 28985944216320.0, + "grad_norm": 2.034085458096377, + "language_loss": 0.81835878, + "learning_rate": 3.439593179516578e-06, + "loss": 0.84102678, + "num_input_tokens_seen": 47639490, + "step": 2220, + "time_per_iteration": 2.85662579536438 + }, + { + "auxiliary_loss_clip": 0.01228002, + "auxiliary_loss_mlp": 0.01038539, + "balance_loss_clip": 1.05821991, + "balance_loss_mlp": 1.02820992, + "epoch": 0.26705946010942105, + "flos": 21507834798720.0, + "grad_norm": 1.915219070968461, + "language_loss": 0.81182206, + "learning_rate": 3.4390523198737524e-06, + "loss": 0.8344875, + "num_input_tokens_seen": 47658650, + "step": 2221, + "time_per_iteration": 2.7167282104492188 + }, + { + "auxiliary_loss_clip": 0.0120928, + "auxiliary_loss_mlp": 0.01380475, + "balance_loss_clip": 1.06094813, + "balance_loss_mlp": 1.00033987, + "epoch": 0.2671797030000601, + "flos": 21471277731840.0, + "grad_norm": 3.9031208073887393, + "language_loss": 0.73586571, + "learning_rate": 3.4385112419278333e-06, + "loss": 0.76176322, + "num_input_tokens_seen": 47679875, + "step": 2222, + "time_per_iteration": 2.708101987838745 + }, + { + "auxiliary_loss_clip": 0.01116443, + "auxiliary_loss_mlp": 0.01000974, + "balance_loss_clip": 1.02653635, + "balance_loss_mlp": 0.99910212, + "epoch": 0.2672999458906992, + "flos": 64189929767040.0, + "grad_norm": 0.7912188360324073, + "language_loss": 0.64871269, + "learning_rate": 3.4379699457609033e-06, + "loss": 0.66988689, + "num_input_tokens_seen": 47737700, + "step": 2223, + "time_per_iteration": 3.198162317276001 + }, + { + "auxiliary_loss_clip": 0.01212678, + "auxiliary_loss_mlp": 0.01033529, + "balance_loss_clip": 1.05464637, + "balance_loss_mlp": 1.02346122, + "epoch": 0.26742018878133833, + "flos": 16909042573440.0, + "grad_norm": 1.8665783665837197, + "language_loss": 0.90143472, + "learning_rate": 3.4374284314550755e-06, + "loss": 0.92389679, + "num_input_tokens_seen": 47756740, + "step": 2224, + "time_per_iteration": 3.608497142791748 + }, + { + "auxiliary_loss_clip": 0.01208459, + "auxiliary_loss_mlp": 0.01033636, + "balance_loss_clip": 1.06146479, + "balance_loss_mlp": 1.02392638, + "epoch": 0.2675404316719774, + "flos": 20667560964480.0, + "grad_norm": 4.637717368239258, + "language_loss": 0.80993462, + "learning_rate": 3.436886699092498e-06, + "loss": 0.8323555, + "num_input_tokens_seen": 47775255, + "step": 2225, + "time_per_iteration": 3.564652681350708 + }, + { + "auxiliary_loss_clip": 0.01211686, + "auxiliary_loss_mlp": 0.01035556, + "balance_loss_clip": 1.06181395, + "balance_loss_mlp": 1.02590597, + "epoch": 0.2676606745626165, + "flos": 17485013157120.0, + "grad_norm": 2.651786450985324, + "language_loss": 0.71823311, + "learning_rate": 3.4363447487553502e-06, + "loss": 0.74070549, + "num_input_tokens_seen": 47788570, + "step": 2226, + "time_per_iteration": 3.476790428161621 + }, + { + "auxiliary_loss_clip": 0.01220444, + "auxiliary_loss_mlp": 0.01033016, + "balance_loss_clip": 1.05818844, + "balance_loss_mlp": 1.0234015, + "epoch": 0.26778091745325555, + "flos": 27852675143040.0, + "grad_norm": 1.7676004851850653, + "language_loss": 0.77940154, + "learning_rate": 3.4358025805258455e-06, + "loss": 0.80193615, + "num_input_tokens_seen": 47808275, + "step": 2227, + "time_per_iteration": 2.7439730167388916 + }, + { + "auxiliary_loss_clip": 0.0124637, + "auxiliary_loss_mlp": 0.01031292, + "balance_loss_clip": 1.05519879, + "balance_loss_mlp": 1.02193379, + "epoch": 0.26790116034389466, + "flos": 20955995176320.0, + "grad_norm": 1.730377142088732, + "language_loss": 0.83152902, + "learning_rate": 3.435260194486232e-06, + "loss": 0.85430568, + "num_input_tokens_seen": 47826245, + "step": 2228, + "time_per_iteration": 2.7925221920013428 + }, + { + "auxiliary_loss_clip": 0.0122432, + "auxiliary_loss_mlp": 0.01034921, + "balance_loss_clip": 1.05712152, + "balance_loss_mlp": 1.02522969, + "epoch": 0.2680214032345338, + "flos": 18040659621120.0, + "grad_norm": 2.3396249851453343, + "language_loss": 0.81878906, + "learning_rate": 3.4347175907187875e-06, + "loss": 0.84138155, + "num_input_tokens_seen": 47843235, + "step": 2229, + "time_per_iteration": 2.6411373615264893 + }, + { + "auxiliary_loss_clip": 0.01212898, + "auxiliary_loss_mlp": 0.01034208, + "balance_loss_clip": 1.05738592, + "balance_loss_mlp": 1.02526748, + "epoch": 0.26814164612517283, + "flos": 22419427086720.0, + "grad_norm": 1.7334530115355395, + "language_loss": 0.88016117, + "learning_rate": 3.4341747693058254e-06, + "loss": 0.90263224, + "num_input_tokens_seen": 47861710, + "step": 2230, + "time_per_iteration": 3.6107335090637207 + }, + { + "auxiliary_loss_clip": 0.01253893, + "auxiliary_loss_mlp": 0.01039985, + "balance_loss_clip": 1.04847157, + "balance_loss_mlp": 1.03044784, + "epoch": 0.26826188901581194, + "flos": 35627371159680.0, + "grad_norm": 1.7302813357756432, + "language_loss": 0.76921201, + "learning_rate": 3.4336317303296916e-06, + "loss": 0.79215074, + "num_input_tokens_seen": 47882685, + "step": 2231, + "time_per_iteration": 3.1316630840301514 + }, + { + "auxiliary_loss_clip": 0.01212541, + "auxiliary_loss_mlp": 0.01030146, + "balance_loss_clip": 1.05815399, + "balance_loss_mlp": 1.02004266, + "epoch": 0.26838213190645105, + "flos": 17639788861440.0, + "grad_norm": 1.9336535108368937, + "language_loss": 0.7501725, + "learning_rate": 3.4330884738727635e-06, + "loss": 0.7725994, + "num_input_tokens_seen": 47900860, + "step": 2232, + "time_per_iteration": 2.9499638080596924 + }, + { + "auxiliary_loss_clip": 0.01227287, + "auxiliary_loss_mlp": 0.01028061, + "balance_loss_clip": 1.0520072, + "balance_loss_mlp": 1.01868558, + "epoch": 0.2685023747970901, + "flos": 22674823764480.0, + "grad_norm": 1.8847510728466983, + "language_loss": 0.70509893, + "learning_rate": 3.4325450000174535e-06, + "loss": 0.72765243, + "num_input_tokens_seen": 47917500, + "step": 2233, + "time_per_iteration": 2.8162424564361572 + }, + { + "auxiliary_loss_clip": 0.0123082, + "auxiliary_loss_mlp": 0.01037504, + "balance_loss_clip": 1.05201149, + "balance_loss_mlp": 1.02785444, + "epoch": 0.2686226176877292, + "flos": 20120533764480.0, + "grad_norm": 1.7502246659383478, + "language_loss": 0.74601918, + "learning_rate": 3.4320013088462067e-06, + "loss": 0.76870245, + "num_input_tokens_seen": 47934860, + "step": 2234, + "time_per_iteration": 2.904542922973633 + }, + { + "auxiliary_loss_clip": 0.01231451, + "auxiliary_loss_mlp": 0.01030613, + "balance_loss_clip": 1.05314589, + "balance_loss_mlp": 1.02115941, + "epoch": 0.2687428605783683, + "flos": 21872040750720.0, + "grad_norm": 2.052557923520032, + "language_loss": 0.81251782, + "learning_rate": 3.431457400441499e-06, + "loss": 0.83513844, + "num_input_tokens_seen": 47955255, + "step": 2235, + "time_per_iteration": 2.8207509517669678 + }, + { + "auxiliary_loss_clip": 0.01147782, + "auxiliary_loss_mlp": 0.01001085, + "balance_loss_clip": 1.02007365, + "balance_loss_mlp": 0.99907058, + "epoch": 0.2688631034690074, + "flos": 69943320766080.0, + "grad_norm": 0.9417848186705835, + "language_loss": 0.60914397, + "learning_rate": 3.4309132748858424e-06, + "loss": 0.63063264, + "num_input_tokens_seen": 48016245, + "step": 2236, + "time_per_iteration": 3.3416733741760254 + }, + { + "auxiliary_loss_clip": 0.01214198, + "auxiliary_loss_mlp": 0.01039569, + "balance_loss_clip": 1.06131995, + "balance_loss_mlp": 1.03012776, + "epoch": 0.2689833463596465, + "flos": 22856639431680.0, + "grad_norm": 1.5792926197653925, + "language_loss": 0.8354094, + "learning_rate": 3.430368932261779e-06, + "loss": 0.85794705, + "num_input_tokens_seen": 48036600, + "step": 2237, + "time_per_iteration": 2.6900312900543213 + }, + { + "auxiliary_loss_clip": 0.0122166, + "auxiliary_loss_mlp": 0.01030487, + "balance_loss_clip": 1.05768943, + "balance_loss_mlp": 1.02125978, + "epoch": 0.2691035892502856, + "flos": 17200242132480.0, + "grad_norm": 1.8813601677230776, + "language_loss": 0.74766016, + "learning_rate": 3.429824372651886e-06, + "loss": 0.77018166, + "num_input_tokens_seen": 48054750, + "step": 2238, + "time_per_iteration": 2.7075231075286865 + }, + { + "auxiliary_loss_clip": 0.01251439, + "auxiliary_loss_mlp": 0.01029858, + "balance_loss_clip": 1.05703259, + "balance_loss_mlp": 1.01986837, + "epoch": 0.26922383214092466, + "flos": 17747484814080.0, + "grad_norm": 2.0118684933918916, + "language_loss": 0.83677709, + "learning_rate": 3.4292795961387732e-06, + "loss": 0.85959011, + "num_input_tokens_seen": 48072650, + "step": 2239, + "time_per_iteration": 2.7277657985687256 + }, + { + "auxiliary_loss_clip": 0.01207416, + "auxiliary_loss_mlp": 0.01031562, + "balance_loss_clip": 1.05978048, + "balance_loss_mlp": 1.02187085, + "epoch": 0.26934407503156377, + "flos": 16173376122240.0, + "grad_norm": 2.1223755697928617, + "language_loss": 0.87406135, + "learning_rate": 3.4287346028050818e-06, + "loss": 0.89645112, + "num_input_tokens_seen": 48088720, + "step": 2240, + "time_per_iteration": 2.6661407947540283 + }, + { + "auxiliary_loss_clip": 0.01221284, + "auxiliary_loss_mlp": 0.01031816, + "balance_loss_clip": 1.05556488, + "balance_loss_mlp": 1.02331018, + "epoch": 0.2694643179222028, + "flos": 23732895715200.0, + "grad_norm": 1.464961558777702, + "language_loss": 0.7928009, + "learning_rate": 3.4281893927334866e-06, + "loss": 0.81533194, + "num_input_tokens_seen": 48108630, + "step": 2241, + "time_per_iteration": 2.6879358291625977 + }, + { + "auxiliary_loss_clip": 0.0121628, + "auxiliary_loss_mlp": 0.01029354, + "balance_loss_clip": 1.05984402, + "balance_loss_mlp": 1.0201447, + "epoch": 0.26958456081284193, + "flos": 24718140840960.0, + "grad_norm": 1.8252118112916393, + "language_loss": 0.75406432, + "learning_rate": 3.4276439660066963e-06, + "loss": 0.77652067, + "num_input_tokens_seen": 48128330, + "step": 2242, + "time_per_iteration": 2.7020914554595947 + }, + { + "auxiliary_loss_clip": 0.01204818, + "auxiliary_loss_mlp": 0.0103591, + "balance_loss_clip": 1.06049347, + "balance_loss_mlp": 1.02689171, + "epoch": 0.26970480370348104, + "flos": 18112588606080.0, + "grad_norm": 2.358869604437762, + "language_loss": 0.8413738, + "learning_rate": 3.427098322707452e-06, + "loss": 0.86378109, + "num_input_tokens_seen": 48144295, + "step": 2243, + "time_per_iteration": 2.5742690563201904 + }, + { + "auxiliary_loss_clip": 0.01217401, + "auxiliary_loss_mlp": 0.01042975, + "balance_loss_clip": 1.06401575, + "balance_loss_mlp": 1.03347397, + "epoch": 0.2698250465941201, + "flos": 10816546250880.0, + "grad_norm": 2.227851845339906, + "language_loss": 0.8974328, + "learning_rate": 3.426552462918526e-06, + "loss": 0.92003655, + "num_input_tokens_seen": 48162230, + "step": 2244, + "time_per_iteration": 2.6588292121887207 + }, + { + "auxiliary_loss_clip": 0.01207353, + "auxiliary_loss_mlp": 0.01041422, + "balance_loss_clip": 1.06364703, + "balance_loss_mlp": 1.03187954, + "epoch": 0.2699452894847592, + "flos": 17308117653120.0, + "grad_norm": 2.431832469679547, + "language_loss": 0.73144555, + "learning_rate": 3.426006386722726e-06, + "loss": 0.75393325, + "num_input_tokens_seen": 48180290, + "step": 2245, + "time_per_iteration": 2.992250680923462 + }, + { + "auxiliary_loss_clip": 0.01236187, + "auxiliary_loss_mlp": 0.01031364, + "balance_loss_clip": 1.06067121, + "balance_loss_mlp": 1.02205992, + "epoch": 0.2700655323753983, + "flos": 18078150441600.0, + "grad_norm": 2.14010540593459, + "language_loss": 0.9237746, + "learning_rate": 3.4254600942028914e-06, + "loss": 0.94645011, + "num_input_tokens_seen": 48198165, + "step": 2246, + "time_per_iteration": 2.7373545169830322 + }, + { + "auxiliary_loss_clip": 0.01222505, + "auxiliary_loss_mlp": 0.01032226, + "balance_loss_clip": 1.06106389, + "balance_loss_mlp": 1.02347052, + "epoch": 0.2701857752660374, + "flos": 18186636493440.0, + "grad_norm": 1.9564595619171876, + "language_loss": 0.82576668, + "learning_rate": 3.424913585441893e-06, + "loss": 0.84831405, + "num_input_tokens_seen": 48216000, + "step": 2247, + "time_per_iteration": 2.691000461578369 + }, + { + "auxiliary_loss_clip": 0.01210871, + "auxiliary_loss_mlp": 0.01032748, + "balance_loss_clip": 1.05920315, + "balance_loss_mlp": 1.02375412, + "epoch": 0.2703060181566765, + "flos": 16319496648960.0, + "grad_norm": 1.889941891612269, + "language_loss": 0.8720066, + "learning_rate": 3.4243668605226374e-06, + "loss": 0.89444274, + "num_input_tokens_seen": 48233025, + "step": 2248, + "time_per_iteration": 2.566197633743286 + }, + { + "auxiliary_loss_clip": 0.01231121, + "auxiliary_loss_mlp": 0.01033006, + "balance_loss_clip": 1.05776298, + "balance_loss_mlp": 1.02342772, + "epoch": 0.2704262610473156, + "flos": 19572357329280.0, + "grad_norm": 2.1012579998695, + "language_loss": 0.8278718, + "learning_rate": 3.423819919528061e-06, + "loss": 0.85051298, + "num_input_tokens_seen": 48251110, + "step": 2249, + "time_per_iteration": 2.780592203140259 + }, + { + "auxiliary_loss_clip": 0.01241265, + "auxiliary_loss_mlp": 0.01025653, + "balance_loss_clip": 1.05294478, + "balance_loss_mlp": 1.01697397, + "epoch": 0.27054650393795465, + "flos": 20740746925440.0, + "grad_norm": 1.8792710575876819, + "language_loss": 0.78463805, + "learning_rate": 3.4232727625411355e-06, + "loss": 0.80730718, + "num_input_tokens_seen": 48270215, + "step": 2250, + "time_per_iteration": 3.6373162269592285 + }, + { + "auxiliary_loss_clip": 0.01239689, + "auxiliary_loss_mlp": 0.01032802, + "balance_loss_clip": 1.05168581, + "balance_loss_mlp": 1.02411771, + "epoch": 0.27066674682859376, + "flos": 18658322916480.0, + "grad_norm": 1.644676930608436, + "language_loss": 0.86448383, + "learning_rate": 3.4227253896448626e-06, + "loss": 0.8872087, + "num_input_tokens_seen": 48288075, + "step": 2251, + "time_per_iteration": 3.715186357498169 + }, + { + "auxiliary_loss_clip": 0.01203489, + "auxiliary_loss_mlp": 0.01030339, + "balance_loss_clip": 1.05930865, + "balance_loss_mlp": 1.02196455, + "epoch": 0.2707869897192329, + "flos": 23002759958400.0, + "grad_norm": 2.8821365116848656, + "language_loss": 0.82100207, + "learning_rate": 3.42217780092228e-06, + "loss": 0.8433404, + "num_input_tokens_seen": 48306415, + "step": 2252, + "time_per_iteration": 3.692326545715332 + }, + { + "auxiliary_loss_clip": 0.01144674, + "auxiliary_loss_mlp": 0.01001986, + "balance_loss_clip": 1.02994144, + "balance_loss_mlp": 1.00015068, + "epoch": 0.27090723260987193, + "flos": 58323240293760.0, + "grad_norm": 0.7907987268918714, + "language_loss": 0.60326415, + "learning_rate": 3.421629996456456e-06, + "loss": 0.62473071, + "num_input_tokens_seen": 48365035, + "step": 2253, + "time_per_iteration": 3.2774696350097656 + }, + { + "auxiliary_loss_clip": 0.01209976, + "auxiliary_loss_mlp": 0.01029759, + "balance_loss_clip": 1.05851531, + "balance_loss_mlp": 1.02076435, + "epoch": 0.27102747550051104, + "flos": 11984540797440.0, + "grad_norm": 2.7614667327872495, + "language_loss": 0.82766324, + "learning_rate": 3.421081976330491e-06, + "loss": 0.85006058, + "num_input_tokens_seen": 48383550, + "step": 2254, + "time_per_iteration": 2.6459879875183105 + }, + { + "auxiliary_loss_clip": 0.0121845, + "auxiliary_loss_mlp": 0.01029651, + "balance_loss_clip": 1.05763245, + "balance_loss_mlp": 1.02136564, + "epoch": 0.27114771839115015, + "flos": 19900401264000.0, + "grad_norm": 1.8530566495870446, + "language_loss": 0.87811887, + "learning_rate": 3.4205337406275207e-06, + "loss": 0.9005999, + "num_input_tokens_seen": 48403670, + "step": 2255, + "time_per_iteration": 2.7682044506073 + }, + { + "auxiliary_loss_clip": 0.01204802, + "auxiliary_loss_mlp": 0.0103388, + "balance_loss_clip": 1.06001782, + "balance_loss_mlp": 1.02530289, + "epoch": 0.2712679612817892, + "flos": 18331966920960.0, + "grad_norm": 2.6975037343653834, + "language_loss": 0.75323999, + "learning_rate": 3.4199852894307114e-06, + "loss": 0.77562678, + "num_input_tokens_seen": 48420420, + "step": 2256, + "time_per_iteration": 3.5066821575164795 + }, + { + "auxiliary_loss_clip": 0.0124915, + "auxiliary_loss_mlp": 0.01037833, + "balance_loss_clip": 1.05493104, + "balance_loss_mlp": 1.02915478, + "epoch": 0.2713882041724283, + "flos": 24460302038400.0, + "grad_norm": 2.0941619096689994, + "language_loss": 0.78465557, + "learning_rate": 3.419436622823262e-06, + "loss": 0.8075254, + "num_input_tokens_seen": 48441140, + "step": 2257, + "time_per_iteration": 2.8758254051208496 + }, + { + "auxiliary_loss_clip": 0.01218794, + "auxiliary_loss_mlp": 0.01033135, + "balance_loss_clip": 1.05922818, + "balance_loss_mlp": 1.02464128, + "epoch": 0.27150844706306737, + "flos": 23039317025280.0, + "grad_norm": 1.5660919746839326, + "language_loss": 0.7419343, + "learning_rate": 3.4188877408884063e-06, + "loss": 0.76445353, + "num_input_tokens_seen": 48461845, + "step": 2258, + "time_per_iteration": 2.674694061279297 + }, + { + "auxiliary_loss_clip": 0.01211518, + "auxiliary_loss_mlp": 0.01035513, + "balance_loss_clip": 1.05341387, + "balance_loss_mlp": 1.02688813, + "epoch": 0.2716286899537065, + "flos": 22563644192640.0, + "grad_norm": 2.537541660354262, + "language_loss": 0.65613341, + "learning_rate": 3.4183386437094088e-06, + "loss": 0.67860377, + "num_input_tokens_seen": 48478510, + "step": 2259, + "time_per_iteration": 2.7007031440734863 + }, + { + "auxiliary_loss_clip": 0.01221728, + "auxiliary_loss_mlp": 0.01026546, + "balance_loss_clip": 1.05600893, + "balance_loss_mlp": 1.01770639, + "epoch": 0.2717489328443456, + "flos": 13115044523520.0, + "grad_norm": 2.653562272051674, + "language_loss": 0.82481122, + "learning_rate": 3.417789331369565e-06, + "loss": 0.84729397, + "num_input_tokens_seen": 48494300, + "step": 2260, + "time_per_iteration": 2.6517090797424316 + }, + { + "auxiliary_loss_clip": 0.01211384, + "auxiliary_loss_mlp": 0.01038033, + "balance_loss_clip": 1.06348968, + "balance_loss_mlp": 1.02886009, + "epoch": 0.27186917573498465, + "flos": 29278688060160.0, + "grad_norm": 1.9343294405934968, + "language_loss": 0.91107309, + "learning_rate": 3.4172398039522088e-06, + "loss": 0.93356729, + "num_input_tokens_seen": 48515585, + "step": 2261, + "time_per_iteration": 2.7302005290985107 + }, + { + "auxiliary_loss_clip": 0.01213039, + "auxiliary_loss_mlp": 0.01028317, + "balance_loss_clip": 1.05954683, + "balance_loss_mlp": 1.01858318, + "epoch": 0.27198941862562376, + "flos": 26032220000640.0, + "grad_norm": 1.705085727206724, + "language_loss": 0.79808784, + "learning_rate": 3.4166900615407e-06, + "loss": 0.82050145, + "num_input_tokens_seen": 48533500, + "step": 2262, + "time_per_iteration": 2.729484796524048 + }, + { + "auxiliary_loss_clip": 0.01214724, + "auxiliary_loss_mlp": 0.01031638, + "balance_loss_clip": 1.06161666, + "balance_loss_mlp": 1.02270877, + "epoch": 0.27210966151626287, + "flos": 32780983760640.0, + "grad_norm": 2.1051953476153367, + "language_loss": 0.75395441, + "learning_rate": 3.416140104218436e-06, + "loss": 0.77641809, + "num_input_tokens_seen": 48552865, + "step": 2263, + "time_per_iteration": 2.795398473739624 + }, + { + "auxiliary_loss_clip": 0.01121429, + "auxiliary_loss_mlp": 0.01375857, + "balance_loss_clip": 1.0249697, + "balance_loss_mlp": 1.00003326, + "epoch": 0.2722299044069019, + "flos": 65471043219840.0, + "grad_norm": 0.8385382019634542, + "language_loss": 0.69673157, + "learning_rate": 3.4155899320688437e-06, + "loss": 0.72170442, + "num_input_tokens_seen": 48618940, + "step": 2264, + "time_per_iteration": 3.3482749462127686 + }, + { + "auxiliary_loss_clip": 0.01250598, + "auxiliary_loss_mlp": 0.01031485, + "balance_loss_clip": 1.05550265, + "balance_loss_mlp": 1.02197766, + "epoch": 0.27235014729754103, + "flos": 15334143782400.0, + "grad_norm": 2.1126238200059713, + "language_loss": 0.73941195, + "learning_rate": 3.415039545175384e-06, + "loss": 0.76223278, + "num_input_tokens_seen": 48634665, + "step": 2265, + "time_per_iteration": 2.7339577674865723 + }, + { + "auxiliary_loss_clip": 0.01217331, + "auxiliary_loss_mlp": 0.01028393, + "balance_loss_clip": 1.06060982, + "balance_loss_mlp": 1.01920819, + "epoch": 0.27247039018818014, + "flos": 21872363973120.0, + "grad_norm": 2.0759934454698272, + "language_loss": 0.6512115, + "learning_rate": 3.414488943621551e-06, + "loss": 0.67366868, + "num_input_tokens_seen": 48653330, + "step": 2266, + "time_per_iteration": 2.6926589012145996 + }, + { + "auxiliary_loss_clip": 0.01211722, + "auxiliary_loss_mlp": 0.0103207, + "balance_loss_clip": 1.05998421, + "balance_loss_mlp": 1.02334356, + "epoch": 0.2725906330788192, + "flos": 18695490514560.0, + "grad_norm": 1.816087192598771, + "language_loss": 0.73705041, + "learning_rate": 3.41393812749087e-06, + "loss": 0.7594884, + "num_input_tokens_seen": 48671375, + "step": 2267, + "time_per_iteration": 2.6714141368865967 + }, + { + "auxiliary_loss_clip": 0.01218992, + "auxiliary_loss_mlp": 0.01026878, + "balance_loss_clip": 1.05876946, + "balance_loss_mlp": 1.0179497, + "epoch": 0.2727108759694583, + "flos": 17886099398400.0, + "grad_norm": 4.579735931280047, + "language_loss": 0.71678078, + "learning_rate": 3.4133870968668984e-06, + "loss": 0.73923945, + "num_input_tokens_seen": 48686175, + "step": 2268, + "time_per_iteration": 2.8589084148406982 + }, + { + "auxiliary_loss_clip": 0.01222122, + "auxiliary_loss_mlp": 0.01032589, + "balance_loss_clip": 1.05753696, + "balance_loss_mlp": 1.02348709, + "epoch": 0.2728311188600974, + "flos": 24461666755200.0, + "grad_norm": 3.0728424886742873, + "language_loss": 0.78489798, + "learning_rate": 3.412835851833229e-06, + "loss": 0.80744505, + "num_input_tokens_seen": 48708370, + "step": 2269, + "time_per_iteration": 2.764896869659424 + }, + { + "auxiliary_loss_clip": 0.01213012, + "auxiliary_loss_mlp": 0.01031864, + "balance_loss_clip": 1.06163502, + "balance_loss_mlp": 1.02315569, + "epoch": 0.2729513617507365, + "flos": 30993314757120.0, + "grad_norm": 1.7517853009206605, + "language_loss": 0.77975738, + "learning_rate": 3.4122843924734834e-06, + "loss": 0.80220616, + "num_input_tokens_seen": 48730670, + "step": 2270, + "time_per_iteration": 2.692556858062744 + }, + { + "auxiliary_loss_clip": 0.01216711, + "auxiliary_loss_mlp": 0.01035399, + "balance_loss_clip": 1.0555048, + "balance_loss_mlp": 1.02565324, + "epoch": 0.2730716046413756, + "flos": 19094637421440.0, + "grad_norm": 1.7409423146137089, + "language_loss": 0.87587607, + "learning_rate": 3.411732718871319e-06, + "loss": 0.89839721, + "num_input_tokens_seen": 48746510, + "step": 2271, + "time_per_iteration": 2.6947543621063232 + }, + { + "auxiliary_loss_clip": 0.01204161, + "auxiliary_loss_mlp": 0.01034722, + "balance_loss_clip": 1.06276095, + "balance_loss_mlp": 1.02658594, + "epoch": 0.27319184753201464, + "flos": 26944566474240.0, + "grad_norm": 1.5629743663411957, + "language_loss": 0.78659356, + "learning_rate": 3.4111808311104227e-06, + "loss": 0.80898237, + "num_input_tokens_seen": 48768825, + "step": 2272, + "time_per_iteration": 2.6975371837615967 + }, + { + "auxiliary_loss_clip": 0.01227769, + "auxiliary_loss_mlp": 0.01031691, + "balance_loss_clip": 1.05643463, + "balance_loss_mlp": 1.02204728, + "epoch": 0.27331209042265375, + "flos": 31759828012800.0, + "grad_norm": 1.7252150974139067, + "language_loss": 0.69475406, + "learning_rate": 3.410628729274517e-06, + "loss": 0.71734869, + "num_input_tokens_seen": 48790345, + "step": 2273, + "time_per_iteration": 2.935181140899658 + }, + { + "auxiliary_loss_clip": 0.01222022, + "auxiliary_loss_mlp": 0.01379945, + "balance_loss_clip": 1.0590049, + "balance_loss_mlp": 1.00003433, + "epoch": 0.27343233331329286, + "flos": 25739081107200.0, + "grad_norm": 1.8185359233623375, + "language_loss": 0.82533401, + "learning_rate": 3.4100764134473546e-06, + "loss": 0.85135365, + "num_input_tokens_seen": 48809630, + "step": 2274, + "time_per_iteration": 2.703814744949341 + }, + { + "auxiliary_loss_clip": 0.01204103, + "auxiliary_loss_mlp": 0.01035657, + "balance_loss_clip": 1.06128383, + "balance_loss_mlp": 1.02709222, + "epoch": 0.2735525762039319, + "flos": 24389414547840.0, + "grad_norm": 3.6128694891857367, + "language_loss": 0.8509599, + "learning_rate": 3.4095238837127215e-06, + "loss": 0.87335747, + "num_input_tokens_seen": 48828770, + "step": 2275, + "time_per_iteration": 2.6719319820404053 + }, + { + "auxiliary_loss_clip": 0.01224556, + "auxiliary_loss_mlp": 0.01033528, + "balance_loss_clip": 1.0545752, + "balance_loss_mlp": 1.02438426, + "epoch": 0.27367281909457103, + "flos": 14465357527680.0, + "grad_norm": 1.9498895731796055, + "language_loss": 0.7935527, + "learning_rate": 3.4089711401544355e-06, + "loss": 0.81613356, + "num_input_tokens_seen": 48846365, + "step": 2276, + "time_per_iteration": 3.7903552055358887 + }, + { + "auxiliary_loss_clip": 0.01209978, + "auxiliary_loss_mlp": 0.01031448, + "balance_loss_clip": 1.05546093, + "balance_loss_mlp": 1.02280545, + "epoch": 0.27379306198521014, + "flos": 23476996247040.0, + "grad_norm": 2.41488385653065, + "language_loss": 0.68025041, + "learning_rate": 3.4084181828563486e-06, + "loss": 0.70266461, + "num_input_tokens_seen": 48863085, + "step": 2277, + "time_per_iteration": 3.5428640842437744 + }, + { + "auxiliary_loss_clip": 0.01234585, + "auxiliary_loss_mlp": 0.01029527, + "balance_loss_clip": 1.05323541, + "balance_loss_mlp": 1.02033615, + "epoch": 0.2739133048758492, + "flos": 17458152762240.0, + "grad_norm": 1.727694624378308, + "language_loss": 0.70483822, + "learning_rate": 3.4078650119023428e-06, + "loss": 0.72747934, + "num_input_tokens_seen": 48881400, + "step": 2278, + "time_per_iteration": 3.6358816623687744 + }, + { + "auxiliary_loss_clip": 0.01244938, + "auxiliary_loss_mlp": 0.01031137, + "balance_loss_clip": 1.04917884, + "balance_loss_mlp": 1.02239323, + "epoch": 0.2740335477664883, + "flos": 19273113123840.0, + "grad_norm": 2.133537269361673, + "language_loss": 0.74310267, + "learning_rate": 3.4073116273763337e-06, + "loss": 0.76586342, + "num_input_tokens_seen": 48895845, + "step": 2279, + "time_per_iteration": 2.7917299270629883 + }, + { + "auxiliary_loss_clip": 0.01224334, + "auxiliary_loss_mlp": 0.01030719, + "balance_loss_clip": 1.05615163, + "balance_loss_mlp": 1.0211103, + "epoch": 0.2741537906571274, + "flos": 26104723603200.0, + "grad_norm": 2.246331476632679, + "language_loss": 0.8132956, + "learning_rate": 3.40675802936227e-06, + "loss": 0.83584613, + "num_input_tokens_seen": 48916630, + "step": 2280, + "time_per_iteration": 2.6846659183502197 + }, + { + "auxiliary_loss_clip": 0.0121158, + "auxiliary_loss_mlp": 0.01027936, + "balance_loss_clip": 1.0558033, + "balance_loss_mlp": 1.01871467, + "epoch": 0.27427403354776647, + "flos": 34164190644480.0, + "grad_norm": 1.9809046668136323, + "language_loss": 0.71751547, + "learning_rate": 3.4062042179441318e-06, + "loss": 0.73991066, + "num_input_tokens_seen": 48937100, + "step": 2281, + "time_per_iteration": 2.8321096897125244 + }, + { + "auxiliary_loss_clip": 0.01208001, + "auxiliary_loss_mlp": 0.0102946, + "balance_loss_clip": 1.05820894, + "balance_loss_mlp": 1.02081156, + "epoch": 0.2743942764384056, + "flos": 18766988536320.0, + "grad_norm": 1.8778471638758545, + "language_loss": 0.80292654, + "learning_rate": 3.4056501932059314e-06, + "loss": 0.82530117, + "num_input_tokens_seen": 48955175, + "step": 2282, + "time_per_iteration": 3.5506722927093506 + }, + { + "auxiliary_loss_clip": 0.01103221, + "auxiliary_loss_mlp": 0.01007706, + "balance_loss_clip": 1.02854335, + "balance_loss_mlp": 1.00579917, + "epoch": 0.2745145193290447, + "flos": 64904048058240.0, + "grad_norm": 0.7857021920005112, + "language_loss": 0.58154351, + "learning_rate": 3.405095955231715e-06, + "loss": 0.60265279, + "num_input_tokens_seen": 49006830, + "step": 2283, + "time_per_iteration": 3.156224012374878 + }, + { + "auxiliary_loss_clip": 0.01215738, + "auxiliary_loss_mlp": 0.0102544, + "balance_loss_clip": 1.05718148, + "balance_loss_mlp": 1.01671994, + "epoch": 0.27463476221968375, + "flos": 16136926796160.0, + "grad_norm": 2.473544597047052, + "language_loss": 0.94245934, + "learning_rate": 3.4045415041055585e-06, + "loss": 0.96487117, + "num_input_tokens_seen": 49022470, + "step": 2284, + "time_per_iteration": 2.646986484527588 + }, + { + "auxiliary_loss_clip": 0.01223965, + "auxiliary_loss_mlp": 0.01027971, + "balance_loss_clip": 1.05582058, + "balance_loss_mlp": 1.0184226, + "epoch": 0.27475500511032286, + "flos": 10376712213120.0, + "grad_norm": 2.157794048412802, + "language_loss": 0.78513455, + "learning_rate": 3.4039868399115728e-06, + "loss": 0.8076539, + "num_input_tokens_seen": 49037110, + "step": 2285, + "time_per_iteration": 2.656999111175537 + }, + { + "auxiliary_loss_clip": 0.01252005, + "auxiliary_loss_mlp": 0.0103784, + "balance_loss_clip": 1.05569315, + "balance_loss_mlp": 1.02907836, + "epoch": 0.27487524800096197, + "flos": 17311062568320.0, + "grad_norm": 1.738597873158568, + "language_loss": 0.80335307, + "learning_rate": 3.4034319627339003e-06, + "loss": 0.82625157, + "num_input_tokens_seen": 49053975, + "step": 2286, + "time_per_iteration": 2.7852394580841064 + }, + { + "auxiliary_loss_clip": 0.01221654, + "auxiliary_loss_mlp": 0.0103286, + "balance_loss_clip": 1.05759156, + "balance_loss_mlp": 1.02408671, + "epoch": 0.274995490891601, + "flos": 27120205002240.0, + "grad_norm": 2.6124741685747352, + "language_loss": 0.69467962, + "learning_rate": 3.402876872656715e-06, + "loss": 0.71722472, + "num_input_tokens_seen": 49072295, + "step": 2287, + "time_per_iteration": 2.7490487098693848 + }, + { + "auxiliary_loss_clip": 0.01221528, + "auxiliary_loss_mlp": 0.01031019, + "balance_loss_clip": 1.05983615, + "balance_loss_mlp": 1.02193499, + "epoch": 0.27511573378224013, + "flos": 23436093634560.0, + "grad_norm": 1.9373289538389182, + "language_loss": 0.89710641, + "learning_rate": 3.402321569764223e-06, + "loss": 0.91963184, + "num_input_tokens_seen": 49091600, + "step": 2288, + "time_per_iteration": 2.763437271118164 + }, + { + "auxiliary_loss_clip": 0.01239566, + "auxiliary_loss_mlp": 0.0138017, + "balance_loss_clip": 1.05405807, + "balance_loss_mlp": 1.00014329, + "epoch": 0.2752359766728792, + "flos": 16722019434240.0, + "grad_norm": 1.8201216090027366, + "language_loss": 0.83747292, + "learning_rate": 3.4017660541406635e-06, + "loss": 0.86367023, + "num_input_tokens_seen": 49107665, + "step": 2289, + "time_per_iteration": 2.8123505115509033 + }, + { + "auxiliary_loss_clip": 0.01226729, + "auxiliary_loss_mlp": 0.0102763, + "balance_loss_clip": 1.05585432, + "balance_loss_mlp": 1.01848102, + "epoch": 0.2753562195635183, + "flos": 25297738698240.0, + "grad_norm": 1.736765377686321, + "language_loss": 0.74167013, + "learning_rate": 3.4012103258703092e-06, + "loss": 0.76421368, + "num_input_tokens_seen": 49126420, + "step": 2290, + "time_per_iteration": 2.830854654312134 + }, + { + "auxiliary_loss_clip": 0.01232275, + "auxiliary_loss_mlp": 0.01035829, + "balance_loss_clip": 1.05748177, + "balance_loss_mlp": 1.02726936, + "epoch": 0.2754764624541574, + "flos": 27338972785920.0, + "grad_norm": 1.9121145424710027, + "language_loss": 0.82983291, + "learning_rate": 3.4006543850374616e-06, + "loss": 0.85251391, + "num_input_tokens_seen": 49141470, + "step": 2291, + "time_per_iteration": 2.7439889907836914 + }, + { + "auxiliary_loss_clip": 0.01215449, + "auxiliary_loss_mlp": 0.0103096, + "balance_loss_clip": 1.05887127, + "balance_loss_mlp": 1.0219003, + "epoch": 0.27559670534479647, + "flos": 17238379397760.0, + "grad_norm": 1.8489505578997543, + "language_loss": 0.74892044, + "learning_rate": 3.400098231726458e-06, + "loss": 0.77138448, + "num_input_tokens_seen": 49158570, + "step": 2292, + "time_per_iteration": 2.768815755844116 + }, + { + "auxiliary_loss_clip": 0.01232309, + "auxiliary_loss_mlp": 0.01031491, + "balance_loss_clip": 1.05343938, + "balance_loss_mlp": 1.02199554, + "epoch": 0.2757169482354356, + "flos": 21939085486080.0, + "grad_norm": 1.7462585414441902, + "language_loss": 0.86774313, + "learning_rate": 3.3995418660216657e-06, + "loss": 0.89038116, + "num_input_tokens_seen": 49176025, + "step": 2293, + "time_per_iteration": 2.840731143951416 + }, + { + "auxiliary_loss_clip": 0.01210843, + "auxiliary_loss_mlp": 0.0103411, + "balance_loss_clip": 1.06187153, + "balance_loss_mlp": 1.02382255, + "epoch": 0.2758371911260747, + "flos": 20850669521280.0, + "grad_norm": 2.955827654208705, + "language_loss": 0.80405891, + "learning_rate": 3.3989852880074848e-06, + "loss": 0.8265084, + "num_input_tokens_seen": 49197455, + "step": 2294, + "time_per_iteration": 2.810312271118164 + }, + { + "auxiliary_loss_clip": 0.01121558, + "auxiliary_loss_mlp": 0.01007616, + "balance_loss_clip": 1.02798271, + "balance_loss_mlp": 1.00555348, + "epoch": 0.27595743401671374, + "flos": 69269063592960.0, + "grad_norm": 0.7965437128027684, + "language_loss": 0.6067785, + "learning_rate": 3.398428497768348e-06, + "loss": 0.62807024, + "num_input_tokens_seen": 49262625, + "step": 2295, + "time_per_iteration": 3.37104868888855 + }, + { + "auxiliary_loss_clip": 0.01234196, + "auxiliary_loss_mlp": 0.01028573, + "balance_loss_clip": 1.05116773, + "balance_loss_mlp": 1.01987684, + "epoch": 0.27607767690735285, + "flos": 21215019127680.0, + "grad_norm": 1.7625445225951628, + "language_loss": 0.71952093, + "learning_rate": 3.3978714953887205e-06, + "loss": 0.74214864, + "num_input_tokens_seen": 49282380, + "step": 2296, + "time_per_iteration": 2.69154691696167 + }, + { + "auxiliary_loss_clip": 0.01221028, + "auxiliary_loss_mlp": 0.01030924, + "balance_loss_clip": 1.04914832, + "balance_loss_mlp": 1.02182209, + "epoch": 0.27619791979799196, + "flos": 24825334003200.0, + "grad_norm": 1.8252634828195549, + "language_loss": 0.85923374, + "learning_rate": 3.397314280953098e-06, + "loss": 0.88175321, + "num_input_tokens_seen": 49303205, + "step": 2297, + "time_per_iteration": 2.804431676864624 + }, + { + "auxiliary_loss_clip": 0.01213656, + "auxiliary_loss_mlp": 0.01036771, + "balance_loss_clip": 1.05629969, + "balance_loss_mlp": 1.02786624, + "epoch": 0.276318162688631, + "flos": 24753548672640.0, + "grad_norm": 2.020357335482745, + "language_loss": 0.80169976, + "learning_rate": 3.3967568545460108e-06, + "loss": 0.82420403, + "num_input_tokens_seen": 49322745, + "step": 2298, + "time_per_iteration": 2.753483295440674 + }, + { + "auxiliary_loss_clip": 0.01210587, + "auxiliary_loss_mlp": 0.01031775, + "balance_loss_clip": 1.05876243, + "balance_loss_mlp": 1.02265, + "epoch": 0.27643840557927013, + "flos": 18150007599360.0, + "grad_norm": 2.2602041346721076, + "language_loss": 0.80666828, + "learning_rate": 3.3961992162520185e-06, + "loss": 0.82909191, + "num_input_tokens_seen": 49341370, + "step": 2299, + "time_per_iteration": 2.7247157096862793 + }, + { + "auxiliary_loss_clip": 0.0121484, + "auxiliary_loss_mlp": 0.01032667, + "balance_loss_clip": 1.05935228, + "balance_loss_mlp": 1.02258146, + "epoch": 0.27655864846990924, + "flos": 24823933372800.0, + "grad_norm": 2.265945689655784, + "language_loss": 0.71892822, + "learning_rate": 3.3956413661557156e-06, + "loss": 0.74140334, + "num_input_tokens_seen": 49361545, + "step": 2300, + "time_per_iteration": 2.7083919048309326 + }, + { + "auxiliary_loss_clip": 0.01237327, + "auxiliary_loss_mlp": 0.01038034, + "balance_loss_clip": 1.05577087, + "balance_loss_mlp": 1.02922463, + "epoch": 0.2766788913605483, + "flos": 20266582464000.0, + "grad_norm": 2.902606956604757, + "language_loss": 0.65881652, + "learning_rate": 3.3950833043417273e-06, + "loss": 0.68157017, + "num_input_tokens_seen": 49379690, + "step": 2301, + "time_per_iteration": 2.7821900844573975 + }, + { + "auxiliary_loss_clip": 0.01216268, + "auxiliary_loss_mlp": 0.01031899, + "balance_loss_clip": 1.06039929, + "balance_loss_mlp": 1.02215338, + "epoch": 0.2767991342511874, + "flos": 21470272151040.0, + "grad_norm": 2.576787123353203, + "language_loss": 0.73837173, + "learning_rate": 3.3945250308947105e-06, + "loss": 0.76085341, + "num_input_tokens_seen": 49395995, + "step": 2302, + "time_per_iteration": 3.5615627765655518 + }, + { + "auxiliary_loss_clip": 0.01117075, + "auxiliary_loss_mlp": 0.01007383, + "balance_loss_clip": 1.02890849, + "balance_loss_mlp": 1.00512981, + "epoch": 0.2769193771418265, + "flos": 66002627571840.0, + "grad_norm": 1.2501680979281562, + "language_loss": 0.68377048, + "learning_rate": 3.3939665458993556e-06, + "loss": 0.70501506, + "num_input_tokens_seen": 49450415, + "step": 2303, + "time_per_iteration": 4.043716669082642 + }, + { + "auxiliary_loss_clip": 0.01233813, + "auxiliary_loss_mlp": 0.01034313, + "balance_loss_clip": 1.05251169, + "balance_loss_mlp": 1.0252527, + "epoch": 0.27703962003246557, + "flos": 20704441253760.0, + "grad_norm": 1.9716901452297064, + "language_loss": 0.76998949, + "learning_rate": 3.3934078494403843e-06, + "loss": 0.79267079, + "num_input_tokens_seen": 49469990, + "step": 2304, + "time_per_iteration": 3.6092042922973633 + }, + { + "auxiliary_loss_clip": 0.01250456, + "auxiliary_loss_mlp": 0.01380132, + "balance_loss_clip": 1.05164146, + "balance_loss_mlp": 1.0000968, + "epoch": 0.2771598629231047, + "flos": 22929897219840.0, + "grad_norm": 1.8472722693881725, + "language_loss": 0.80938268, + "learning_rate": 3.3928489416025495e-06, + "loss": 0.83568859, + "num_input_tokens_seen": 49490835, + "step": 2305, + "time_per_iteration": 3.011733055114746 + }, + { + "auxiliary_loss_clip": 0.01217372, + "auxiliary_loss_mlp": 0.01035416, + "balance_loss_clip": 1.05550885, + "balance_loss_mlp": 1.02574861, + "epoch": 0.27728010581374374, + "flos": 18369457741440.0, + "grad_norm": 2.4253355068363174, + "language_loss": 0.79026318, + "learning_rate": 3.392289822470638e-06, + "loss": 0.81279105, + "num_input_tokens_seen": 49508815, + "step": 2306, + "time_per_iteration": 3.1087934970855713 + }, + { + "auxiliary_loss_clip": 0.01218565, + "auxiliary_loss_mlp": 0.01034514, + "balance_loss_clip": 1.05501497, + "balance_loss_mlp": 1.02500725, + "epoch": 0.27740034870438285, + "flos": 19427637432960.0, + "grad_norm": 2.0772232339417838, + "language_loss": 0.76057863, + "learning_rate": 3.3917304921294674e-06, + "loss": 0.78310943, + "num_input_tokens_seen": 49526980, + "step": 2307, + "time_per_iteration": 2.6583311557769775 + }, + { + "auxiliary_loss_clip": 0.01212537, + "auxiliary_loss_mlp": 0.01029253, + "balance_loss_clip": 1.05589676, + "balance_loss_mlp": 1.01968682, + "epoch": 0.27752059159502196, + "flos": 21614776565760.0, + "grad_norm": 2.062515473977052, + "language_loss": 0.80698073, + "learning_rate": 3.3911709506638876e-06, + "loss": 0.82939863, + "num_input_tokens_seen": 49546290, + "step": 2308, + "time_per_iteration": 3.609266757965088 + }, + { + "auxiliary_loss_clip": 0.01216144, + "auxiliary_loss_mlp": 0.01380004, + "balance_loss_clip": 1.04968786, + "balance_loss_mlp": 1.00007725, + "epoch": 0.277640834485661, + "flos": 26608011016320.0, + "grad_norm": 2.571164402616193, + "language_loss": 0.80691469, + "learning_rate": 3.390611198158781e-06, + "loss": 0.83287621, + "num_input_tokens_seen": 49564165, + "step": 2309, + "time_per_iteration": 2.7433359622955322 + }, + { + "auxiliary_loss_clip": 0.01210571, + "auxiliary_loss_mlp": 0.01034241, + "balance_loss_clip": 1.06314695, + "balance_loss_mlp": 1.02514553, + "epoch": 0.2777610773763001, + "flos": 19492814661120.0, + "grad_norm": 2.0417909295064245, + "language_loss": 0.900958, + "learning_rate": 3.3900512346990612e-06, + "loss": 0.92340612, + "num_input_tokens_seen": 49580155, + "step": 2310, + "time_per_iteration": 2.635300397872925 + }, + { + "auxiliary_loss_clip": 0.01232496, + "auxiliary_loss_mlp": 0.01034435, + "balance_loss_clip": 1.04857814, + "balance_loss_mlp": 1.02525556, + "epoch": 0.27788132026693924, + "flos": 38290650001920.0, + "grad_norm": 1.868808929883222, + "language_loss": 0.65884972, + "learning_rate": 3.389491060369674e-06, + "loss": 0.68151903, + "num_input_tokens_seen": 49605830, + "step": 2311, + "time_per_iteration": 2.8728458881378174 + }, + { + "auxiliary_loss_clip": 0.01230094, + "auxiliary_loss_mlp": 0.01032419, + "balance_loss_clip": 1.05105543, + "balance_loss_mlp": 1.02318001, + "epoch": 0.2780015631575783, + "flos": 22382546797440.0, + "grad_norm": 2.031955788202339, + "language_loss": 0.89562213, + "learning_rate": 3.388930675255598e-06, + "loss": 0.91824722, + "num_input_tokens_seen": 49625680, + "step": 2312, + "time_per_iteration": 2.7731430530548096 + }, + { + "auxiliary_loss_clip": 0.01225655, + "auxiliary_loss_mlp": 0.0103115, + "balance_loss_clip": 1.05644155, + "balance_loss_mlp": 1.0212791, + "epoch": 0.2781218060482174, + "flos": 12203200840320.0, + "grad_norm": 2.242999477285709, + "language_loss": 0.79570079, + "learning_rate": 3.388370079441843e-06, + "loss": 0.81826878, + "num_input_tokens_seen": 49641195, + "step": 2313, + "time_per_iteration": 2.652651071548462 + }, + { + "auxiliary_loss_clip": 0.01230977, + "auxiliary_loss_mlp": 0.01029917, + "balance_loss_clip": 1.05810523, + "balance_loss_mlp": 1.0203805, + "epoch": 0.2782420489388565, + "flos": 18107632529280.0, + "grad_norm": 2.055029652350873, + "language_loss": 0.92519802, + "learning_rate": 3.3878092730134505e-06, + "loss": 0.94780695, + "num_input_tokens_seen": 49659180, + "step": 2314, + "time_per_iteration": 3.0184760093688965 + }, + { + "auxiliary_loss_clip": 0.01204641, + "auxiliary_loss_mlp": 0.01035946, + "balance_loss_clip": 1.05802119, + "balance_loss_mlp": 1.02624798, + "epoch": 0.27836229182949557, + "flos": 18514752255360.0, + "grad_norm": 1.4930426646923418, + "language_loss": 0.80579978, + "learning_rate": 3.3872482560554947e-06, + "loss": 0.82820565, + "num_input_tokens_seen": 49677955, + "step": 2315, + "time_per_iteration": 2.652627468109131 + }, + { + "auxiliary_loss_clip": 0.01116827, + "auxiliary_loss_mlp": 0.01002535, + "balance_loss_clip": 1.0299747, + "balance_loss_mlp": 1.00046086, + "epoch": 0.2784825347201347, + "flos": 67079230940160.0, + "grad_norm": 0.801790536329611, + "language_loss": 0.5693661, + "learning_rate": 3.386687028653082e-06, + "loss": 0.59055972, + "num_input_tokens_seen": 49740800, + "step": 2316, + "time_per_iteration": 3.199491024017334 + }, + { + "auxiliary_loss_clip": 0.01239183, + "auxiliary_loss_mlp": 0.01028667, + "balance_loss_clip": 1.05353165, + "balance_loss_mlp": 1.01929128, + "epoch": 0.2786027776107738, + "flos": 22631119891200.0, + "grad_norm": 5.442474292019069, + "language_loss": 0.84963435, + "learning_rate": 3.386125590891349e-06, + "loss": 0.87231284, + "num_input_tokens_seen": 49757675, + "step": 2317, + "time_per_iteration": 2.759061574935913 + }, + { + "auxiliary_loss_clip": 0.01205963, + "auxiliary_loss_mlp": 0.01032722, + "balance_loss_clip": 1.05194986, + "balance_loss_mlp": 1.02354288, + "epoch": 0.27872302050141284, + "flos": 15778826156160.0, + "grad_norm": 2.2481109805830877, + "language_loss": 0.83295691, + "learning_rate": 3.3855639428554657e-06, + "loss": 0.8553437, + "num_input_tokens_seen": 49775205, + "step": 2318, + "time_per_iteration": 2.607816219329834 + }, + { + "auxiliary_loss_clip": 0.01217425, + "auxiliary_loss_mlp": 0.01029058, + "balance_loss_clip": 1.05221891, + "balance_loss_mlp": 1.02038538, + "epoch": 0.27884326339205195, + "flos": 22126970551680.0, + "grad_norm": 1.8458070367580257, + "language_loss": 0.79919279, + "learning_rate": 3.385002084630635e-06, + "loss": 0.82165766, + "num_input_tokens_seen": 49794175, + "step": 2319, + "time_per_iteration": 2.699173927307129 + }, + { + "auxiliary_loss_clip": 0.01215707, + "auxiliary_loss_mlp": 0.01027724, + "balance_loss_clip": 1.05700541, + "balance_loss_mlp": 1.01851523, + "epoch": 0.278963506282691, + "flos": 20558715776640.0, + "grad_norm": 2.50216864949375, + "language_loss": 0.84952307, + "learning_rate": 3.384440016302088e-06, + "loss": 0.87195742, + "num_input_tokens_seen": 49812850, + "step": 2320, + "time_per_iteration": 2.6281192302703857 + }, + { + "auxiliary_loss_clip": 0.01209906, + "auxiliary_loss_mlp": 0.01025697, + "balance_loss_clip": 1.05648649, + "balance_loss_mlp": 1.01625609, + "epoch": 0.2790837491733301, + "flos": 21942928241280.0, + "grad_norm": 2.0700300450371523, + "language_loss": 0.62308902, + "learning_rate": 3.3838777379550923e-06, + "loss": 0.64544511, + "num_input_tokens_seen": 49832295, + "step": 2321, + "time_per_iteration": 2.671311140060425 + }, + { + "auxiliary_loss_clip": 0.01224704, + "auxiliary_loss_mlp": 0.01029402, + "balance_loss_clip": 1.05778408, + "balance_loss_mlp": 1.01947737, + "epoch": 0.27920399206396923, + "flos": 26286790665600.0, + "grad_norm": 2.1938047760100945, + "language_loss": 0.78425449, + "learning_rate": 3.383315249674944e-06, + "loss": 0.80679554, + "num_input_tokens_seen": 49850860, + "step": 2322, + "time_per_iteration": 2.7817273139953613 + }, + { + "auxiliary_loss_clip": 0.01232701, + "auxiliary_loss_mlp": 0.010398, + "balance_loss_clip": 1.05488849, + "balance_loss_mlp": 1.03094244, + "epoch": 0.2793242349546083, + "flos": 25400981364480.0, + "grad_norm": 2.012253950540485, + "language_loss": 0.86195892, + "learning_rate": 3.3827525515469715e-06, + "loss": 0.88468391, + "num_input_tokens_seen": 49865765, + "step": 2323, + "time_per_iteration": 2.700026512145996 + }, + { + "auxiliary_loss_clip": 0.01218561, + "auxiliary_loss_mlp": 0.01034172, + "balance_loss_clip": 1.05077755, + "balance_loss_mlp": 1.02383614, + "epoch": 0.2794444778452474, + "flos": 20850346298880.0, + "grad_norm": 2.2132914931704017, + "language_loss": 0.70464796, + "learning_rate": 3.3821896436565367e-06, + "loss": 0.7271753, + "num_input_tokens_seen": 49885425, + "step": 2324, + "time_per_iteration": 2.7599430084228516 + }, + { + "auxiliary_loss_clip": 0.01219304, + "auxiliary_loss_mlp": 0.01033645, + "balance_loss_clip": 1.06168199, + "balance_loss_mlp": 1.0239954, + "epoch": 0.2795647207358865, + "flos": 21576244250880.0, + "grad_norm": 1.6619666009899285, + "language_loss": 0.70320886, + "learning_rate": 3.381626526089032e-06, + "loss": 0.72573841, + "num_input_tokens_seen": 49904990, + "step": 2325, + "time_per_iteration": 2.672116279602051 + }, + { + "auxiliary_loss_clip": 0.01215602, + "auxiliary_loss_mlp": 0.01029183, + "balance_loss_clip": 1.05224729, + "balance_loss_mlp": 1.02007496, + "epoch": 0.27968496362652556, + "flos": 21471744608640.0, + "grad_norm": 2.458918904947792, + "language_loss": 0.78994811, + "learning_rate": 3.3810631989298815e-06, + "loss": 0.81239593, + "num_input_tokens_seen": 49924600, + "step": 2326, + "time_per_iteration": 2.8022398948669434 + }, + { + "auxiliary_loss_clip": 0.01244208, + "auxiliary_loss_mlp": 0.01031781, + "balance_loss_clip": 1.05392528, + "balance_loss_mlp": 1.02211928, + "epoch": 0.2798052065171647, + "flos": 23258695340160.0, + "grad_norm": 2.314491857076046, + "language_loss": 0.84251058, + "learning_rate": 3.3804996622645423e-06, + "loss": 0.8652705, + "num_input_tokens_seen": 49942600, + "step": 2327, + "time_per_iteration": 2.7431676387786865 + }, + { + "auxiliary_loss_clip": 0.01206133, + "auxiliary_loss_mlp": 0.01028042, + "balance_loss_clip": 1.06035304, + "balance_loss_mlp": 1.01861238, + "epoch": 0.2799254494078038, + "flos": 21539328048000.0, + "grad_norm": 1.7348493291203684, + "language_loss": 0.89593601, + "learning_rate": 3.3799359161785015e-06, + "loss": 0.91827774, + "num_input_tokens_seen": 49962250, + "step": 2328, + "time_per_iteration": 3.5567150115966797 + }, + { + "auxiliary_loss_clip": 0.01209009, + "auxiliary_loss_mlp": 0.01029771, + "balance_loss_clip": 1.05508351, + "balance_loss_mlp": 1.02031171, + "epoch": 0.28004569229844284, + "flos": 26393912000640.0, + "grad_norm": 3.38894702072747, + "language_loss": 0.85504252, + "learning_rate": 3.3793719607572798e-06, + "loss": 0.87743032, + "num_input_tokens_seen": 49983215, + "step": 2329, + "time_per_iteration": 4.485391616821289 + }, + { + "auxiliary_loss_clip": 0.01199088, + "auxiliary_loss_mlp": 0.01035566, + "balance_loss_clip": 1.0511626, + "balance_loss_mlp": 1.02676868, + "epoch": 0.28016593518908195, + "flos": 33547676584320.0, + "grad_norm": 2.6668071818570973, + "language_loss": 0.77228457, + "learning_rate": 3.378807796086428e-06, + "loss": 0.79463106, + "num_input_tokens_seen": 50006075, + "step": 2330, + "time_per_iteration": 2.748683214187622 + }, + { + "auxiliary_loss_clip": 0.01208275, + "auxiliary_loss_mlp": 0.01034407, + "balance_loss_clip": 1.06216943, + "balance_loss_mlp": 1.0246433, + "epoch": 0.28028617807972106, + "flos": 15340823712000.0, + "grad_norm": 1.9608700844504456, + "language_loss": 0.76916385, + "learning_rate": 3.37824342225153e-06, + "loss": 0.79159063, + "num_input_tokens_seen": 50022495, + "step": 2331, + "time_per_iteration": 2.6478493213653564 + }, + { + "auxiliary_loss_clip": 0.01238615, + "auxiliary_loss_mlp": 0.01033706, + "balance_loss_clip": 1.05326462, + "balance_loss_mlp": 1.02492046, + "epoch": 0.2804064209703601, + "flos": 25520277409920.0, + "grad_norm": 1.9920698930315375, + "language_loss": 0.77634323, + "learning_rate": 3.3776788393382006e-06, + "loss": 0.79906642, + "num_input_tokens_seen": 50041975, + "step": 2332, + "time_per_iteration": 2.7739858627319336 + }, + { + "auxiliary_loss_clip": 0.01206169, + "auxiliary_loss_mlp": 0.01028967, + "balance_loss_clip": 1.06021547, + "balance_loss_mlp": 1.01947832, + "epoch": 0.2805266638609992, + "flos": 29351766280320.0, + "grad_norm": 2.5062203220034607, + "language_loss": 0.76892459, + "learning_rate": 3.3771140474320872e-06, + "loss": 0.79127592, + "num_input_tokens_seen": 50061925, + "step": 2333, + "time_per_iteration": 2.7966437339782715 + }, + { + "auxiliary_loss_clip": 0.01236065, + "auxiliary_loss_mlp": 0.01032313, + "balance_loss_clip": 1.05457723, + "balance_loss_mlp": 1.02306831, + "epoch": 0.28064690675163834, + "flos": 21463735875840.0, + "grad_norm": 1.8968303039876893, + "language_loss": 0.792979, + "learning_rate": 3.3765490466188664e-06, + "loss": 0.81566274, + "num_input_tokens_seen": 50079325, + "step": 2334, + "time_per_iteration": 3.7056937217712402 + }, + { + "auxiliary_loss_clip": 0.0122733, + "auxiliary_loss_mlp": 0.01033843, + "balance_loss_clip": 1.05481958, + "balance_loss_mlp": 1.02424097, + "epoch": 0.2807671496422774, + "flos": 20995640812800.0, + "grad_norm": 2.4380540573199743, + "language_loss": 0.74002779, + "learning_rate": 3.3759838369842508e-06, + "loss": 0.76263952, + "num_input_tokens_seen": 50097400, + "step": 2335, + "time_per_iteration": 2.688199520111084 + }, + { + "auxiliary_loss_clip": 0.01229786, + "auxiliary_loss_mlp": 0.01033786, + "balance_loss_clip": 1.05478024, + "balance_loss_mlp": 1.02439213, + "epoch": 0.2808873925329165, + "flos": 21506577822720.0, + "grad_norm": 1.8216959050000416, + "language_loss": 0.73035097, + "learning_rate": 3.375418418613981e-06, + "loss": 0.75298667, + "num_input_tokens_seen": 50116425, + "step": 2336, + "time_per_iteration": 2.774513006210327 + }, + { + "auxiliary_loss_clip": 0.01221603, + "auxiliary_loss_mlp": 0.01033215, + "balance_loss_clip": 1.0565747, + "balance_loss_mlp": 1.02313614, + "epoch": 0.28100763542355556, + "flos": 16070815814400.0, + "grad_norm": 2.3719493451081495, + "language_loss": 0.83368206, + "learning_rate": 3.374852791593831e-06, + "loss": 0.8562302, + "num_input_tokens_seen": 50132625, + "step": 2337, + "time_per_iteration": 2.6544272899627686 + }, + { + "auxiliary_loss_clip": 0.01246957, + "auxiliary_loss_mlp": 0.01033365, + "balance_loss_clip": 1.05283165, + "balance_loss_mlp": 1.02277911, + "epoch": 0.28112787831419467, + "flos": 19062605468160.0, + "grad_norm": 14.790534690629919, + "language_loss": 0.54146677, + "learning_rate": 3.374286956009605e-06, + "loss": 0.56427002, + "num_input_tokens_seen": 50151190, + "step": 2338, + "time_per_iteration": 2.7874555587768555 + }, + { + "auxiliary_loss_clip": 0.01214227, + "auxiliary_loss_mlp": 0.01032963, + "balance_loss_clip": 1.06289148, + "balance_loss_mlp": 1.02370071, + "epoch": 0.2812481212048338, + "flos": 12823629482880.0, + "grad_norm": 4.968798997415987, + "language_loss": 0.75370783, + "learning_rate": 3.3737209119471405e-06, + "loss": 0.77617973, + "num_input_tokens_seen": 50167700, + "step": 2339, + "time_per_iteration": 2.6607003211975098 + }, + { + "auxiliary_loss_clip": 0.01220748, + "auxiliary_loss_mlp": 0.01038256, + "balance_loss_clip": 1.0597558, + "balance_loss_mlp": 1.02794993, + "epoch": 0.28136836409547283, + "flos": 15633064765440.0, + "grad_norm": 2.5508870925103544, + "language_loss": 0.64022887, + "learning_rate": 3.373154659492306e-06, + "loss": 0.66281897, + "num_input_tokens_seen": 50185840, + "step": 2340, + "time_per_iteration": 2.6232736110687256 + }, + { + "auxiliary_loss_clip": 0.01227434, + "auxiliary_loss_mlp": 0.01034117, + "balance_loss_clip": 1.05964303, + "balance_loss_mlp": 1.02438331, + "epoch": 0.28148860698611194, + "flos": 19933726106880.0, + "grad_norm": 2.15011983900978, + "language_loss": 0.85285336, + "learning_rate": 3.3725881987310016e-06, + "loss": 0.87546885, + "num_input_tokens_seen": 50203375, + "step": 2341, + "time_per_iteration": 2.678112268447876 + }, + { + "auxiliary_loss_clip": 0.01220238, + "auxiliary_loss_mlp": 0.01028043, + "balance_loss_clip": 1.05572534, + "balance_loss_mlp": 1.01883411, + "epoch": 0.28160884987675106, + "flos": 17457219008640.0, + "grad_norm": 2.8821654475169987, + "language_loss": 0.87307614, + "learning_rate": 3.372021529749159e-06, + "loss": 0.89555889, + "num_input_tokens_seen": 50222435, + "step": 2342, + "time_per_iteration": 2.6533751487731934 + }, + { + "auxiliary_loss_clip": 0.01248408, + "auxiliary_loss_mlp": 0.0103326, + "balance_loss_clip": 1.05446553, + "balance_loss_mlp": 1.02431917, + "epoch": 0.2817290927673901, + "flos": 16834743290880.0, + "grad_norm": 2.4822505186135975, + "language_loss": 0.92525351, + "learning_rate": 3.3714546526327405e-06, + "loss": 0.94807023, + "num_input_tokens_seen": 50240435, + "step": 2343, + "time_per_iteration": 2.7720916271209717 + }, + { + "auxiliary_loss_clip": 0.0123333, + "auxiliary_loss_mlp": 0.01028577, + "balance_loss_clip": 1.05486894, + "balance_loss_mlp": 1.01936841, + "epoch": 0.2818493356580292, + "flos": 15414081500160.0, + "grad_norm": 2.2367313023065063, + "language_loss": 0.87942624, + "learning_rate": 3.3708875674677423e-06, + "loss": 0.90204525, + "num_input_tokens_seen": 50258410, + "step": 2344, + "time_per_iteration": 2.7331817150115967 + }, + { + "auxiliary_loss_clip": 0.01231708, + "auxiliary_loss_mlp": 0.01034168, + "balance_loss_clip": 1.06037712, + "balance_loss_mlp": 1.02468467, + "epoch": 0.28196957854866833, + "flos": 20412451595520.0, + "grad_norm": 1.9851974524795906, + "language_loss": 0.83261144, + "learning_rate": 3.37032027434019e-06, + "loss": 0.85527015, + "num_input_tokens_seen": 50277930, + "step": 2345, + "time_per_iteration": 2.7090821266174316 + }, + { + "auxiliary_loss_clip": 0.01222809, + "auxiliary_loss_mlp": 0.01033868, + "balance_loss_clip": 1.05840433, + "balance_loss_mlp": 1.023592, + "epoch": 0.2820898214393074, + "flos": 19973120348160.0, + "grad_norm": 1.7506918683543233, + "language_loss": 0.83144259, + "learning_rate": 3.369752773336141e-06, + "loss": 0.85400939, + "num_input_tokens_seen": 50297410, + "step": 2346, + "time_per_iteration": 2.6668503284454346 + }, + { + "auxiliary_loss_clip": 0.01222455, + "auxiliary_loss_mlp": 0.01034745, + "balance_loss_clip": 1.0562675, + "balance_loss_mlp": 1.02463627, + "epoch": 0.2822100643299465, + "flos": 22528308188160.0, + "grad_norm": 2.4157597939876907, + "language_loss": 0.78226352, + "learning_rate": 3.3691850645416864e-06, + "loss": 0.80483544, + "num_input_tokens_seen": 50317120, + "step": 2347, + "time_per_iteration": 2.731656312942505 + }, + { + "auxiliary_loss_clip": 0.01218813, + "auxiliary_loss_mlp": 0.01035105, + "balance_loss_clip": 1.05950618, + "balance_loss_mlp": 1.0254972, + "epoch": 0.2823303072205856, + "flos": 11546682007680.0, + "grad_norm": 2.280426382993318, + "language_loss": 0.82997525, + "learning_rate": 3.368617148042945e-06, + "loss": 0.85251445, + "num_input_tokens_seen": 50334790, + "step": 2348, + "time_per_iteration": 2.59771466255188 + }, + { + "auxiliary_loss_clip": 0.01216243, + "auxiliary_loss_mlp": 0.01039825, + "balance_loss_clip": 1.05591202, + "balance_loss_mlp": 1.03051472, + "epoch": 0.28245055011122466, + "flos": 18259894281600.0, + "grad_norm": 1.6222683952262877, + "language_loss": 0.84284925, + "learning_rate": 3.368049023926071e-06, + "loss": 0.86540991, + "num_input_tokens_seen": 50353785, + "step": 2349, + "time_per_iteration": 2.667797327041626 + }, + { + "auxiliary_loss_clip": 0.0121594, + "auxiliary_loss_mlp": 0.01027954, + "balance_loss_clip": 1.06192625, + "balance_loss_mlp": 1.0189178, + "epoch": 0.2825707930018638, + "flos": 24608110504320.0, + "grad_norm": 1.7798233145776976, + "language_loss": 0.83439445, + "learning_rate": 3.3674806922772476e-06, + "loss": 0.85683346, + "num_input_tokens_seen": 50374670, + "step": 2350, + "time_per_iteration": 2.6699366569519043 + }, + { + "auxiliary_loss_clip": 0.01234071, + "auxiliary_loss_mlp": 0.01035102, + "balance_loss_clip": 1.05561435, + "balance_loss_mlp": 1.02588129, + "epoch": 0.28269103589250283, + "flos": 25226994862080.0, + "grad_norm": 2.2917625327982942, + "language_loss": 0.75038135, + "learning_rate": 3.3669121531826904e-06, + "loss": 0.77307314, + "num_input_tokens_seen": 50395650, + "step": 2351, + "time_per_iteration": 2.7907357215881348 + }, + { + "auxiliary_loss_clip": 0.01229386, + "auxiliary_loss_mlp": 0.01033811, + "balance_loss_clip": 1.06127357, + "balance_loss_mlp": 1.02408993, + "epoch": 0.28281127878314194, + "flos": 19281552819840.0, + "grad_norm": 2.771892648004035, + "language_loss": 0.82998246, + "learning_rate": 3.366343406728647e-06, + "loss": 0.8526144, + "num_input_tokens_seen": 50415100, + "step": 2352, + "time_per_iteration": 2.707109212875366 + }, + { + "auxiliary_loss_clip": 0.01206927, + "auxiliary_loss_mlp": 0.01030897, + "balance_loss_clip": 1.05695939, + "balance_loss_mlp": 1.02160454, + "epoch": 0.28293152167378105, + "flos": 23878405710720.0, + "grad_norm": 11.354858892523156, + "language_loss": 0.68170208, + "learning_rate": 3.3657744530013946e-06, + "loss": 0.70408034, + "num_input_tokens_seen": 50434335, + "step": 2353, + "time_per_iteration": 3.606046676635742 + }, + { + "auxiliary_loss_clip": 0.01219559, + "auxiliary_loss_mlp": 0.01032389, + "balance_loss_clip": 1.06044805, + "balance_loss_mlp": 1.02303708, + "epoch": 0.2830517645644201, + "flos": 43866965928960.0, + "grad_norm": 2.012528650258542, + "language_loss": 0.71485603, + "learning_rate": 3.3652052920872437e-06, + "loss": 0.7373755, + "num_input_tokens_seen": 50457200, + "step": 2354, + "time_per_iteration": 2.8588130474090576 + }, + { + "auxiliary_loss_clip": 0.01224866, + "auxiliary_loss_mlp": 0.01038762, + "balance_loss_clip": 1.05762744, + "balance_loss_mlp": 1.02895725, + "epoch": 0.2831720074550592, + "flos": 26651750803200.0, + "grad_norm": 2.134598054833611, + "language_loss": 0.85608864, + "learning_rate": 3.3646359240725355e-06, + "loss": 0.87872487, + "num_input_tokens_seen": 50476390, + "step": 2355, + "time_per_iteration": 3.6597249507904053 + }, + { + "auxiliary_loss_clip": 0.0120966, + "auxiliary_loss_mlp": 0.01380561, + "balance_loss_clip": 1.05926776, + "balance_loss_mlp": 1.00017965, + "epoch": 0.2832922503456983, + "flos": 31029979564800.0, + "grad_norm": 2.415371877684896, + "language_loss": 0.67181802, + "learning_rate": 3.364066349043643e-06, + "loss": 0.69772029, + "num_input_tokens_seen": 50497595, + "step": 2356, + "time_per_iteration": 2.717766761779785 + }, + { + "auxiliary_loss_clip": 0.01222658, + "auxiliary_loss_mlp": 0.01029693, + "balance_loss_clip": 1.05970275, + "balance_loss_mlp": 1.02088392, + "epoch": 0.2834124932363374, + "flos": 20405699838720.0, + "grad_norm": 2.423275798308675, + "language_loss": 0.82090718, + "learning_rate": 3.363496567086969e-06, + "loss": 0.84343064, + "num_input_tokens_seen": 50514690, + "step": 2357, + "time_per_iteration": 2.6528687477111816 + }, + { + "auxiliary_loss_clip": 0.01205122, + "auxiliary_loss_mlp": 0.01032488, + "balance_loss_clip": 1.05982435, + "balance_loss_mlp": 1.02307081, + "epoch": 0.2835327361269765, + "flos": 39384848056320.0, + "grad_norm": 1.9102232584170982, + "language_loss": 0.75752068, + "learning_rate": 3.3629265782889506e-06, + "loss": 0.7798968, + "num_input_tokens_seen": 50536515, + "step": 2358, + "time_per_iteration": 2.7209692001342773 + }, + { + "auxiliary_loss_clip": 0.0122244, + "auxiliary_loss_mlp": 0.01031988, + "balance_loss_clip": 1.05173886, + "balance_loss_mlp": 1.0219444, + "epoch": 0.2836529790176156, + "flos": 30261598801920.0, + "grad_norm": 2.3621820140610614, + "language_loss": 0.71806669, + "learning_rate": 3.362356382736054e-06, + "loss": 0.74061096, + "num_input_tokens_seen": 50557120, + "step": 2359, + "time_per_iteration": 2.781256914138794 + }, + { + "auxiliary_loss_clip": 0.01230909, + "auxiliary_loss_mlp": 0.01029211, + "balance_loss_clip": 1.05264354, + "balance_loss_mlp": 1.01975727, + "epoch": 0.28377322190825466, + "flos": 12677796264960.0, + "grad_norm": 2.040746365668196, + "language_loss": 0.9083901, + "learning_rate": 3.361785980514777e-06, + "loss": 0.93099129, + "num_input_tokens_seen": 50573320, + "step": 2360, + "time_per_iteration": 3.5499138832092285 + }, + { + "auxiliary_loss_clip": 0.01240559, + "auxiliary_loss_mlp": 0.01035441, + "balance_loss_clip": 1.05295813, + "balance_loss_mlp": 1.0256716, + "epoch": 0.28389346479889377, + "flos": 18296666830080.0, + "grad_norm": 1.6555045836858444, + "language_loss": 0.76424611, + "learning_rate": 3.361215371711649e-06, + "loss": 0.78700602, + "num_input_tokens_seen": 50592415, + "step": 2361, + "time_per_iteration": 2.753805160522461 + }, + { + "auxiliary_loss_clip": 0.01222569, + "auxiliary_loss_mlp": 0.0103243, + "balance_loss_clip": 1.0533464, + "balance_loss_mlp": 1.02274466, + "epoch": 0.2840137076895329, + "flos": 20406992728320.0, + "grad_norm": 1.800646032893291, + "language_loss": 0.833987, + "learning_rate": 3.3606445564132326e-06, + "loss": 0.85653704, + "num_input_tokens_seen": 50609710, + "step": 2362, + "time_per_iteration": 2.7320468425750732 + }, + { + "auxiliary_loss_clip": 0.01209368, + "auxiliary_loss_mlp": 0.01380067, + "balance_loss_clip": 1.06439769, + "balance_loss_mlp": 1.000036, + "epoch": 0.28413395058017193, + "flos": 20048030161920.0, + "grad_norm": 1.9271324919253492, + "language_loss": 0.82276958, + "learning_rate": 3.360073534706118e-06, + "loss": 0.84866393, + "num_input_tokens_seen": 50626865, + "step": 2363, + "time_per_iteration": 2.6140172481536865 + }, + { + "auxiliary_loss_clip": 0.01226395, + "auxiliary_loss_mlp": 0.01038969, + "balance_loss_clip": 1.05972993, + "balance_loss_mlp": 1.02945566, + "epoch": 0.28425419347081105, + "flos": 37663613256960.0, + "grad_norm": 2.43823565008204, + "language_loss": 0.75523281, + "learning_rate": 3.35950230667693e-06, + "loss": 0.77788645, + "num_input_tokens_seen": 50648560, + "step": 2364, + "time_per_iteration": 2.8373007774353027 + }, + { + "auxiliary_loss_clip": 0.01216737, + "auxiliary_loss_mlp": 0.01031962, + "balance_loss_clip": 1.05900288, + "balance_loss_mlp": 1.02259827, + "epoch": 0.28437443636145016, + "flos": 13845072539520.0, + "grad_norm": 2.1204171064678237, + "language_loss": 0.86160332, + "learning_rate": 3.358930872412323e-06, + "loss": 0.8840903, + "num_input_tokens_seen": 50665725, + "step": 2365, + "time_per_iteration": 2.612334728240967 + }, + { + "auxiliary_loss_clip": 0.01214238, + "auxiliary_loss_mlp": 0.01037096, + "balance_loss_clip": 1.06012273, + "balance_loss_mlp": 1.02875161, + "epoch": 0.2844946792520892, + "flos": 22747794243840.0, + "grad_norm": 1.5846261801429833, + "language_loss": 0.80776811, + "learning_rate": 3.3583592319989825e-06, + "loss": 0.83028144, + "num_input_tokens_seen": 50685095, + "step": 2366, + "time_per_iteration": 2.6957335472106934 + }, + { + "auxiliary_loss_clip": 0.01223029, + "auxiliary_loss_mlp": 0.01034429, + "balance_loss_clip": 1.06056654, + "balance_loss_mlp": 1.02421856, + "epoch": 0.2846149221427283, + "flos": 32415987709440.0, + "grad_norm": 2.393248713132126, + "language_loss": 0.6863783, + "learning_rate": 3.357787385523627e-06, + "loss": 0.70895284, + "num_input_tokens_seen": 50706500, + "step": 2367, + "time_per_iteration": 2.705846071243286 + }, + { + "auxiliary_loss_clip": 0.01256304, + "auxiliary_loss_mlp": 0.0103313, + "balance_loss_clip": 1.05255675, + "balance_loss_mlp": 1.02361691, + "epoch": 0.2847351650333674, + "flos": 28475976873600.0, + "grad_norm": 2.902204505822212, + "language_loss": 0.82617891, + "learning_rate": 3.3572153330730048e-06, + "loss": 0.84907317, + "num_input_tokens_seen": 50727595, + "step": 2368, + "time_per_iteration": 2.847848892211914 + }, + { + "auxiliary_loss_clip": 0.01138603, + "auxiliary_loss_mlp": 0.01003412, + "balance_loss_clip": 1.02612388, + "balance_loss_mlp": 1.00111163, + "epoch": 0.2848554079240065, + "flos": 55753399704960.0, + "grad_norm": 0.8319324792780064, + "language_loss": 0.64722049, + "learning_rate": 3.3566430747338956e-06, + "loss": 0.66864061, + "num_input_tokens_seen": 50782800, + "step": 2369, + "time_per_iteration": 3.097316026687622 + }, + { + "auxiliary_loss_clip": 0.0121715, + "auxiliary_loss_mlp": 0.01032041, + "balance_loss_clip": 1.05790997, + "balance_loss_mlp": 1.02298713, + "epoch": 0.2849756508146456, + "flos": 11836875985920.0, + "grad_norm": 2.078659937862253, + "language_loss": 0.86398441, + "learning_rate": 3.35607061059311e-06, + "loss": 0.88647634, + "num_input_tokens_seen": 50797730, + "step": 2370, + "time_per_iteration": 2.655349016189575 + }, + { + "auxiliary_loss_clip": 0.01204913, + "auxiliary_loss_mlp": 0.01028188, + "balance_loss_clip": 1.061795, + "balance_loss_mlp": 1.01885402, + "epoch": 0.28509589370528465, + "flos": 25155209531520.0, + "grad_norm": 1.8247850159218262, + "language_loss": 0.74997622, + "learning_rate": 3.3554979407374917e-06, + "loss": 0.77230728, + "num_input_tokens_seen": 50819840, + "step": 2371, + "time_per_iteration": 2.6444530487060547 + }, + { + "auxiliary_loss_clip": 0.01213485, + "auxiliary_loss_mlp": 0.01032095, + "balance_loss_clip": 1.05715823, + "balance_loss_mlp": 1.02301717, + "epoch": 0.28521613659592376, + "flos": 19974808287360.0, + "grad_norm": 1.5420202699423278, + "language_loss": 0.73441273, + "learning_rate": 3.3549250652539134e-06, + "loss": 0.75686854, + "num_input_tokens_seen": 50838935, + "step": 2372, + "time_per_iteration": 2.666677713394165 + }, + { + "auxiliary_loss_clip": 0.01220286, + "auxiliary_loss_mlp": 0.01034958, + "balance_loss_clip": 1.05426407, + "balance_loss_mlp": 1.02489662, + "epoch": 0.2853363794865629, + "flos": 23367971491200.0, + "grad_norm": 1.9547689809781412, + "language_loss": 0.81820965, + "learning_rate": 3.3543519842292794e-06, + "loss": 0.84076208, + "num_input_tokens_seen": 50858590, + "step": 2373, + "time_per_iteration": 2.749878168106079 + }, + { + "auxiliary_loss_clip": 0.01205296, + "auxiliary_loss_mlp": 0.01380055, + "balance_loss_clip": 1.06012106, + "balance_loss_mlp": 1.0001179, + "epoch": 0.28545662237720193, + "flos": 19861940776320.0, + "grad_norm": 2.4019401445000312, + "language_loss": 0.83595598, + "learning_rate": 3.353778697750527e-06, + "loss": 0.86180943, + "num_input_tokens_seen": 50876995, + "step": 2374, + "time_per_iteration": 2.5831334590911865 + }, + { + "auxiliary_loss_clip": 0.01213468, + "auxiliary_loss_mlp": 0.01029798, + "balance_loss_clip": 1.05514979, + "balance_loss_mlp": 1.01955831, + "epoch": 0.28557686526784104, + "flos": 23879016241920.0, + "grad_norm": 1.9691453086505577, + "language_loss": 0.89233273, + "learning_rate": 3.353205205904622e-06, + "loss": 0.91476542, + "num_input_tokens_seen": 50896105, + "step": 2375, + "time_per_iteration": 2.7195322513580322 + }, + { + "auxiliary_loss_clip": 0.0122118, + "auxiliary_loss_mlp": 0.01032175, + "balance_loss_clip": 1.05649018, + "balance_loss_mlp": 1.02269793, + "epoch": 0.28569710815848015, + "flos": 44890384233600.0, + "grad_norm": 1.9128719171031634, + "language_loss": 0.71895146, + "learning_rate": 3.3526315087785637e-06, + "loss": 0.74148512, + "num_input_tokens_seen": 50917220, + "step": 2376, + "time_per_iteration": 2.8252525329589844 + }, + { + "auxiliary_loss_clip": 0.01217387, + "auxiliary_loss_mlp": 0.0103083, + "balance_loss_clip": 1.05151629, + "balance_loss_mlp": 1.02120972, + "epoch": 0.2858173510491192, + "flos": 26829759628800.0, + "grad_norm": 1.5581993417525248, + "language_loss": 0.80911243, + "learning_rate": 3.3520576064593805e-06, + "loss": 0.83159459, + "num_input_tokens_seen": 50937175, + "step": 2377, + "time_per_iteration": 2.777764081954956 + }, + { + "auxiliary_loss_clip": 0.01218716, + "auxiliary_loss_mlp": 0.0103568, + "balance_loss_clip": 1.06020987, + "balance_loss_mlp": 1.02622032, + "epoch": 0.2859375939397583, + "flos": 23148916398720.0, + "grad_norm": 1.673828337163378, + "language_loss": 0.81472099, + "learning_rate": 3.3514834990341337e-06, + "loss": 0.8372649, + "num_input_tokens_seen": 50957500, + "step": 2378, + "time_per_iteration": 2.6961512565612793 + }, + { + "auxiliary_loss_clip": 0.01227603, + "auxiliary_loss_mlp": 0.01030481, + "balance_loss_clip": 1.0573113, + "balance_loss_mlp": 1.02192163, + "epoch": 0.2860578368303974, + "flos": 12129799397760.0, + "grad_norm": 3.150411132963878, + "language_loss": 0.92766213, + "learning_rate": 3.3509091865899144e-06, + "loss": 0.950243, + "num_input_tokens_seen": 50972690, + "step": 2379, + "time_per_iteration": 3.672611951828003 + }, + { + "auxiliary_loss_clip": 0.01205702, + "auxiliary_loss_mlp": 0.01032534, + "balance_loss_clip": 1.05949867, + "balance_loss_mlp": 1.02332509, + "epoch": 0.2861780797210365, + "flos": 19938035738880.0, + "grad_norm": 2.0258592617411795, + "language_loss": 0.70531082, + "learning_rate": 3.350334669213846e-06, + "loss": 0.7276932, + "num_input_tokens_seen": 50990095, + "step": 2380, + "time_per_iteration": 3.4919369220733643 + }, + { + "auxiliary_loss_clip": 0.01215171, + "auxiliary_loss_mlp": 0.01028587, + "balance_loss_clip": 1.0612551, + "balance_loss_mlp": 1.01949739, + "epoch": 0.2862983226116756, + "flos": 27563127609600.0, + "grad_norm": 3.9001212209711484, + "language_loss": 0.75642526, + "learning_rate": 3.3497599469930816e-06, + "loss": 0.77886283, + "num_input_tokens_seen": 51008305, + "step": 2381, + "time_per_iteration": 3.582291841506958 + }, + { + "auxiliary_loss_clip": 0.01206749, + "auxiliary_loss_mlp": 0.01031392, + "balance_loss_clip": 1.05907607, + "balance_loss_mlp": 1.02188468, + "epoch": 0.28641856550231465, + "flos": 22053964158720.0, + "grad_norm": 2.3462691724513527, + "language_loss": 0.83234316, + "learning_rate": 3.349185020014807e-06, + "loss": 0.85472459, + "num_input_tokens_seen": 51025570, + "step": 2382, + "time_per_iteration": 2.686267852783203 + }, + { + "auxiliary_loss_clip": 0.01213728, + "auxiliary_loss_mlp": 0.01038053, + "balance_loss_clip": 1.05678821, + "balance_loss_mlp": 1.02864766, + "epoch": 0.28653880839295376, + "flos": 22378775869440.0, + "grad_norm": 1.8013771772031417, + "language_loss": 0.74542189, + "learning_rate": 3.348609888366237e-06, + "loss": 0.76793975, + "num_input_tokens_seen": 51044585, + "step": 2383, + "time_per_iteration": 2.6826119422912598 + }, + { + "auxiliary_loss_clip": 0.01239877, + "auxiliary_loss_mlp": 0.01029805, + "balance_loss_clip": 1.05029047, + "balance_loss_mlp": 1.02086425, + "epoch": 0.28665905128359287, + "flos": 23367971491200.0, + "grad_norm": 2.295764284468142, + "language_loss": 0.62806195, + "learning_rate": 3.348034552134619e-06, + "loss": 0.65075874, + "num_input_tokens_seen": 51063990, + "step": 2384, + "time_per_iteration": 2.7770605087280273 + }, + { + "auxiliary_loss_clip": 0.01231907, + "auxiliary_loss_mlp": 0.01032092, + "balance_loss_clip": 1.05448127, + "balance_loss_mlp": 1.02271664, + "epoch": 0.2867792941742319, + "flos": 20881695893760.0, + "grad_norm": 2.7571849198004315, + "language_loss": 0.84471631, + "learning_rate": 3.3474590114072316e-06, + "loss": 0.8673563, + "num_input_tokens_seen": 51081990, + "step": 2385, + "time_per_iteration": 2.815758466720581 + }, + { + "auxiliary_loss_clip": 0.0122515, + "auxiliary_loss_mlp": 0.01031373, + "balance_loss_clip": 1.05493808, + "balance_loss_mlp": 1.02163935, + "epoch": 0.28689953706487104, + "flos": 20664005518080.0, + "grad_norm": 1.8575972726752232, + "language_loss": 0.8330242, + "learning_rate": 3.3468832662713836e-06, + "loss": 0.85558945, + "num_input_tokens_seen": 51100235, + "step": 2386, + "time_per_iteration": 3.6393048763275146 + }, + { + "auxiliary_loss_clip": 0.01229452, + "auxiliary_loss_mlp": 0.01029786, + "balance_loss_clip": 1.05740833, + "balance_loss_mlp": 1.02074945, + "epoch": 0.28701977995551015, + "flos": 12675533708160.0, + "grad_norm": 2.2284896176649114, + "language_loss": 0.84048867, + "learning_rate": 3.346307316814415e-06, + "loss": 0.8630811, + "num_input_tokens_seen": 51115405, + "step": 2387, + "time_per_iteration": 2.720770835876465 + }, + { + "auxiliary_loss_clip": 0.01214003, + "auxiliary_loss_mlp": 0.01029979, + "balance_loss_clip": 1.06097949, + "balance_loss_mlp": 1.02085972, + "epoch": 0.2871400228461492, + "flos": 21252366293760.0, + "grad_norm": 2.832086208679477, + "language_loss": 0.75687516, + "learning_rate": 3.3457311631236965e-06, + "loss": 0.77931494, + "num_input_tokens_seen": 51136390, + "step": 2388, + "time_per_iteration": 2.6624503135681152 + }, + { + "auxiliary_loss_clip": 0.01208104, + "auxiliary_loss_mlp": 0.01032625, + "balance_loss_clip": 1.05484736, + "balance_loss_mlp": 1.02264166, + "epoch": 0.2872602657367883, + "flos": 25119262995840.0, + "grad_norm": 1.816322089077284, + "language_loss": 0.8465482, + "learning_rate": 3.345154805286631e-06, + "loss": 0.86895555, + "num_input_tokens_seen": 51156650, + "step": 2389, + "time_per_iteration": 2.6940369606018066 + }, + { + "auxiliary_loss_clip": 0.01204469, + "auxiliary_loss_mlp": 0.01033743, + "balance_loss_clip": 1.05572128, + "balance_loss_mlp": 1.02416492, + "epoch": 0.2873805086274274, + "flos": 16646606830080.0, + "grad_norm": 2.430692164811192, + "language_loss": 0.7624647, + "learning_rate": 3.344578243390651e-06, + "loss": 0.78484678, + "num_input_tokens_seen": 51172210, + "step": 2390, + "time_per_iteration": 2.5523781776428223 + }, + { + "auxiliary_loss_clip": 0.01220081, + "auxiliary_loss_mlp": 0.01028051, + "balance_loss_clip": 1.05724633, + "balance_loss_mlp": 1.01901472, + "epoch": 0.2875007515180665, + "flos": 17420123237760.0, + "grad_norm": 2.0488560403071543, + "language_loss": 0.77997434, + "learning_rate": 3.3440014775232206e-06, + "loss": 0.80245566, + "num_input_tokens_seen": 51190265, + "step": 2391, + "time_per_iteration": 2.6645820140838623 + }, + { + "auxiliary_loss_clip": 0.01232088, + "auxiliary_loss_mlp": 0.01028117, + "balance_loss_clip": 1.05539131, + "balance_loss_mlp": 1.01874113, + "epoch": 0.2876209944087056, + "flos": 23434190213760.0, + "grad_norm": 1.895288333373904, + "language_loss": 0.71323943, + "learning_rate": 3.343424507771834e-06, + "loss": 0.73584151, + "num_input_tokens_seen": 51208475, + "step": 2392, + "time_per_iteration": 2.7121400833129883 + }, + { + "auxiliary_loss_clip": 0.01230531, + "auxiliary_loss_mlp": 0.01036451, + "balance_loss_clip": 1.05639625, + "balance_loss_mlp": 1.02731919, + "epoch": 0.2877412372993447, + "flos": 13735509079680.0, + "grad_norm": 1.76467506732025, + "language_loss": 0.86410391, + "learning_rate": 3.342847334224018e-06, + "loss": 0.88677371, + "num_input_tokens_seen": 51225875, + "step": 2393, + "time_per_iteration": 2.657792806625366 + }, + { + "auxiliary_loss_clip": 0.01117214, + "auxiliary_loss_mlp": 0.01003273, + "balance_loss_clip": 1.03016686, + "balance_loss_mlp": 1.00107956, + "epoch": 0.28786148018998375, + "flos": 58079695104000.0, + "grad_norm": 0.9501949188716489, + "language_loss": 0.62425488, + "learning_rate": 3.342269956967329e-06, + "loss": 0.64545971, + "num_input_tokens_seen": 51287780, + "step": 2394, + "time_per_iteration": 3.2834243774414062 + }, + { + "auxiliary_loss_clip": 0.01218141, + "auxiliary_loss_mlp": 0.0103846, + "balance_loss_clip": 1.06026924, + "balance_loss_mlp": 1.02835655, + "epoch": 0.28798172308062286, + "flos": 23435052140160.0, + "grad_norm": 3.0867639025209868, + "language_loss": 0.71682572, + "learning_rate": 3.341692376089355e-06, + "loss": 0.7393918, + "num_input_tokens_seen": 51303335, + "step": 2395, + "time_per_iteration": 2.7027153968811035 + }, + { + "auxiliary_loss_clip": 0.01208597, + "auxiliary_loss_mlp": 0.01028425, + "balance_loss_clip": 1.05697954, + "balance_loss_mlp": 1.01988411, + "epoch": 0.288101965971262, + "flos": 25110033200640.0, + "grad_norm": 2.8372787262771233, + "language_loss": 0.84106529, + "learning_rate": 3.3411145916777146e-06, + "loss": 0.86343551, + "num_input_tokens_seen": 51317495, + "step": 2396, + "time_per_iteration": 2.6953964233398438 + }, + { + "auxiliary_loss_clip": 0.01209512, + "auxiliary_loss_mlp": 0.01037596, + "balance_loss_clip": 1.05438101, + "balance_loss_mlp": 1.02793419, + "epoch": 0.28822220886190103, + "flos": 16252559654400.0, + "grad_norm": 2.131425976508854, + "language_loss": 0.90757358, + "learning_rate": 3.3405366038200566e-06, + "loss": 0.93004459, + "num_input_tokens_seen": 51336430, + "step": 2397, + "time_per_iteration": 2.650428533554077 + }, + { + "auxiliary_loss_clip": 0.01228834, + "auxiliary_loss_mlp": 0.01043184, + "balance_loss_clip": 1.0637784, + "balance_loss_mlp": 1.03340912, + "epoch": 0.28834245175254014, + "flos": 24535642815360.0, + "grad_norm": 2.291762257666975, + "language_loss": 0.84829807, + "learning_rate": 3.3399584126040617e-06, + "loss": 0.87101829, + "num_input_tokens_seen": 51355930, + "step": 2398, + "time_per_iteration": 2.898505210876465 + }, + { + "auxiliary_loss_clip": 0.01206215, + "auxiliary_loss_mlp": 0.01380039, + "balance_loss_clip": 1.06180263, + "balance_loss_mlp": 1.00002813, + "epoch": 0.2884626946431792, + "flos": 24571445696640.0, + "grad_norm": 1.875310941788629, + "language_loss": 0.90567338, + "learning_rate": 3.339380018117441e-06, + "loss": 0.93153596, + "num_input_tokens_seen": 51376765, + "step": 2399, + "time_per_iteration": 2.7224628925323486 + }, + { + "auxiliary_loss_clip": 0.01209598, + "auxiliary_loss_mlp": 0.01027058, + "balance_loss_clip": 1.05979323, + "balance_loss_mlp": 1.01775336, + "epoch": 0.2885829375338183, + "flos": 16544657053440.0, + "grad_norm": 2.843893576429507, + "language_loss": 0.7827, + "learning_rate": 3.3388014204479366e-06, + "loss": 0.80506659, + "num_input_tokens_seen": 51394570, + "step": 2400, + "time_per_iteration": 2.6412947177886963 + }, + { + "auxiliary_loss_clip": 0.01207986, + "auxiliary_loss_mlp": 0.01036022, + "balance_loss_clip": 1.06165099, + "balance_loss_mlp": 1.02615118, + "epoch": 0.2887031804244574, + "flos": 24061226958720.0, + "grad_norm": 1.9500336933166713, + "language_loss": 0.91454399, + "learning_rate": 3.338222619683321e-06, + "loss": 0.93698412, + "num_input_tokens_seen": 51414535, + "step": 2401, + "time_per_iteration": 2.653395652770996 + }, + { + "auxiliary_loss_clip": 0.01224241, + "auxiliary_loss_mlp": 0.01024488, + "balance_loss_clip": 1.058393, + "balance_loss_mlp": 1.01545238, + "epoch": 0.2888234233150965, + "flos": 23330696152320.0, + "grad_norm": 2.498055129211515, + "language_loss": 0.74116147, + "learning_rate": 3.337643615911398e-06, + "loss": 0.76364887, + "num_input_tokens_seen": 51434160, + "step": 2402, + "time_per_iteration": 2.7371368408203125 + }, + { + "auxiliary_loss_clip": 0.0121534, + "auxiliary_loss_mlp": 0.01025149, + "balance_loss_clip": 1.05842066, + "balance_loss_mlp": 1.01672709, + "epoch": 0.2889436662057356, + "flos": 22272767856000.0, + "grad_norm": 2.0668696771336967, + "language_loss": 0.78640544, + "learning_rate": 3.3370644092200026e-06, + "loss": 0.80881035, + "num_input_tokens_seen": 51451435, + "step": 2403, + "time_per_iteration": 2.715911388397217 + }, + { + "auxiliary_loss_clip": 0.01213234, + "auxiliary_loss_mlp": 0.01034546, + "balance_loss_clip": 1.05186224, + "balance_loss_mlp": 1.02548003, + "epoch": 0.2890639090963747, + "flos": 21616931381760.0, + "grad_norm": 2.3600460914611516, + "language_loss": 0.78328913, + "learning_rate": 3.3364849996969985e-06, + "loss": 0.80576694, + "num_input_tokens_seen": 51471455, + "step": 2404, + "time_per_iteration": 2.696101427078247 + }, + { + "auxiliary_loss_clip": 0.01211805, + "auxiliary_loss_mlp": 0.01034901, + "balance_loss_clip": 1.05943811, + "balance_loss_mlp": 1.02621651, + "epoch": 0.28918415198701375, + "flos": 28585540333440.0, + "grad_norm": 1.837447814127416, + "language_loss": 0.85664058, + "learning_rate": 3.335905387430283e-06, + "loss": 0.87910759, + "num_input_tokens_seen": 51492890, + "step": 2405, + "time_per_iteration": 3.5428707599639893 + }, + { + "auxiliary_loss_clip": 0.01225061, + "auxiliary_loss_mlp": 0.01030988, + "balance_loss_clip": 1.05547953, + "balance_loss_mlp": 1.02211833, + "epoch": 0.28930439487765286, + "flos": 21944688007680.0, + "grad_norm": 1.7297545300796917, + "language_loss": 0.83052957, + "learning_rate": 3.335325572507782e-06, + "loss": 0.85309005, + "num_input_tokens_seen": 51513390, + "step": 2406, + "time_per_iteration": 3.6179041862487793 + }, + { + "auxiliary_loss_clip": 0.01209258, + "auxiliary_loss_mlp": 0.01380169, + "balance_loss_clip": 1.06497228, + "balance_loss_mlp": 1.00018311, + "epoch": 0.28942463776829197, + "flos": 19281911955840.0, + "grad_norm": 1.736164628897546, + "language_loss": 0.7375679, + "learning_rate": 3.3347455550174537e-06, + "loss": 0.76346219, + "num_input_tokens_seen": 51532730, + "step": 2407, + "time_per_iteration": 3.5188801288604736 + }, + { + "auxiliary_loss_clip": 0.01223387, + "auxiliary_loss_mlp": 0.01027108, + "balance_loss_clip": 1.05168784, + "balance_loss_mlp": 1.01837587, + "epoch": 0.289544880658931, + "flos": 14645700737280.0, + "grad_norm": 1.874862999581222, + "language_loss": 0.68118495, + "learning_rate": 3.3341653350472864e-06, + "loss": 0.70368993, + "num_input_tokens_seen": 51549560, + "step": 2408, + "time_per_iteration": 2.7056663036346436 + }, + { + "auxiliary_loss_clip": 0.01212173, + "auxiliary_loss_mlp": 0.01035128, + "balance_loss_clip": 1.06095099, + "balance_loss_mlp": 1.0243156, + "epoch": 0.28966512354957014, + "flos": 28621881918720.0, + "grad_norm": 3.497410654486015, + "language_loss": 0.69123948, + "learning_rate": 3.333584912685298e-06, + "loss": 0.71371251, + "num_input_tokens_seen": 51568180, + "step": 2409, + "time_per_iteration": 2.68296480178833 + }, + { + "auxiliary_loss_clip": 0.01139491, + "auxiliary_loss_mlp": 0.01002158, + "balance_loss_clip": 1.0297519, + "balance_loss_mlp": 0.999798, + "epoch": 0.28978536644020925, + "flos": 64711784511360.0, + "grad_norm": 0.8840731022287995, + "language_loss": 0.55613959, + "learning_rate": 3.3330042880195385e-06, + "loss": 0.57755613, + "num_input_tokens_seen": 51622530, + "step": 2410, + "time_per_iteration": 3.245706081390381 + }, + { + "auxiliary_loss_clip": 0.0122098, + "auxiliary_loss_mlp": 0.01029359, + "balance_loss_clip": 1.05580258, + "balance_loss_mlp": 1.02042389, + "epoch": 0.2899056093308483, + "flos": 18624638937600.0, + "grad_norm": 2.37209632434107, + "language_loss": 0.78507888, + "learning_rate": 3.3324234611380888e-06, + "loss": 0.80758226, + "num_input_tokens_seen": 51641260, + "step": 2411, + "time_per_iteration": 2.7640485763549805 + }, + { + "auxiliary_loss_clip": 0.01222991, + "auxiliary_loss_mlp": 0.01034459, + "balance_loss_clip": 1.05584145, + "balance_loss_mlp": 1.0253098, + "epoch": 0.2900258522214874, + "flos": 22893735202560.0, + "grad_norm": 1.8163830464546888, + "language_loss": 0.82036686, + "learning_rate": 3.3318424321290596e-06, + "loss": 0.84294128, + "num_input_tokens_seen": 51660975, + "step": 2412, + "time_per_iteration": 3.649555206298828 + }, + { + "auxiliary_loss_clip": 0.01133513, + "auxiliary_loss_mlp": 0.01005751, + "balance_loss_clip": 1.02647066, + "balance_loss_mlp": 1.00360513, + "epoch": 0.2901460951121265, + "flos": 71106036013440.0, + "grad_norm": 0.8313801820391905, + "language_loss": 0.59951842, + "learning_rate": 3.3312612010805917e-06, + "loss": 0.62091112, + "num_input_tokens_seen": 51720550, + "step": 2413, + "time_per_iteration": 3.302919387817383 + }, + { + "auxiliary_loss_clip": 0.01208759, + "auxiliary_loss_mlp": 0.01030028, + "balance_loss_clip": 1.05674815, + "balance_loss_mlp": 1.02061021, + "epoch": 0.2902663380027656, + "flos": 32160986081280.0, + "grad_norm": 1.7005155613856369, + "language_loss": 0.7015599, + "learning_rate": 3.330679768080858e-06, + "loss": 0.72394776, + "num_input_tokens_seen": 51744435, + "step": 2414, + "time_per_iteration": 2.8127551078796387 + }, + { + "auxiliary_loss_clip": 0.01209971, + "auxiliary_loss_mlp": 0.01037719, + "balance_loss_clip": 1.05958557, + "balance_loss_mlp": 1.02888536, + "epoch": 0.2903865808934047, + "flos": 29351658539520.0, + "grad_norm": 2.1272960332310484, + "language_loss": 0.83077085, + "learning_rate": 3.3300981332180627e-06, + "loss": 0.8532477, + "num_input_tokens_seen": 51763640, + "step": 2415, + "time_per_iteration": 2.6869685649871826 + }, + { + "auxiliary_loss_clip": 0.0123711, + "auxiliary_loss_mlp": 0.01034133, + "balance_loss_clip": 1.05792785, + "balance_loss_mlp": 1.02492404, + "epoch": 0.29050682378404374, + "flos": 17089026647040.0, + "grad_norm": 1.9965812537101113, + "language_loss": 0.79963219, + "learning_rate": 3.3295162965804373e-06, + "loss": 0.8223446, + "num_input_tokens_seen": 51782135, + "step": 2416, + "time_per_iteration": 2.7649874687194824 + }, + { + "auxiliary_loss_clip": 0.01227126, + "auxiliary_loss_mlp": 0.01029067, + "balance_loss_clip": 1.05638397, + "balance_loss_mlp": 1.02034736, + "epoch": 0.29062706667468285, + "flos": 17858233422720.0, + "grad_norm": 2.864193625378686, + "language_loss": 0.78477824, + "learning_rate": 3.328934258256247e-06, + "loss": 0.80734015, + "num_input_tokens_seen": 51800200, + "step": 2417, + "time_per_iteration": 2.7022953033447266 + }, + { + "auxiliary_loss_clip": 0.01209609, + "auxiliary_loss_mlp": 0.0103243, + "balance_loss_clip": 1.05509472, + "balance_loss_mlp": 1.02310824, + "epoch": 0.29074730956532197, + "flos": 24279815174400.0, + "grad_norm": 2.284608658000821, + "language_loss": 0.67512894, + "learning_rate": 3.3283520183337856e-06, + "loss": 0.69754934, + "num_input_tokens_seen": 51819905, + "step": 2418, + "time_per_iteration": 2.6958203315734863 + }, + { + "auxiliary_loss_clip": 0.01215694, + "auxiliary_loss_mlp": 0.01032153, + "balance_loss_clip": 1.05387771, + "balance_loss_mlp": 1.0231111, + "epoch": 0.290867552455961, + "flos": 22340961826560.0, + "grad_norm": 1.695815746033169, + "language_loss": 0.69165051, + "learning_rate": 3.3277695769013797e-06, + "loss": 0.71412897, + "num_input_tokens_seen": 51839350, + "step": 2419, + "time_per_iteration": 2.6420066356658936 + }, + { + "auxiliary_loss_clip": 0.01214352, + "auxiliary_loss_mlp": 0.01030652, + "balance_loss_clip": 1.06088471, + "balance_loss_mlp": 1.02143741, + "epoch": 0.29098779534660013, + "flos": 23186155824000.0, + "grad_norm": 1.9112108137223451, + "language_loss": 0.77648813, + "learning_rate": 3.327186934047385e-06, + "loss": 0.79893816, + "num_input_tokens_seen": 51858045, + "step": 2420, + "time_per_iteration": 2.7145438194274902 + }, + { + "auxiliary_loss_clip": 0.01207375, + "auxiliary_loss_mlp": 0.01031828, + "balance_loss_clip": 1.05240965, + "balance_loss_mlp": 1.02329803, + "epoch": 0.29110803823723924, + "flos": 15304194817920.0, + "grad_norm": 2.119748557997067, + "language_loss": 0.6523295, + "learning_rate": 3.3266040898601877e-06, + "loss": 0.67472154, + "num_input_tokens_seen": 51875880, + "step": 2421, + "time_per_iteration": 2.6554315090179443 + }, + { + "auxiliary_loss_clip": 0.01228277, + "auxiliary_loss_mlp": 0.0103516, + "balance_loss_clip": 1.05051315, + "balance_loss_mlp": 1.02604699, + "epoch": 0.2912282811278783, + "flos": 22595352923520.0, + "grad_norm": 1.8748854033898488, + "language_loss": 0.78177857, + "learning_rate": 3.3260210444282045e-06, + "loss": 0.80441296, + "num_input_tokens_seen": 51893835, + "step": 2422, + "time_per_iteration": 2.9354188442230225 + }, + { + "auxiliary_loss_clip": 0.0120466, + "auxiliary_loss_mlp": 0.01034705, + "balance_loss_clip": 1.0559473, + "balance_loss_mlp": 1.02567458, + "epoch": 0.2913485240185174, + "flos": 24497900599680.0, + "grad_norm": 2.002949287828724, + "language_loss": 0.73203063, + "learning_rate": 3.325437797839883e-06, + "loss": 0.75442427, + "num_input_tokens_seen": 51912205, + "step": 2423, + "time_per_iteration": 2.6664352416992188 + }, + { + "auxiliary_loss_clip": 0.01203844, + "auxiliary_loss_mlp": 0.01032864, + "balance_loss_clip": 1.05888164, + "balance_loss_mlp": 1.02380466, + "epoch": 0.2914687669091565, + "flos": 17931024334080.0, + "grad_norm": 2.425699786709941, + "language_loss": 0.75166285, + "learning_rate": 3.3248543501837015e-06, + "loss": 0.77402991, + "num_input_tokens_seen": 51929410, + "step": 2424, + "time_per_iteration": 2.6882846355438232 + }, + { + "auxiliary_loss_clip": 0.01242149, + "auxiliary_loss_mlp": 0.01036687, + "balance_loss_clip": 1.0561558, + "balance_loss_mlp": 1.02721632, + "epoch": 0.2915890097997956, + "flos": 22529313768960.0, + "grad_norm": 1.9913424605743397, + "language_loss": 0.77277321, + "learning_rate": 3.3242707015481684e-06, + "loss": 0.79556155, + "num_input_tokens_seen": 51949345, + "step": 2425, + "time_per_iteration": 2.7978873252868652 + }, + { + "auxiliary_loss_clip": 0.01214218, + "auxiliary_loss_mlp": 0.01029884, + "balance_loss_clip": 1.05126882, + "balance_loss_mlp": 1.02148008, + "epoch": 0.2917092526904347, + "flos": 13845216193920.0, + "grad_norm": 10.057324004936287, + "language_loss": 0.8075695, + "learning_rate": 3.323686852021823e-06, + "loss": 0.83001053, + "num_input_tokens_seen": 51966855, + "step": 2426, + "time_per_iteration": 2.6303610801696777 + }, + { + "auxiliary_loss_clip": 0.01233399, + "auxiliary_loss_mlp": 0.01032921, + "balance_loss_clip": 1.05442023, + "balance_loss_mlp": 1.02338398, + "epoch": 0.2918294955810738, + "flos": 22674859678080.0, + "grad_norm": 1.8786335406817058, + "language_loss": 0.79517949, + "learning_rate": 3.323102801693235e-06, + "loss": 0.81784272, + "num_input_tokens_seen": 51985620, + "step": 2427, + "time_per_iteration": 2.7592716217041016 + }, + { + "auxiliary_loss_clip": 0.01204115, + "auxiliary_loss_mlp": 0.01032058, + "balance_loss_clip": 1.05615497, + "balance_loss_mlp": 1.02292025, + "epoch": 0.29194973847171285, + "flos": 23438284364160.0, + "grad_norm": 2.315705522621508, + "language_loss": 0.81051242, + "learning_rate": 3.322518550651003e-06, + "loss": 0.83287418, + "num_input_tokens_seen": 52004930, + "step": 2428, + "time_per_iteration": 2.6973695755004883 + }, + { + "auxiliary_loss_clip": 0.01224357, + "auxiliary_loss_mlp": 0.01033457, + "balance_loss_clip": 1.05484462, + "balance_loss_mlp": 1.02448058, + "epoch": 0.29206998136235196, + "flos": 21909064694400.0, + "grad_norm": 1.8353468118838805, + "language_loss": 0.81232607, + "learning_rate": 3.3219340989837586e-06, + "loss": 0.83490419, + "num_input_tokens_seen": 52024920, + "step": 2429, + "time_per_iteration": 2.7022111415863037 + }, + { + "auxiliary_loss_clip": 0.01219478, + "auxiliary_loss_mlp": 0.01031692, + "balance_loss_clip": 1.05713069, + "balance_loss_mlp": 1.0228827, + "epoch": 0.292190224252991, + "flos": 23215925220480.0, + "grad_norm": 1.7606065979960783, + "language_loss": 0.8046273, + "learning_rate": 3.3213494467801625e-06, + "loss": 0.8271389, + "num_input_tokens_seen": 52044095, + "step": 2430, + "time_per_iteration": 3.6297245025634766 + }, + { + "auxiliary_loss_clip": 0.01250499, + "auxiliary_loss_mlp": 0.01025844, + "balance_loss_clip": 1.04793286, + "balance_loss_mlp": 1.01727235, + "epoch": 0.2923104671436301, + "flos": 20740818752640.0, + "grad_norm": 2.0143309587752856, + "language_loss": 0.7123301, + "learning_rate": 3.3207645941289063e-06, + "loss": 0.73509347, + "num_input_tokens_seen": 52062440, + "step": 2431, + "time_per_iteration": 2.988915205001831 + }, + { + "auxiliary_loss_clip": 0.01210826, + "auxiliary_loss_mlp": 0.0137976, + "balance_loss_clip": 1.05866659, + "balance_loss_mlp": 0.99999106, + "epoch": 0.29243071003426924, + "flos": 35809114999680.0, + "grad_norm": 1.7714636907902594, + "language_loss": 0.80121601, + "learning_rate": 3.320179541118711e-06, + "loss": 0.82712185, + "num_input_tokens_seen": 52084940, + "step": 2432, + "time_per_iteration": 3.0306248664855957 + }, + { + "auxiliary_loss_clip": 0.01118295, + "auxiliary_loss_mlp": 0.01008056, + "balance_loss_clip": 1.03056312, + "balance_loss_mlp": 1.00571966, + "epoch": 0.2925509529249083, + "flos": 58081598524800.0, + "grad_norm": 0.9999367689279309, + "language_loss": 0.60296929, + "learning_rate": 3.3195942878383293e-06, + "loss": 0.62423283, + "num_input_tokens_seen": 52141040, + "step": 2433, + "time_per_iteration": 5.02247953414917 + }, + { + "auxiliary_loss_clip": 0.01213178, + "auxiliary_loss_mlp": 0.01027886, + "balance_loss_clip": 1.05847692, + "balance_loss_mlp": 1.01939881, + "epoch": 0.2926711958155474, + "flos": 21397122103680.0, + "grad_norm": 1.975089247291234, + "language_loss": 0.77946544, + "learning_rate": 3.319008834376543e-06, + "loss": 0.80187607, + "num_input_tokens_seen": 52160730, + "step": 2434, + "time_per_iteration": 2.69765043258667 + }, + { + "auxiliary_loss_clip": 0.01234292, + "auxiliary_loss_mlp": 0.01033331, + "balance_loss_clip": 1.05289125, + "balance_loss_mlp": 1.02411675, + "epoch": 0.2927914387061865, + "flos": 23185796688000.0, + "grad_norm": 2.857891434790069, + "language_loss": 0.88457912, + "learning_rate": 3.3184231808221654e-06, + "loss": 0.90725541, + "num_input_tokens_seen": 52175055, + "step": 2435, + "time_per_iteration": 2.7738730907440186 + }, + { + "auxiliary_loss_clip": 0.01231708, + "auxiliary_loss_mlp": 0.01039912, + "balance_loss_clip": 1.05606437, + "balance_loss_mlp": 1.03071463, + "epoch": 0.29291168159682557, + "flos": 22455553190400.0, + "grad_norm": 2.343232198092561, + "language_loss": 0.62885439, + "learning_rate": 3.3178373272640394e-06, + "loss": 0.65157056, + "num_input_tokens_seen": 52194150, + "step": 2436, + "time_per_iteration": 2.709099531173706 + }, + { + "auxiliary_loss_clip": 0.01203024, + "auxiliary_loss_mlp": 0.01026267, + "balance_loss_clip": 1.06076431, + "balance_loss_mlp": 1.01767802, + "epoch": 0.2930319244874647, + "flos": 21170632896000.0, + "grad_norm": 2.0404714771522383, + "language_loss": 0.84623986, + "learning_rate": 3.3172512737910387e-06, + "loss": 0.86853278, + "num_input_tokens_seen": 52211660, + "step": 2437, + "time_per_iteration": 2.685051918029785 + }, + { + "auxiliary_loss_clip": 0.01210906, + "auxiliary_loss_mlp": 0.01027848, + "balance_loss_clip": 1.0581162, + "balance_loss_mlp": 1.01892483, + "epoch": 0.2931521673781038, + "flos": 31357843931520.0, + "grad_norm": 2.218471350780363, + "language_loss": 0.88233638, + "learning_rate": 3.3166650204920674e-06, + "loss": 0.90472394, + "num_input_tokens_seen": 52232830, + "step": 2438, + "time_per_iteration": 3.65644907951355 + }, + { + "auxiliary_loss_clip": 0.01214608, + "auxiliary_loss_mlp": 0.01030814, + "balance_loss_clip": 1.06311488, + "balance_loss_mlp": 1.02106273, + "epoch": 0.29327241026874284, + "flos": 24200990778240.0, + "grad_norm": 1.669763251101543, + "language_loss": 0.8164711, + "learning_rate": 3.316078567456059e-06, + "loss": 0.83892536, + "num_input_tokens_seen": 52250670, + "step": 2439, + "time_per_iteration": 2.631990671157837 + }, + { + "auxiliary_loss_clip": 0.01250373, + "auxiliary_loss_mlp": 0.01033509, + "balance_loss_clip": 1.05281687, + "balance_loss_mlp": 1.02464581, + "epoch": 0.29339265315938196, + "flos": 24242611662720.0, + "grad_norm": 1.5809137399775508, + "language_loss": 0.75585067, + "learning_rate": 3.3154919147719786e-06, + "loss": 0.7786895, + "num_input_tokens_seen": 52271685, + "step": 2440, + "time_per_iteration": 2.8177683353424072 + }, + { + "auxiliary_loss_clip": 0.01210316, + "auxiliary_loss_mlp": 0.01030387, + "balance_loss_clip": 1.05738568, + "balance_loss_mlp": 1.02090383, + "epoch": 0.29351289605002107, + "flos": 16946641134720.0, + "grad_norm": 2.0254531790613957, + "language_loss": 0.86321849, + "learning_rate": 3.31490506252882e-06, + "loss": 0.88562554, + "num_input_tokens_seen": 52291065, + "step": 2441, + "time_per_iteration": 2.8643219470977783 + }, + { + "auxiliary_loss_clip": 0.0121595, + "auxiliary_loss_mlp": 0.01031103, + "balance_loss_clip": 1.05025625, + "balance_loss_mlp": 1.02224541, + "epoch": 0.2936331389406601, + "flos": 19829082810240.0, + "grad_norm": 1.843703671201291, + "language_loss": 0.84036541, + "learning_rate": 3.31431801081561e-06, + "loss": 0.862836, + "num_input_tokens_seen": 52310000, + "step": 2442, + "time_per_iteration": 2.7195632457733154 + }, + { + "auxiliary_loss_clip": 0.01121977, + "auxiliary_loss_mlp": 0.01002296, + "balance_loss_clip": 1.02911174, + "balance_loss_mlp": 1.00021005, + "epoch": 0.29375338183129923, + "flos": 71416844398080.0, + "grad_norm": 0.9068556192862038, + "language_loss": 0.67935568, + "learning_rate": 3.313730759721402e-06, + "loss": 0.70059842, + "num_input_tokens_seen": 52372930, + "step": 2443, + "time_per_iteration": 3.3580892086029053 + }, + { + "auxiliary_loss_clip": 0.01218386, + "auxiliary_loss_mlp": 0.01033499, + "balance_loss_clip": 1.057271, + "balance_loss_mlp": 1.02473748, + "epoch": 0.29387362472193834, + "flos": 22054502862720.0, + "grad_norm": 2.359811211649124, + "language_loss": 0.86325324, + "learning_rate": 3.313143309335282e-06, + "loss": 0.88577211, + "num_input_tokens_seen": 52391420, + "step": 2444, + "time_per_iteration": 2.781806230545044 + }, + { + "auxiliary_loss_clip": 0.01224144, + "auxiliary_loss_mlp": 0.01036044, + "balance_loss_clip": 1.05476117, + "balance_loss_mlp": 1.02756262, + "epoch": 0.2939938676125774, + "flos": 22966418373120.0, + "grad_norm": 2.679103682580223, + "language_loss": 0.84709895, + "learning_rate": 3.3125556597463665e-06, + "loss": 0.86970079, + "num_input_tokens_seen": 52410725, + "step": 2445, + "time_per_iteration": 2.7404913902282715 + }, + { + "auxiliary_loss_clip": 0.01212124, + "auxiliary_loss_mlp": 0.01030187, + "balance_loss_clip": 1.06066012, + "balance_loss_mlp": 1.02140152, + "epoch": 0.2941141105032165, + "flos": 31358705857920.0, + "grad_norm": 1.7260660843464888, + "language_loss": 0.66074532, + "learning_rate": 3.311967811043801e-06, + "loss": 0.68316841, + "num_input_tokens_seen": 52432645, + "step": 2446, + "time_per_iteration": 2.693917989730835 + }, + { + "auxiliary_loss_clip": 0.01209446, + "auxiliary_loss_mlp": 0.01030709, + "balance_loss_clip": 1.05774415, + "balance_loss_mlp": 1.02236462, + "epoch": 0.29423435339385556, + "flos": 23222138273280.0, + "grad_norm": 2.4532277459964926, + "language_loss": 0.81554055, + "learning_rate": 3.3113797633167617e-06, + "loss": 0.83794212, + "num_input_tokens_seen": 52450940, + "step": 2447, + "time_per_iteration": 2.70440411567688 + }, + { + "auxiliary_loss_clip": 0.01201192, + "auxiliary_loss_mlp": 0.01029142, + "balance_loss_clip": 1.05783153, + "balance_loss_mlp": 1.02030826, + "epoch": 0.2943545962844947, + "flos": 26864054138880.0, + "grad_norm": 2.4537019242520364, + "language_loss": 0.68684042, + "learning_rate": 3.310791516654455e-06, + "loss": 0.70914376, + "num_input_tokens_seen": 52468000, + "step": 2448, + "time_per_iteration": 2.6310291290283203 + }, + { + "auxiliary_loss_clip": 0.01229454, + "auxiliary_loss_mlp": 0.01027193, + "balance_loss_clip": 1.05152726, + "balance_loss_mlp": 1.01819241, + "epoch": 0.2944748391751338, + "flos": 20231677422720.0, + "grad_norm": 1.9404745383997837, + "language_loss": 0.79702508, + "learning_rate": 3.3102030711461177e-06, + "loss": 0.81959152, + "num_input_tokens_seen": 52487575, + "step": 2449, + "time_per_iteration": 2.74090838432312 + }, + { + "auxiliary_loss_clip": 0.01231486, + "auxiliary_loss_mlp": 0.01037267, + "balance_loss_clip": 1.05515957, + "balance_loss_mlp": 1.02818942, + "epoch": 0.29459508206577284, + "flos": 15960965045760.0, + "grad_norm": 2.3845237578961105, + "language_loss": 0.67938781, + "learning_rate": 3.3096144268810156e-06, + "loss": 0.70207536, + "num_input_tokens_seen": 52506335, + "step": 2450, + "time_per_iteration": 2.688969373703003 + }, + { + "auxiliary_loss_clip": 0.01196576, + "auxiliary_loss_mlp": 0.01034389, + "balance_loss_clip": 1.05334759, + "balance_loss_mlp": 1.02518046, + "epoch": 0.29471532495641195, + "flos": 20412882558720.0, + "grad_norm": 2.0733625259231285, + "language_loss": 0.73528552, + "learning_rate": 3.3090255839484462e-06, + "loss": 0.75759518, + "num_input_tokens_seen": 52524330, + "step": 2451, + "time_per_iteration": 2.717583417892456 + }, + { + "auxiliary_loss_clip": 0.01219234, + "auxiliary_loss_mlp": 0.01031628, + "balance_loss_clip": 1.05314338, + "balance_loss_mlp": 1.02258039, + "epoch": 0.29483556784705106, + "flos": 20376576887040.0, + "grad_norm": 2.378547304987405, + "language_loss": 0.85972309, + "learning_rate": 3.3084365424377366e-06, + "loss": 0.88223171, + "num_input_tokens_seen": 52543095, + "step": 2452, + "time_per_iteration": 2.6836588382720947 + }, + { + "auxiliary_loss_clip": 0.01146635, + "auxiliary_loss_mlp": 0.01015343, + "balance_loss_clip": 1.02666163, + "balance_loss_mlp": 1.01353121, + "epoch": 0.2949558107376901, + "flos": 68555660595840.0, + "grad_norm": 0.7262856586493552, + "language_loss": 0.55957353, + "learning_rate": 3.307847302438245e-06, + "loss": 0.58119333, + "num_input_tokens_seen": 52597075, + "step": 2453, + "time_per_iteration": 3.198324680328369 + }, + { + "auxiliary_loss_clip": 0.0121753, + "auxiliary_loss_mlp": 0.01037268, + "balance_loss_clip": 1.04776335, + "balance_loss_mlp": 1.02760637, + "epoch": 0.2950760536283292, + "flos": 16107085572480.0, + "grad_norm": 2.284077705764094, + "language_loss": 0.77890074, + "learning_rate": 3.3072578640393562e-06, + "loss": 0.8014487, + "num_input_tokens_seen": 52614410, + "step": 2454, + "time_per_iteration": 2.657956123352051 + }, + { + "auxiliary_loss_clip": 0.01218478, + "auxiliary_loss_mlp": 0.01029271, + "balance_loss_clip": 1.0553844, + "balance_loss_mlp": 1.01999092, + "epoch": 0.29519629651896834, + "flos": 20483626394880.0, + "grad_norm": 1.807414456814609, + "language_loss": 0.79527807, + "learning_rate": 3.3066682273304886e-06, + "loss": 0.81775558, + "num_input_tokens_seen": 52632055, + "step": 2455, + "time_per_iteration": 2.6951043605804443 + }, + { + "auxiliary_loss_clip": 0.01218765, + "auxiliary_loss_mlp": 0.01380526, + "balance_loss_clip": 1.05867314, + "balance_loss_mlp": 1.00006163, + "epoch": 0.2953165394096074, + "flos": 18916484941440.0, + "grad_norm": 2.7104917591656754, + "language_loss": 0.78937662, + "learning_rate": 3.3060783924010904e-06, + "loss": 0.81536955, + "num_input_tokens_seen": 52649980, + "step": 2456, + "time_per_iteration": 3.58481502532959 + }, + { + "auxiliary_loss_clip": 0.01226078, + "auxiliary_loss_mlp": 0.01035221, + "balance_loss_clip": 1.05389714, + "balance_loss_mlp": 1.02587473, + "epoch": 0.2954367823002465, + "flos": 20624467622400.0, + "grad_norm": 2.210100781905182, + "language_loss": 0.85125065, + "learning_rate": 3.3054883593406387e-06, + "loss": 0.87386358, + "num_input_tokens_seen": 52664730, + "step": 2457, + "time_per_iteration": 2.746699571609497 + }, + { + "auxiliary_loss_clip": 0.01219792, + "auxiliary_loss_mlp": 0.01035911, + "balance_loss_clip": 1.05377293, + "balance_loss_mlp": 1.0271256, + "epoch": 0.2955570251908856, + "flos": 31175525473920.0, + "grad_norm": 2.209411431112123, + "language_loss": 0.65105057, + "learning_rate": 3.3048981282386404e-06, + "loss": 0.67360753, + "num_input_tokens_seen": 52686040, + "step": 2458, + "time_per_iteration": 3.7138710021972656 + }, + { + "auxiliary_loss_clip": 0.01212396, + "auxiliary_loss_mlp": 0.01037645, + "balance_loss_clip": 1.05358458, + "balance_loss_mlp": 1.02856696, + "epoch": 0.29567726808152467, + "flos": 21650328051840.0, + "grad_norm": 3.384657440247524, + "language_loss": 0.82483846, + "learning_rate": 3.304307699184634e-06, + "loss": 0.84733886, + "num_input_tokens_seen": 52704630, + "step": 2459, + "time_per_iteration": 3.674903392791748 + }, + { + "auxiliary_loss_clip": 0.01224509, + "auxiliary_loss_mlp": 0.0103666, + "balance_loss_clip": 1.06042969, + "balance_loss_mlp": 1.02780867, + "epoch": 0.2957975109721638, + "flos": 24243868638720.0, + "grad_norm": 1.70124474359242, + "language_loss": 0.7877835, + "learning_rate": 3.3037170722681866e-06, + "loss": 0.81039524, + "num_input_tokens_seen": 52725465, + "step": 2460, + "time_per_iteration": 2.73710298538208 + }, + { + "auxiliary_loss_clip": 0.01217842, + "auxiliary_loss_mlp": 0.01031605, + "balance_loss_clip": 1.05470645, + "balance_loss_mlp": 1.02231908, + "epoch": 0.29591775386280283, + "flos": 13479717352320.0, + "grad_norm": 2.131203258878132, + "language_loss": 0.68274218, + "learning_rate": 3.3031262475788956e-06, + "loss": 0.70523667, + "num_input_tokens_seen": 52742405, + "step": 2461, + "time_per_iteration": 2.744605779647827 + }, + { + "auxiliary_loss_clip": 0.01212871, + "auxiliary_loss_mlp": 0.01033206, + "balance_loss_clip": 1.05238962, + "balance_loss_mlp": 1.02427125, + "epoch": 0.29603799675344195, + "flos": 17749783284480.0, + "grad_norm": 2.1060443378350224, + "language_loss": 0.73101616, + "learning_rate": 3.3025352252063897e-06, + "loss": 0.75347692, + "num_input_tokens_seen": 52761100, + "step": 2462, + "time_per_iteration": 2.6372525691986084 + }, + { + "auxiliary_loss_clip": 0.01210428, + "auxiliary_loss_mlp": 0.01034688, + "balance_loss_clip": 1.05975914, + "balance_loss_mlp": 1.02499032, + "epoch": 0.29615823964408106, + "flos": 22783920347520.0, + "grad_norm": 1.6645675348863598, + "language_loss": 0.74913961, + "learning_rate": 3.3019440052403252e-06, + "loss": 0.77159077, + "num_input_tokens_seen": 52780965, + "step": 2463, + "time_per_iteration": 2.7092883586883545 + }, + { + "auxiliary_loss_clip": 0.01220315, + "auxiliary_loss_mlp": 0.01031785, + "balance_loss_clip": 1.05557156, + "balance_loss_mlp": 1.02282619, + "epoch": 0.2962784825347201, + "flos": 23514199758720.0, + "grad_norm": 4.50921847397841, + "language_loss": 0.70748335, + "learning_rate": 3.30135258777039e-06, + "loss": 0.73000431, + "num_input_tokens_seen": 52800335, + "step": 2464, + "time_per_iteration": 3.6058261394500732 + }, + { + "auxiliary_loss_clip": 0.01212761, + "auxiliary_loss_mlp": 0.01379922, + "balance_loss_clip": 1.05495954, + "balance_loss_mlp": 1.0000453, + "epoch": 0.2963987254253592, + "flos": 16362769559040.0, + "grad_norm": 1.852181376747576, + "language_loss": 0.70496106, + "learning_rate": 3.3007609728863024e-06, + "loss": 0.73088789, + "num_input_tokens_seen": 52818425, + "step": 2465, + "time_per_iteration": 2.6391849517822266 + }, + { + "auxiliary_loss_clip": 0.01236805, + "auxiliary_loss_mlp": 0.0102652, + "balance_loss_clip": 1.05312669, + "balance_loss_mlp": 1.01729941, + "epoch": 0.29651896831599833, + "flos": 33472263980160.0, + "grad_norm": 1.743494772366078, + "language_loss": 0.73381162, + "learning_rate": 3.300169160677809e-06, + "loss": 0.75644493, + "num_input_tokens_seen": 52842340, + "step": 2466, + "time_per_iteration": 2.8674521446228027 + }, + { + "auxiliary_loss_clip": 0.01235902, + "auxiliary_loss_mlp": 0.01039021, + "balance_loss_clip": 1.05628502, + "balance_loss_mlp": 1.0292877, + "epoch": 0.2966392112066374, + "flos": 23805363404160.0, + "grad_norm": 2.6753292899104473, + "language_loss": 0.77581495, + "learning_rate": 3.2995771512346878e-06, + "loss": 0.7985642, + "num_input_tokens_seen": 52860690, + "step": 2467, + "time_per_iteration": 2.6436822414398193 + }, + { + "auxiliary_loss_clip": 0.01205573, + "auxiliary_loss_mlp": 0.0138006, + "balance_loss_clip": 1.060583, + "balance_loss_mlp": 1.00010276, + "epoch": 0.2967594540972765, + "flos": 19938466702080.0, + "grad_norm": 2.071601604562695, + "language_loss": 0.73039234, + "learning_rate": 3.298984944646746e-06, + "loss": 0.75624871, + "num_input_tokens_seen": 52879370, + "step": 2468, + "time_per_iteration": 2.5362823009490967 + }, + { + "auxiliary_loss_clip": 0.01217753, + "auxiliary_loss_mlp": 0.01379632, + "balance_loss_clip": 1.0629524, + "balance_loss_mlp": 1.00005889, + "epoch": 0.2968796969879156, + "flos": 23732823888000.0, + "grad_norm": 3.3213043980275585, + "language_loss": 0.8152923, + "learning_rate": 3.298392541003822e-06, + "loss": 0.84126616, + "num_input_tokens_seen": 52898775, + "step": 2469, + "time_per_iteration": 2.6370272636413574 + }, + { + "auxiliary_loss_clip": 0.01218273, + "auxiliary_loss_mlp": 0.01029965, + "balance_loss_clip": 1.05550349, + "balance_loss_mlp": 1.02098894, + "epoch": 0.29699993987855466, + "flos": 22893699288960.0, + "grad_norm": 1.9163638241648038, + "language_loss": 0.8997376, + "learning_rate": 3.2977999403957806e-06, + "loss": 0.92221999, + "num_input_tokens_seen": 52917535, + "step": 2470, + "time_per_iteration": 2.684574842453003 + }, + { + "auxiliary_loss_clip": 0.01202803, + "auxiliary_loss_mlp": 0.01033267, + "balance_loss_clip": 1.06032896, + "balance_loss_mlp": 1.0235393, + "epoch": 0.2971201827691938, + "flos": 33832555349760.0, + "grad_norm": 7.6458533944819616, + "language_loss": 0.67229986, + "learning_rate": 3.2972071429125207e-06, + "loss": 0.69466054, + "num_input_tokens_seen": 52938755, + "step": 2471, + "time_per_iteration": 2.6958255767822266 + }, + { + "auxiliary_loss_clip": 0.01223865, + "auxiliary_loss_mlp": 0.01031528, + "balance_loss_clip": 1.0547905, + "balance_loss_mlp": 1.02217007, + "epoch": 0.2972404256598329, + "flos": 22054359208320.0, + "grad_norm": 2.062942137757452, + "language_loss": 0.88275361, + "learning_rate": 3.2966141486439682e-06, + "loss": 0.90530747, + "num_input_tokens_seen": 52957945, + "step": 2472, + "time_per_iteration": 2.7460601329803467 + }, + { + "auxiliary_loss_clip": 0.01243412, + "auxiliary_loss_mlp": 0.01033857, + "balance_loss_clip": 1.04939711, + "balance_loss_mlp": 1.02451706, + "epoch": 0.29736066855047194, + "flos": 31978595796480.0, + "grad_norm": 2.949221893849347, + "language_loss": 0.64434034, + "learning_rate": 3.29602095768008e-06, + "loss": 0.66711307, + "num_input_tokens_seen": 52978460, + "step": 2473, + "time_per_iteration": 2.8135077953338623 + }, + { + "auxiliary_loss_clip": 0.01212827, + "auxiliary_loss_mlp": 0.01031062, + "balance_loss_clip": 1.05714774, + "balance_loss_mlp": 1.02214527, + "epoch": 0.29748091144111105, + "flos": 33510401245440.0, + "grad_norm": 1.9269300465318984, + "language_loss": 0.63657564, + "learning_rate": 3.2954275701108437e-06, + "loss": 0.65901452, + "num_input_tokens_seen": 52999640, + "step": 2474, + "time_per_iteration": 2.749992847442627 + }, + { + "auxiliary_loss_clip": 0.0123288, + "auxiliary_loss_mlp": 0.01035362, + "balance_loss_clip": 1.053195, + "balance_loss_mlp": 1.02577138, + "epoch": 0.29760115433175016, + "flos": 41283373409280.0, + "grad_norm": 1.9655205166703849, + "language_loss": 0.68870497, + "learning_rate": 3.294833986026275e-06, + "loss": 0.7113874, + "num_input_tokens_seen": 53022880, + "step": 2475, + "time_per_iteration": 2.9124903678894043 + }, + { + "auxiliary_loss_clip": 0.01224053, + "auxiliary_loss_mlp": 0.01036285, + "balance_loss_clip": 1.056288, + "balance_loss_mlp": 1.02761865, + "epoch": 0.2977213972223892, + "flos": 24493339572480.0, + "grad_norm": 1.9020884723716964, + "language_loss": 0.85318494, + "learning_rate": 3.29424020551642e-06, + "loss": 0.87578833, + "num_input_tokens_seen": 53041515, + "step": 2476, + "time_per_iteration": 2.7448134422302246 + }, + { + "auxiliary_loss_clip": 0.01209252, + "auxiliary_loss_mlp": 0.0104139, + "balance_loss_clip": 1.0633297, + "balance_loss_mlp": 1.03193665, + "epoch": 0.2978416401130283, + "flos": 21285116519040.0, + "grad_norm": 2.284183761059837, + "language_loss": 0.72071576, + "learning_rate": 3.2936462286713546e-06, + "loss": 0.74322218, + "num_input_tokens_seen": 53059865, + "step": 2477, + "time_per_iteration": 2.5983593463897705 + }, + { + "auxiliary_loss_clip": 0.01212992, + "auxiliary_loss_mlp": 0.01031164, + "balance_loss_clip": 1.05935705, + "balance_loss_mlp": 1.02180648, + "epoch": 0.2979618830036674, + "flos": 25772154554880.0, + "grad_norm": 1.8912574334219188, + "language_loss": 0.7722367, + "learning_rate": 3.2930520555811846e-06, + "loss": 0.79467827, + "num_input_tokens_seen": 53079490, + "step": 2478, + "time_per_iteration": 2.6733498573303223 + }, + { + "auxiliary_loss_clip": 0.01227491, + "auxiliary_loss_mlp": 0.0138037, + "balance_loss_clip": 1.04645479, + "balance_loss_mlp": 1.00003934, + "epoch": 0.2980821258943065, + "flos": 23476996247040.0, + "grad_norm": 1.878689969527119, + "language_loss": 0.80277884, + "learning_rate": 3.292457686336046e-06, + "loss": 0.82885742, + "num_input_tokens_seen": 53098810, + "step": 2479, + "time_per_iteration": 2.9046051502227783 + }, + { + "auxiliary_loss_clip": 0.01135262, + "auxiliary_loss_mlp": 0.01006755, + "balance_loss_clip": 1.03504133, + "balance_loss_mlp": 1.00457382, + "epoch": 0.2982023687849456, + "flos": 69752314195200.0, + "grad_norm": 0.8641097136367936, + "language_loss": 0.61241555, + "learning_rate": 3.291863121026105e-06, + "loss": 0.63383573, + "num_input_tokens_seen": 53162590, + "step": 2480, + "time_per_iteration": 3.4919204711914062 + }, + { + "auxiliary_loss_clip": 0.01215815, + "auxiliary_loss_mlp": 0.01038956, + "balance_loss_clip": 1.06114745, + "balance_loss_mlp": 1.02953875, + "epoch": 0.29832261167558466, + "flos": 29825930741760.0, + "grad_norm": 2.4134012109660223, + "language_loss": 0.7709614, + "learning_rate": 3.2912683597415547e-06, + "loss": 0.79350919, + "num_input_tokens_seen": 53186675, + "step": 2481, + "time_per_iteration": 2.757460832595825 + }, + { + "auxiliary_loss_clip": 0.01231857, + "auxiliary_loss_mlp": 0.01029918, + "balance_loss_clip": 1.05757427, + "balance_loss_mlp": 1.02110243, + "epoch": 0.29844285456622377, + "flos": 33910158683520.0, + "grad_norm": 1.9972446822227905, + "language_loss": 0.78255451, + "learning_rate": 3.2906734025726213e-06, + "loss": 0.80517232, + "num_input_tokens_seen": 53205940, + "step": 2482, + "time_per_iteration": 3.7284090518951416 + }, + { + "auxiliary_loss_clip": 0.01218673, + "auxiliary_loss_mlp": 0.01030266, + "balance_loss_clip": 1.06174481, + "balance_loss_mlp": 1.0196867, + "epoch": 0.2985630974568629, + "flos": 23876933253120.0, + "grad_norm": 1.9561153811917176, + "language_loss": 0.87941802, + "learning_rate": 3.290078249609559e-06, + "loss": 0.90190744, + "num_input_tokens_seen": 53225360, + "step": 2483, + "time_per_iteration": 2.644796133041382 + }, + { + "auxiliary_loss_clip": 0.01213556, + "auxiliary_loss_mlp": 0.01038331, + "balance_loss_clip": 1.06315732, + "balance_loss_mlp": 1.02945566, + "epoch": 0.29868334034750194, + "flos": 21799106184960.0, + "grad_norm": 1.9847465249713352, + "language_loss": 0.87847161, + "learning_rate": 3.2894829009426514e-06, + "loss": 0.90099049, + "num_input_tokens_seen": 53243195, + "step": 2484, + "time_per_iteration": 3.5639820098876953 + }, + { + "auxiliary_loss_clip": 0.01212282, + "auxiliary_loss_mlp": 0.01027716, + "balance_loss_clip": 1.06135798, + "balance_loss_mlp": 1.01894808, + "epoch": 0.29880358323814105, + "flos": 25666649331840.0, + "grad_norm": 1.972495446239297, + "language_loss": 0.77761352, + "learning_rate": 3.288887356662213e-06, + "loss": 0.80001342, + "num_input_tokens_seen": 53264530, + "step": 2485, + "time_per_iteration": 2.717129945755005 + }, + { + "auxiliary_loss_clip": 0.01116731, + "auxiliary_loss_mlp": 0.01002109, + "balance_loss_clip": 1.03457499, + "balance_loss_mlp": 0.99995154, + "epoch": 0.29892382612878016, + "flos": 71005846003200.0, + "grad_norm": 0.772299289089044, + "language_loss": 0.59691155, + "learning_rate": 3.288291616858588e-06, + "loss": 0.61809993, + "num_input_tokens_seen": 53319920, + "step": 2486, + "time_per_iteration": 3.0943639278411865 + }, + { + "auxiliary_loss_clip": 0.01233586, + "auxiliary_loss_mlp": 0.01034773, + "balance_loss_clip": 1.05752814, + "balance_loss_mlp": 1.02640486, + "epoch": 0.2990440690194192, + "flos": 25481134563840.0, + "grad_norm": 1.7706889874262783, + "language_loss": 0.76888466, + "learning_rate": 3.287695681622149e-06, + "loss": 0.79156828, + "num_input_tokens_seen": 53339270, + "step": 2487, + "time_per_iteration": 2.985262155532837 + }, + { + "auxiliary_loss_clip": 0.01228762, + "auxiliary_loss_mlp": 0.01027771, + "balance_loss_clip": 1.06058669, + "balance_loss_mlp": 1.01936102, + "epoch": 0.2991643119100583, + "flos": 23732357011200.0, + "grad_norm": 1.8083072243924003, + "language_loss": 0.80817378, + "learning_rate": 3.2870995510432982e-06, + "loss": 0.83073914, + "num_input_tokens_seen": 53357750, + "step": 2488, + "time_per_iteration": 2.7861900329589844 + }, + { + "auxiliary_loss_clip": 0.01204204, + "auxiliary_loss_mlp": 0.01031456, + "balance_loss_clip": 1.05896759, + "balance_loss_mlp": 1.02248514, + "epoch": 0.29928455480069743, + "flos": 27417545786880.0, + "grad_norm": 1.7749849701174993, + "language_loss": 0.76538169, + "learning_rate": 3.2865032252124697e-06, + "loss": 0.78773832, + "num_input_tokens_seen": 53378265, + "step": 2489, + "time_per_iteration": 2.6855616569519043 + }, + { + "auxiliary_loss_clip": 0.01217776, + "auxiliary_loss_mlp": 0.01032818, + "balance_loss_clip": 1.0575105, + "balance_loss_mlp": 1.02470005, + "epoch": 0.2994047976913365, + "flos": 33692935184640.0, + "grad_norm": 2.2702197402473776, + "language_loss": 0.77372992, + "learning_rate": 3.2859067042201243e-06, + "loss": 0.79623586, + "num_input_tokens_seen": 53400305, + "step": 2490, + "time_per_iteration": 3.6433932781219482 + }, + { + "auxiliary_loss_clip": 0.01219781, + "auxiliary_loss_mlp": 0.01025546, + "balance_loss_clip": 1.05005932, + "balance_loss_mlp": 1.01708817, + "epoch": 0.2995250405819756, + "flos": 16763963541120.0, + "grad_norm": 1.9659575323099951, + "language_loss": 0.778036, + "learning_rate": 3.2853099881567544e-06, + "loss": 0.80048925, + "num_input_tokens_seen": 53418705, + "step": 2491, + "time_per_iteration": 2.7573671340942383 + }, + { + "auxiliary_loss_clip": 0.01202349, + "auxiliary_loss_mlp": 0.01027162, + "balance_loss_clip": 1.06284022, + "balance_loss_mlp": 1.01857877, + "epoch": 0.29964528347261465, + "flos": 22963976248320.0, + "grad_norm": 2.674893458355977, + "language_loss": 0.79472649, + "learning_rate": 3.284713077112881e-06, + "loss": 0.81702155, + "num_input_tokens_seen": 53438135, + "step": 2492, + "time_per_iteration": 2.6841835975646973 + }, + { + "auxiliary_loss_clip": 0.01237975, + "auxiliary_loss_mlp": 0.01030281, + "balance_loss_clip": 1.06137478, + "balance_loss_mlp": 1.02083397, + "epoch": 0.29976552636325376, + "flos": 16938021870720.0, + "grad_norm": 3.0026408165505414, + "language_loss": 0.86560059, + "learning_rate": 3.284115971179056e-06, + "loss": 0.88828313, + "num_input_tokens_seen": 53452165, + "step": 2493, + "time_per_iteration": 2.679147958755493 + }, + { + "auxiliary_loss_clip": 0.01248514, + "auxiliary_loss_mlp": 0.0102799, + "balance_loss_clip": 1.05515885, + "balance_loss_mlp": 1.01925755, + "epoch": 0.2998857692538929, + "flos": 17056455989760.0, + "grad_norm": 1.8867608330976178, + "language_loss": 0.78543186, + "learning_rate": 3.283518670445859e-06, + "loss": 0.8081969, + "num_input_tokens_seen": 53470075, + "step": 2494, + "time_per_iteration": 2.806652784347534 + }, + { + "auxiliary_loss_clip": 0.01120277, + "auxiliary_loss_mlp": 0.01375218, + "balance_loss_clip": 1.02976418, + "balance_loss_mlp": 0.9998011, + "epoch": 0.30000601214453193, + "flos": 68831528025600.0, + "grad_norm": 0.7227835387638678, + "language_loss": 0.54306728, + "learning_rate": 3.2829211750038995e-06, + "loss": 0.56802225, + "num_input_tokens_seen": 53538705, + "step": 2495, + "time_per_iteration": 3.2909882068634033 + }, + { + "auxiliary_loss_clip": 0.01221299, + "auxiliary_loss_mlp": 0.01031109, + "balance_loss_clip": 1.05305958, + "balance_loss_mlp": 1.0224781, + "epoch": 0.30012625503517104, + "flos": 17603267708160.0, + "grad_norm": 3.214666206923121, + "language_loss": 0.89048803, + "learning_rate": 3.2823234849438183e-06, + "loss": 0.91301215, + "num_input_tokens_seen": 53556740, + "step": 2496, + "time_per_iteration": 2.6771655082702637 + }, + { + "auxiliary_loss_clip": 0.01226518, + "auxiliary_loss_mlp": 0.01025861, + "balance_loss_clip": 1.06132078, + "balance_loss_mlp": 1.01697421, + "epoch": 0.30024649792581015, + "flos": 21252581775360.0, + "grad_norm": 2.565178589269314, + "language_loss": 0.75540721, + "learning_rate": 3.2817256003562836e-06, + "loss": 0.77793097, + "num_input_tokens_seen": 53577115, + "step": 2497, + "time_per_iteration": 2.699556589126587 + }, + { + "auxiliary_loss_clip": 0.01247186, + "auxiliary_loss_mlp": 0.01028545, + "balance_loss_clip": 1.05353558, + "balance_loss_mlp": 1.01993227, + "epoch": 0.3003667408164492, + "flos": 23003262748800.0, + "grad_norm": 1.7344409423958744, + "language_loss": 0.65975571, + "learning_rate": 3.281127521331995e-06, + "loss": 0.682513, + "num_input_tokens_seen": 53598295, + "step": 2498, + "time_per_iteration": 2.814861297607422 + }, + { + "auxiliary_loss_clip": 0.01105016, + "auxiliary_loss_mlp": 0.01009629, + "balance_loss_clip": 1.03243041, + "balance_loss_mlp": 1.00757813, + "epoch": 0.3004869837070883, + "flos": 64232340750720.0, + "grad_norm": 0.8830431842828873, + "language_loss": 0.60672915, + "learning_rate": 3.2805292479616798e-06, + "loss": 0.62787563, + "num_input_tokens_seen": 53657160, + "step": 2499, + "time_per_iteration": 3.151538372039795 + }, + { + "auxiliary_loss_clip": 0.01222458, + "auxiliary_loss_mlp": 0.01034145, + "balance_loss_clip": 1.05974865, + "balance_loss_mlp": 1.02532363, + "epoch": 0.30060722659772743, + "flos": 26248653400320.0, + "grad_norm": 2.5939974267371313, + "language_loss": 0.92108124, + "learning_rate": 3.2799307803360955e-06, + "loss": 0.94364727, + "num_input_tokens_seen": 53673090, + "step": 2500, + "time_per_iteration": 2.6779556274414062 + }, + { + "auxiliary_loss_clip": 0.01199807, + "auxiliary_loss_mlp": 0.01026878, + "balance_loss_clip": 1.06071067, + "balance_loss_mlp": 1.01834822, + "epoch": 0.3007274694883665, + "flos": 24970879912320.0, + "grad_norm": 1.4072846161446702, + "language_loss": 0.81421149, + "learning_rate": 3.27933211854603e-06, + "loss": 0.83647835, + "num_input_tokens_seen": 53692145, + "step": 2501, + "time_per_iteration": 2.654597759246826 + }, + { + "auxiliary_loss_clip": 0.01222863, + "auxiliary_loss_mlp": 0.01033176, + "balance_loss_clip": 1.06052065, + "balance_loss_mlp": 1.02419996, + "epoch": 0.3008477123790056, + "flos": 17055845458560.0, + "grad_norm": 1.605236383211186, + "language_loss": 0.86683404, + "learning_rate": 3.278733262682299e-06, + "loss": 0.88939452, + "num_input_tokens_seen": 53710000, + "step": 2502, + "time_per_iteration": 2.587254047393799 + }, + { + "auxiliary_loss_clip": 0.0120162, + "auxiliary_loss_mlp": 0.01026506, + "balance_loss_clip": 1.05867147, + "balance_loss_mlp": 1.01813745, + "epoch": 0.3009679552696447, + "flos": 21506398254720.0, + "grad_norm": 2.359402702316474, + "language_loss": 0.82610464, + "learning_rate": 3.2781342128357484e-06, + "loss": 0.84838593, + "num_input_tokens_seen": 53729355, + "step": 2503, + "time_per_iteration": 2.607727289199829 + }, + { + "auxiliary_loss_clip": 0.01225674, + "auxiliary_loss_mlp": 0.01028063, + "balance_loss_clip": 1.05382991, + "balance_loss_mlp": 1.01959276, + "epoch": 0.30108819816028376, + "flos": 21134004001920.0, + "grad_norm": 2.3237249805348896, + "language_loss": 0.8055625, + "learning_rate": 3.2775349690972547e-06, + "loss": 0.82809985, + "num_input_tokens_seen": 53743505, + "step": 2504, + "time_per_iteration": 2.65681529045105 + }, + { + "auxiliary_loss_clip": 0.01108522, + "auxiliary_loss_mlp": 0.01002179, + "balance_loss_clip": 1.0286442, + "balance_loss_mlp": 1.00017631, + "epoch": 0.30120844105092287, + "flos": 71126434938240.0, + "grad_norm": 0.7761315199020282, + "language_loss": 0.51867783, + "learning_rate": 3.276935531557722e-06, + "loss": 0.53978485, + "num_input_tokens_seen": 53808725, + "step": 2505, + "time_per_iteration": 3.26265811920166 + }, + { + "auxiliary_loss_clip": 0.01240331, + "auxiliary_loss_mlp": 0.01026368, + "balance_loss_clip": 1.0554378, + "balance_loss_mlp": 1.01733232, + "epoch": 0.301328683941562, + "flos": 20264571302400.0, + "grad_norm": 2.1457171792717533, + "language_loss": 0.79357266, + "learning_rate": 3.2763359003080837e-06, + "loss": 0.81623971, + "num_input_tokens_seen": 53825680, + "step": 2506, + "time_per_iteration": 2.730485439300537 + }, + { + "auxiliary_loss_clip": 0.01126456, + "auxiliary_loss_mlp": 0.0100433, + "balance_loss_clip": 1.02858233, + "balance_loss_mlp": 1.00227928, + "epoch": 0.30144892683220104, + "flos": 70648212240000.0, + "grad_norm": 0.8013330160700936, + "language_loss": 0.62464476, + "learning_rate": 3.2757360754393047e-06, + "loss": 0.64595264, + "num_input_tokens_seen": 53889750, + "step": 2507, + "time_per_iteration": 3.283052444458008 + }, + { + "auxiliary_loss_clip": 0.01210142, + "auxiliary_loss_mlp": 0.01024788, + "balance_loss_clip": 1.05959809, + "balance_loss_mlp": 1.01617527, + "epoch": 0.30156916972284015, + "flos": 22820549241600.0, + "grad_norm": 3.381230222181369, + "language_loss": 0.63548142, + "learning_rate": 3.2751360570423767e-06, + "loss": 0.65783072, + "num_input_tokens_seen": 53908135, + "step": 2508, + "time_per_iteration": 3.6357476711273193 + }, + { + "auxiliary_loss_clip": 0.01217043, + "auxiliary_loss_mlp": 0.01033374, + "balance_loss_clip": 1.05717194, + "balance_loss_mlp": 1.02446866, + "epoch": 0.3016894126134792, + "flos": 29899188529920.0, + "grad_norm": 2.1759909497932104, + "language_loss": 0.75948954, + "learning_rate": 3.2745358452083236e-06, + "loss": 0.78199363, + "num_input_tokens_seen": 53931035, + "step": 2509, + "time_per_iteration": 2.7086408138275146 + }, + { + "auxiliary_loss_clip": 0.01212011, + "auxiliary_loss_mlp": 0.01033645, + "balance_loss_clip": 1.05974007, + "balance_loss_mlp": 1.02537799, + "epoch": 0.3018096555041183, + "flos": 21546331200000.0, + "grad_norm": 1.431392809881689, + "language_loss": 0.82326353, + "learning_rate": 3.2739354400281955e-06, + "loss": 0.84572017, + "num_input_tokens_seen": 53952255, + "step": 2510, + "time_per_iteration": 3.615422248840332 + }, + { + "auxiliary_loss_clip": 0.01133422, + "auxiliary_loss_mlp": 0.01375298, + "balance_loss_clip": 1.02478337, + "balance_loss_mlp": 0.99981374, + "epoch": 0.3019298983947574, + "flos": 59136294597120.0, + "grad_norm": 0.8859019930585876, + "language_loss": 0.63688493, + "learning_rate": 3.2733348415930744e-06, + "loss": 0.66197217, + "num_input_tokens_seen": 54014125, + "step": 2511, + "time_per_iteration": 3.300745964050293 + }, + { + "auxiliary_loss_clip": 0.01223605, + "auxiliary_loss_mlp": 0.01026162, + "balance_loss_clip": 1.05582142, + "balance_loss_mlp": 1.01799572, + "epoch": 0.3020501412853965, + "flos": 34423070941440.0, + "grad_norm": 3.7300352030480193, + "language_loss": 0.80488288, + "learning_rate": 3.27273404999407e-06, + "loss": 0.82738054, + "num_input_tokens_seen": 54036345, + "step": 2512, + "time_per_iteration": 2.7873520851135254 + }, + { + "auxiliary_loss_clip": 0.01122881, + "auxiliary_loss_mlp": 0.01000228, + "balance_loss_clip": 1.02574229, + "balance_loss_mlp": 0.99823725, + "epoch": 0.3021703841760356, + "flos": 71008288128000.0, + "grad_norm": 0.8079357419291955, + "language_loss": 0.60508943, + "learning_rate": 3.272133065322322e-06, + "loss": 0.62632048, + "num_input_tokens_seen": 54094615, + "step": 2513, + "time_per_iteration": 3.2163193225860596 + }, + { + "auxiliary_loss_clip": 0.01195737, + "auxiliary_loss_mlp": 0.01032622, + "balance_loss_clip": 1.05603051, + "balance_loss_mlp": 1.02383602, + "epoch": 0.3022906270666747, + "flos": 21510528318720.0, + "grad_norm": 1.7119293942288878, + "language_loss": 0.79250634, + "learning_rate": 3.271531887669e-06, + "loss": 0.81478989, + "num_input_tokens_seen": 54114675, + "step": 2514, + "time_per_iteration": 2.6433675289154053 + }, + { + "auxiliary_loss_clip": 0.01238913, + "auxiliary_loss_mlp": 0.0102566, + "balance_loss_clip": 1.051579, + "balance_loss_mlp": 1.01632547, + "epoch": 0.30241086995731375, + "flos": 31132001168640.0, + "grad_norm": 2.229888572555217, + "language_loss": 0.62984252, + "learning_rate": 3.2709305171253015e-06, + "loss": 0.65248823, + "num_input_tokens_seen": 54134795, + "step": 2515, + "time_per_iteration": 3.690758466720581 + }, + { + "auxiliary_loss_clip": 0.01209138, + "auxiliary_loss_mlp": 0.01034903, + "balance_loss_clip": 1.05952656, + "balance_loss_mlp": 1.02617049, + "epoch": 0.30253111284795287, + "flos": 23511542152320.0, + "grad_norm": 1.9610276644444053, + "language_loss": 0.77757037, + "learning_rate": 3.2703289537824536e-06, + "loss": 0.80001068, + "num_input_tokens_seen": 54154595, + "step": 2516, + "time_per_iteration": 2.6826701164245605 + }, + { + "auxiliary_loss_clip": 0.01233338, + "auxiliary_loss_mlp": 0.01032747, + "balance_loss_clip": 1.0499177, + "balance_loss_mlp": 1.02399683, + "epoch": 0.302651355738592, + "flos": 18725367651840.0, + "grad_norm": 2.790861640384183, + "language_loss": 0.78055, + "learning_rate": 3.269727197731714e-06, + "loss": 0.80321085, + "num_input_tokens_seen": 54167360, + "step": 2517, + "time_per_iteration": 2.667639970779419 + }, + { + "auxiliary_loss_clip": 0.01225172, + "auxiliary_loss_mlp": 0.01026926, + "balance_loss_clip": 1.05227685, + "balance_loss_mlp": 1.01806831, + "epoch": 0.30277159862923103, + "flos": 22418888382720.0, + "grad_norm": 1.6857493856142947, + "language_loss": 0.77875459, + "learning_rate": 3.269125249064367e-06, + "loss": 0.80127561, + "num_input_tokens_seen": 54187055, + "step": 2518, + "time_per_iteration": 2.750701904296875 + }, + { + "auxiliary_loss_clip": 0.0120446, + "auxiliary_loss_mlp": 0.01028848, + "balance_loss_clip": 1.06000757, + "balance_loss_mlp": 1.0194068, + "epoch": 0.30289184151987014, + "flos": 22273126992000.0, + "grad_norm": 1.6373143431311363, + "language_loss": 0.83202946, + "learning_rate": 3.2685231078717297e-06, + "loss": 0.85436261, + "num_input_tokens_seen": 54207245, + "step": 2519, + "time_per_iteration": 2.5966691970825195 + }, + { + "auxiliary_loss_clip": 0.01216552, + "auxiliary_loss_mlp": 0.01379864, + "balance_loss_clip": 1.05404627, + "balance_loss_mlp": 1.00003421, + "epoch": 0.30301208441050925, + "flos": 25225594231680.0, + "grad_norm": 3.6854539208179196, + "language_loss": 0.75619566, + "learning_rate": 3.267920774245145e-06, + "loss": 0.78215981, + "num_input_tokens_seen": 54226650, + "step": 2520, + "time_per_iteration": 2.884669780731201 + }, + { + "auxiliary_loss_clip": 0.01211608, + "auxiliary_loss_mlp": 0.01033941, + "balance_loss_clip": 1.06085777, + "balance_loss_mlp": 1.02445805, + "epoch": 0.3031323273011483, + "flos": 23039245198080.0, + "grad_norm": 1.712914277731595, + "language_loss": 0.8433429, + "learning_rate": 3.2673182482759876e-06, + "loss": 0.86579835, + "num_input_tokens_seen": 54245765, + "step": 2521, + "time_per_iteration": 2.711239814758301 + }, + { + "auxiliary_loss_clip": 0.01211523, + "auxiliary_loss_mlp": 0.010362, + "balance_loss_clip": 1.06019044, + "balance_loss_mlp": 1.02731884, + "epoch": 0.3032525701917874, + "flos": 18876695650560.0, + "grad_norm": 1.7741927837009388, + "language_loss": 0.66231918, + "learning_rate": 3.266715530055659e-06, + "loss": 0.68479633, + "num_input_tokens_seen": 54263915, + "step": 2522, + "time_per_iteration": 2.695819854736328 + }, + { + "auxiliary_loss_clip": 0.01195474, + "auxiliary_loss_mlp": 0.01026347, + "balance_loss_clip": 1.05412042, + "balance_loss_mlp": 1.01738286, + "epoch": 0.30337281308242653, + "flos": 17782641250560.0, + "grad_norm": 1.6369372453008475, + "language_loss": 0.80659503, + "learning_rate": 3.2661126196755927e-06, + "loss": 0.82881325, + "num_input_tokens_seen": 54283025, + "step": 2523, + "time_per_iteration": 2.62471866607666 + }, + { + "auxiliary_loss_clip": 0.01097061, + "auxiliary_loss_mlp": 0.01004609, + "balance_loss_clip": 1.02449512, + "balance_loss_mlp": 1.00251055, + "epoch": 0.3034930559730656, + "flos": 57824298426240.0, + "grad_norm": 0.7773054726721672, + "language_loss": 0.55934751, + "learning_rate": 3.265509517227248e-06, + "loss": 0.58036423, + "num_input_tokens_seen": 54339840, + "step": 2524, + "time_per_iteration": 3.2297523021698 + }, + { + "auxiliary_loss_clip": 0.01212657, + "auxiliary_loss_mlp": 0.01031615, + "balance_loss_clip": 1.05079842, + "balance_loss_mlp": 1.02306128, + "epoch": 0.3036132988637047, + "flos": 14755587419520.0, + "grad_norm": 3.3815907498192117, + "language_loss": 0.81363416, + "learning_rate": 3.264906222802115e-06, + "loss": 0.83607686, + "num_input_tokens_seen": 54357690, + "step": 2525, + "time_per_iteration": 2.670013427734375 + }, + { + "auxiliary_loss_clip": 0.01202245, + "auxiliary_loss_mlp": 0.01034796, + "balance_loss_clip": 1.05860496, + "balance_loss_mlp": 1.02556336, + "epoch": 0.30373354175434375, + "flos": 21033203460480.0, + "grad_norm": 2.7458873118899234, + "language_loss": 0.78474468, + "learning_rate": 3.264302736491715e-06, + "loss": 0.80711508, + "num_input_tokens_seen": 54377810, + "step": 2526, + "time_per_iteration": 2.633845090866089 + }, + { + "auxiliary_loss_clip": 0.01201193, + "auxiliary_loss_mlp": 0.010282, + "balance_loss_clip": 1.05500317, + "balance_loss_mlp": 1.0192709, + "epoch": 0.30385378464498286, + "flos": 21143233797120.0, + "grad_norm": 1.972625926736232, + "language_loss": 0.87589455, + "learning_rate": 3.263699058387594e-06, + "loss": 0.89818847, + "num_input_tokens_seen": 54395245, + "step": 2527, + "time_per_iteration": 2.6265475749969482 + }, + { + "auxiliary_loss_clip": 0.01217823, + "auxiliary_loss_mlp": 0.01034835, + "balance_loss_clip": 1.05038869, + "balance_loss_mlp": 1.0255363, + "epoch": 0.30397402753562197, + "flos": 20629244131200.0, + "grad_norm": 4.554473303256122, + "language_loss": 0.90163386, + "learning_rate": 3.2630951885813315e-06, + "loss": 0.92416042, + "num_input_tokens_seen": 54412640, + "step": 2528, + "time_per_iteration": 2.653531074523926 + }, + { + "auxiliary_loss_clip": 0.01217869, + "auxiliary_loss_mlp": 0.01034633, + "balance_loss_clip": 1.05335474, + "balance_loss_mlp": 1.02628255, + "epoch": 0.304094270426261, + "flos": 15085678429440.0, + "grad_norm": 2.004306966631651, + "language_loss": 0.78362072, + "learning_rate": 3.262491127164533e-06, + "loss": 0.80614567, + "num_input_tokens_seen": 54431455, + "step": 2529, + "time_per_iteration": 2.65470814704895 + }, + { + "auxiliary_loss_clip": 0.01222046, + "auxiliary_loss_mlp": 0.01379897, + "balance_loss_clip": 1.05396056, + "balance_loss_mlp": 1.0001092, + "epoch": 0.30421451331690014, + "flos": 13845216193920.0, + "grad_norm": 16.327430208113412, + "language_loss": 0.80293053, + "learning_rate": 3.2618868742288337e-06, + "loss": 0.82894993, + "num_input_tokens_seen": 54448380, + "step": 2530, + "time_per_iteration": 2.638547420501709 + }, + { + "auxiliary_loss_clip": 0.01208696, + "auxiliary_loss_mlp": 0.01022976, + "balance_loss_clip": 1.05765665, + "balance_loss_mlp": 1.01455998, + "epoch": 0.30433475620753925, + "flos": 17384212615680.0, + "grad_norm": 1.809136095988698, + "language_loss": 0.72818995, + "learning_rate": 3.261282429865899e-06, + "loss": 0.75050664, + "num_input_tokens_seen": 54466385, + "step": 2531, + "time_per_iteration": 2.587249994277954 + }, + { + "auxiliary_loss_clip": 0.01221085, + "auxiliary_loss_mlp": 0.01379351, + "balance_loss_clip": 1.05679417, + "balance_loss_mlp": 1.00002599, + "epoch": 0.3044549990981783, + "flos": 18916951818240.0, + "grad_norm": 1.6412833742231971, + "language_loss": 0.72445893, + "learning_rate": 3.2606777941674225e-06, + "loss": 0.75046325, + "num_input_tokens_seen": 54485040, + "step": 2532, + "time_per_iteration": 2.6883935928344727 + }, + { + "auxiliary_loss_clip": 0.01223228, + "auxiliary_loss_mlp": 0.01026611, + "balance_loss_clip": 1.0512588, + "balance_loss_mlp": 1.01786113, + "epoch": 0.3045752419888174, + "flos": 21068431724160.0, + "grad_norm": 2.102175215552908, + "language_loss": 0.84572184, + "learning_rate": 3.2600729672251276e-06, + "loss": 0.86822021, + "num_input_tokens_seen": 54502755, + "step": 2533, + "time_per_iteration": 2.8050618171691895 + }, + { + "auxiliary_loss_clip": 0.01200511, + "auxiliary_loss_mlp": 0.01379307, + "balance_loss_clip": 1.05950797, + "balance_loss_mlp": 1.00003362, + "epoch": 0.3046954848794565, + "flos": 29096405516160.0, + "grad_norm": 2.1621126325845768, + "language_loss": 0.65733886, + "learning_rate": 3.259467949130765e-06, + "loss": 0.683137, + "num_input_tokens_seen": 54524165, + "step": 2534, + "time_per_iteration": 3.5475363731384277 + }, + { + "auxiliary_loss_clip": 0.01218695, + "auxiliary_loss_mlp": 0.01025216, + "balance_loss_clip": 1.05657387, + "balance_loss_mlp": 1.01632905, + "epoch": 0.3048157277700956, + "flos": 20295346279680.0, + "grad_norm": 2.3252199465892383, + "language_loss": 0.82486796, + "learning_rate": 3.2588627399761164e-06, + "loss": 0.84730709, + "num_input_tokens_seen": 54540160, + "step": 2535, + "time_per_iteration": 2.654892683029175 + }, + { + "auxiliary_loss_clip": 0.01212991, + "auxiliary_loss_mlp": 0.01029659, + "balance_loss_clip": 1.05360615, + "balance_loss_mlp": 1.02173162, + "epoch": 0.3049359706607347, + "flos": 22739929165440.0, + "grad_norm": 1.7112686888613076, + "language_loss": 0.70776975, + "learning_rate": 3.2582573398529903e-06, + "loss": 0.7301963, + "num_input_tokens_seen": 54557515, + "step": 2536, + "time_per_iteration": 4.512095212936401 + }, + { + "auxiliary_loss_clip": 0.01222985, + "auxiliary_loss_mlp": 0.01038247, + "balance_loss_clip": 1.05366707, + "balance_loss_mlp": 1.02934194, + "epoch": 0.3050562135513738, + "flos": 18434634969600.0, + "grad_norm": 6.5250700156815515, + "language_loss": 0.74399889, + "learning_rate": 3.2576517488532265e-06, + "loss": 0.76661122, + "num_input_tokens_seen": 54573865, + "step": 2537, + "time_per_iteration": 2.730543851852417 + }, + { + "auxiliary_loss_clip": 0.01205881, + "auxiliary_loss_mlp": 0.01033568, + "balance_loss_clip": 1.05486727, + "balance_loss_mlp": 1.02517009, + "epoch": 0.30517645644201286, + "flos": 20370327920640.0, + "grad_norm": 2.1532246486823343, + "language_loss": 0.87518758, + "learning_rate": 3.257045967068692e-06, + "loss": 0.89758205, + "num_input_tokens_seen": 54593120, + "step": 2538, + "time_per_iteration": 2.585540533065796 + }, + { + "auxiliary_loss_clip": 0.01202503, + "auxiliary_loss_mlp": 0.01028592, + "balance_loss_clip": 1.06024361, + "balance_loss_mlp": 1.02013373, + "epoch": 0.30529669933265197, + "flos": 21945118970880.0, + "grad_norm": 1.5370070370213105, + "language_loss": 0.81865859, + "learning_rate": 3.2564399945912848e-06, + "loss": 0.84096956, + "num_input_tokens_seen": 54612910, + "step": 2539, + "time_per_iteration": 2.5983388423919678 + }, + { + "auxiliary_loss_clip": 0.01234978, + "auxiliary_loss_mlp": 0.01032567, + "balance_loss_clip": 1.05089295, + "balance_loss_mlp": 1.02443683, + "epoch": 0.305416942223291, + "flos": 21835411856640.0, + "grad_norm": 2.1890457597766884, + "language_loss": 0.82190096, + "learning_rate": 3.2558338315129287e-06, + "loss": 0.84457642, + "num_input_tokens_seen": 54631055, + "step": 2540, + "time_per_iteration": 2.716942310333252 + }, + { + "auxiliary_loss_clip": 0.0120243, + "auxiliary_loss_mlp": 0.01038753, + "balance_loss_clip": 1.05652118, + "balance_loss_mlp": 1.02938342, + "epoch": 0.30553718511393013, + "flos": 33911810709120.0, + "grad_norm": 2.453617797313978, + "language_loss": 0.7598837, + "learning_rate": 3.2552274779255785e-06, + "loss": 0.78229553, + "num_input_tokens_seen": 54651985, + "step": 2541, + "time_per_iteration": 3.5765438079833984 + }, + { + "auxiliary_loss_clip": 0.01205599, + "auxiliary_loss_mlp": 0.01027095, + "balance_loss_clip": 1.05658817, + "balance_loss_mlp": 1.0180707, + "epoch": 0.30565742800456924, + "flos": 22268530051200.0, + "grad_norm": 2.117220401373692, + "language_loss": 0.76892948, + "learning_rate": 3.2546209339212184e-06, + "loss": 0.79125643, + "num_input_tokens_seen": 54671005, + "step": 2542, + "time_per_iteration": 2.6427860260009766 + }, + { + "auxiliary_loss_clip": 0.0121631, + "auxiliary_loss_mlp": 0.01026558, + "balance_loss_clip": 1.05376267, + "balance_loss_mlp": 1.01857138, + "epoch": 0.3057776708952083, + "flos": 22565044823040.0, + "grad_norm": 1.639700399558147, + "language_loss": 0.77596557, + "learning_rate": 3.25401419959186e-06, + "loss": 0.7983942, + "num_input_tokens_seen": 54691615, + "step": 2543, + "time_per_iteration": 2.7306833267211914 + }, + { + "auxiliary_loss_clip": 0.01229257, + "auxiliary_loss_mlp": 0.01032516, + "balance_loss_clip": 1.06058371, + "balance_loss_mlp": 1.02463579, + "epoch": 0.3058979137858474, + "flos": 21799213925760.0, + "grad_norm": 1.7480123651220734, + "language_loss": 0.75645626, + "learning_rate": 3.253407275029545e-06, + "loss": 0.77907401, + "num_input_tokens_seen": 54710520, + "step": 2544, + "time_per_iteration": 2.7539401054382324 + }, + { + "auxiliary_loss_clip": 0.01233276, + "auxiliary_loss_mlp": 0.01036251, + "balance_loss_clip": 1.05877805, + "balance_loss_mlp": 1.02704191, + "epoch": 0.3060181566764865, + "flos": 26979435601920.0, + "grad_norm": 1.7428599280547996, + "language_loss": 0.80700731, + "learning_rate": 3.2528001603263425e-06, + "loss": 0.82970262, + "num_input_tokens_seen": 54732590, + "step": 2545, + "time_per_iteration": 2.7275609970092773 + }, + { + "auxiliary_loss_clip": 0.01208663, + "auxiliary_loss_mlp": 0.0103274, + "balance_loss_clip": 1.05847621, + "balance_loss_mlp": 1.02381158, + "epoch": 0.3061383995671256, + "flos": 19865101173120.0, + "grad_norm": 1.8479946228730884, + "language_loss": 0.8129341, + "learning_rate": 3.2521928555743514e-06, + "loss": 0.83534813, + "num_input_tokens_seen": 54749935, + "step": 2546, + "time_per_iteration": 2.6298320293426514 + }, + { + "auxiliary_loss_clip": 0.01211588, + "auxiliary_loss_mlp": 0.01379671, + "balance_loss_clip": 1.05485392, + "balance_loss_mlp": 1.00000405, + "epoch": 0.3062586424577647, + "flos": 22127509255680.0, + "grad_norm": 2.3451404491578183, + "language_loss": 0.67380714, + "learning_rate": 3.2515853608657e-06, + "loss": 0.69971967, + "num_input_tokens_seen": 54767935, + "step": 2547, + "time_per_iteration": 2.6485326290130615 + }, + { + "auxiliary_loss_clip": 0.01204342, + "auxiliary_loss_mlp": 0.01030592, + "balance_loss_clip": 1.05723512, + "balance_loss_mlp": 1.02144289, + "epoch": 0.3063788853484038, + "flos": 20845497962880.0, + "grad_norm": 2.5711473301223, + "language_loss": 0.74381554, + "learning_rate": 3.250977676292545e-06, + "loss": 0.7661649, + "num_input_tokens_seen": 54786175, + "step": 2548, + "time_per_iteration": 2.767514944076538 + }, + { + "auxiliary_loss_clip": 0.01219022, + "auxiliary_loss_mlp": 0.01031209, + "balance_loss_clip": 1.05655909, + "balance_loss_mlp": 1.02244675, + "epoch": 0.30649912823904285, + "flos": 16209717707520.0, + "grad_norm": 2.0638103085842467, + "language_loss": 0.79524708, + "learning_rate": 3.2503698019470712e-06, + "loss": 0.81774938, + "num_input_tokens_seen": 54801945, + "step": 2549, + "time_per_iteration": 2.6119346618652344 + }, + { + "auxiliary_loss_clip": 0.01206349, + "auxiliary_loss_mlp": 0.01028246, + "balance_loss_clip": 1.05441654, + "balance_loss_mlp": 1.02002621, + "epoch": 0.30661937112968196, + "flos": 18617815353600.0, + "grad_norm": 2.419348361355445, + "language_loss": 0.7815367, + "learning_rate": 3.249761737921492e-06, + "loss": 0.80388266, + "num_input_tokens_seen": 54818475, + "step": 2550, + "time_per_iteration": 2.6873910427093506 + }, + { + "auxiliary_loss_clip": 0.01215178, + "auxiliary_loss_mlp": 0.01028783, + "balance_loss_clip": 1.05667591, + "balance_loss_mlp": 1.02070081, + "epoch": 0.30673961402032107, + "flos": 31390809638400.0, + "grad_norm": 2.3357635703360082, + "language_loss": 0.7459743, + "learning_rate": 3.249153484308051e-06, + "loss": 0.7684139, + "num_input_tokens_seen": 54837090, + "step": 2551, + "time_per_iteration": 2.7863638401031494 + }, + { + "auxiliary_loss_clip": 0.0122131, + "auxiliary_loss_mlp": 0.01029221, + "balance_loss_clip": 1.05019915, + "balance_loss_mlp": 1.02026868, + "epoch": 0.3068598569109601, + "flos": 20229809915520.0, + "grad_norm": 1.8352785731077292, + "language_loss": 0.77874494, + "learning_rate": 3.2485450411990194e-06, + "loss": 0.80125022, + "num_input_tokens_seen": 54856445, + "step": 2552, + "time_per_iteration": 2.7447235584259033 + }, + { + "auxiliary_loss_clip": 0.01198686, + "auxiliary_loss_mlp": 0.01030021, + "balance_loss_clip": 1.05573165, + "balance_loss_mlp": 1.02090704, + "epoch": 0.30698009980159924, + "flos": 29601991399680.0, + "grad_norm": 1.7000982156780038, + "language_loss": 0.8215158, + "learning_rate": 3.2479364086866983e-06, + "loss": 0.84380293, + "num_input_tokens_seen": 54876700, + "step": 2553, + "time_per_iteration": 2.6598031520843506 + }, + { + "auxiliary_loss_clip": 0.01219117, + "auxiliary_loss_mlp": 0.0137969, + "balance_loss_clip": 1.05928838, + "balance_loss_mlp": 1.00009918, + "epoch": 0.30710034269223835, + "flos": 23842423261440.0, + "grad_norm": 1.774207324605309, + "language_loss": 0.81503785, + "learning_rate": 3.247327586863416e-06, + "loss": 0.84102589, + "num_input_tokens_seen": 54897580, + "step": 2554, + "time_per_iteration": 2.6963882446289062 + }, + { + "auxiliary_loss_clip": 0.01227229, + "auxiliary_loss_mlp": 0.01030196, + "balance_loss_clip": 1.05214024, + "balance_loss_mlp": 1.0216012, + "epoch": 0.3072205855828774, + "flos": 25884986152320.0, + "grad_norm": 2.0313482610685245, + "language_loss": 0.77025837, + "learning_rate": 3.2467185758215304e-06, + "loss": 0.79283261, + "num_input_tokens_seen": 54917320, + "step": 2555, + "time_per_iteration": 2.7917027473449707 + }, + { + "auxiliary_loss_clip": 0.01228731, + "auxiliary_loss_mlp": 0.01379499, + "balance_loss_clip": 1.05484009, + "balance_loss_mlp": 1.0001018, + "epoch": 0.3073408284735165, + "flos": 22236390357120.0, + "grad_norm": 2.387132093202862, + "language_loss": 0.85657346, + "learning_rate": 3.246109375653428e-06, + "loss": 0.88265574, + "num_input_tokens_seen": 54934085, + "step": 2556, + "time_per_iteration": 2.741791248321533 + }, + { + "auxiliary_loss_clip": 0.01197668, + "auxiliary_loss_mlp": 0.01029511, + "balance_loss_clip": 1.05783105, + "balance_loss_mlp": 1.02064192, + "epoch": 0.30746107136415557, + "flos": 19500284689920.0, + "grad_norm": 2.020207938899999, + "language_loss": 0.78547812, + "learning_rate": 3.2454999864515243e-06, + "loss": 0.80774993, + "num_input_tokens_seen": 54953460, + "step": 2557, + "time_per_iteration": 2.5967578887939453 + }, + { + "auxiliary_loss_clip": 0.01208916, + "auxiliary_loss_mlp": 0.01379357, + "balance_loss_clip": 1.05411935, + "balance_loss_mlp": 1.00006938, + "epoch": 0.3075813142547947, + "flos": 21724806902400.0, + "grad_norm": 1.705107457466735, + "language_loss": 0.69439495, + "learning_rate": 3.244890408308263e-06, + "loss": 0.72027767, + "num_input_tokens_seen": 54974165, + "step": 2558, + "time_per_iteration": 2.6910555362701416 + }, + { + "auxiliary_loss_clip": 0.012316, + "auxiliary_loss_mlp": 0.01025306, + "balance_loss_clip": 1.04962111, + "balance_loss_mlp": 1.01669335, + "epoch": 0.3077015571454338, + "flos": 24097963593600.0, + "grad_norm": 2.4679955434397516, + "language_loss": 0.61046273, + "learning_rate": 3.2442806413161165e-06, + "loss": 0.63303185, + "num_input_tokens_seen": 54993810, + "step": 2559, + "time_per_iteration": 2.7885842323303223 + }, + { + "auxiliary_loss_clip": 0.01232785, + "auxiliary_loss_mlp": 0.01037615, + "balance_loss_clip": 1.04931974, + "balance_loss_mlp": 1.02856135, + "epoch": 0.30782180003607285, + "flos": 18405476104320.0, + "grad_norm": 1.9528041138980452, + "language_loss": 0.75804925, + "learning_rate": 3.243670685567586e-06, + "loss": 0.78075325, + "num_input_tokens_seen": 55011210, + "step": 2560, + "time_per_iteration": 3.6002118587493896 + }, + { + "auxiliary_loss_clip": 0.0121563, + "auxiliary_loss_mlp": 0.0137943, + "balance_loss_clip": 1.05448198, + "balance_loss_mlp": 1.00010324, + "epoch": 0.30794204292671196, + "flos": 23878549365120.0, + "grad_norm": 2.296988189627587, + "language_loss": 0.80333513, + "learning_rate": 3.2430605411552012e-06, + "loss": 0.82928574, + "num_input_tokens_seen": 55031325, + "step": 2561, + "time_per_iteration": 2.709308385848999 + }, + { + "auxiliary_loss_clip": 0.01139569, + "auxiliary_loss_mlp": 0.0100746, + "balance_loss_clip": 1.02761769, + "balance_loss_mlp": 1.00539768, + "epoch": 0.30806228581735107, + "flos": 67927800816000.0, + "grad_norm": 0.9002051547286404, + "language_loss": 0.70617008, + "learning_rate": 3.2424502081715205e-06, + "loss": 0.72764039, + "num_input_tokens_seen": 55094440, + "step": 2562, + "time_per_iteration": 5.1435136795043945 + }, + { + "auxiliary_loss_clip": 0.01220371, + "auxiliary_loss_mlp": 0.01029704, + "balance_loss_clip": 1.05639672, + "balance_loss_mlp": 1.02065563, + "epoch": 0.3081825287079901, + "flos": 23843213360640.0, + "grad_norm": 1.652228021888808, + "language_loss": 0.77957332, + "learning_rate": 3.241839686709132e-06, + "loss": 0.80207407, + "num_input_tokens_seen": 55115375, + "step": 2563, + "time_per_iteration": 2.753427505493164 + }, + { + "auxiliary_loss_clip": 0.01204462, + "auxiliary_loss_mlp": 0.01036455, + "balance_loss_clip": 1.05446696, + "balance_loss_mlp": 1.02728176, + "epoch": 0.30830277159862923, + "flos": 16209969102720.0, + "grad_norm": 2.482540490182887, + "language_loss": 0.81544816, + "learning_rate": 3.2412289768606495e-06, + "loss": 0.83785737, + "num_input_tokens_seen": 55131945, + "step": 2564, + "time_per_iteration": 2.6044108867645264 + }, + { + "auxiliary_loss_clip": 0.01210973, + "auxiliary_loss_mlp": 0.01039095, + "balance_loss_clip": 1.05851626, + "balance_loss_mlp": 1.03004146, + "epoch": 0.30842301448926834, + "flos": 29349503723520.0, + "grad_norm": 1.9665680159776109, + "language_loss": 0.83050156, + "learning_rate": 3.240618078718718e-06, + "loss": 0.85300225, + "num_input_tokens_seen": 55153405, + "step": 2565, + "time_per_iteration": 2.745103597640991 + }, + { + "auxiliary_loss_clip": 0.01217609, + "auxiliary_loss_mlp": 0.01032275, + "balance_loss_clip": 1.0500524, + "balance_loss_mlp": 1.02287531, + "epoch": 0.3085432573799074, + "flos": 21945190798080.0, + "grad_norm": 1.9555778087908176, + "language_loss": 0.73977911, + "learning_rate": 3.240006992376011e-06, + "loss": 0.76227796, + "num_input_tokens_seen": 55173030, + "step": 2566, + "time_per_iteration": 2.7309553623199463 + }, + { + "auxiliary_loss_clip": 0.01219089, + "auxiliary_loss_mlp": 0.01030153, + "balance_loss_clip": 1.05515313, + "balance_loss_mlp": 1.02197504, + "epoch": 0.3086635002705465, + "flos": 22054718344320.0, + "grad_norm": 2.263247877749846, + "language_loss": 0.76260316, + "learning_rate": 3.2393957179252284e-06, + "loss": 0.78509557, + "num_input_tokens_seen": 55189565, + "step": 2567, + "time_per_iteration": 3.5804443359375 + }, + { + "auxiliary_loss_clip": 0.01205717, + "auxiliary_loss_mlp": 0.01031631, + "balance_loss_clip": 1.06207943, + "balance_loss_mlp": 1.02265418, + "epoch": 0.3087837431611856, + "flos": 32665925520000.0, + "grad_norm": 2.0271978522011938, + "language_loss": 0.8074615, + "learning_rate": 3.2387842554591016e-06, + "loss": 0.82983494, + "num_input_tokens_seen": 55210380, + "step": 2568, + "time_per_iteration": 2.6937592029571533 + }, + { + "auxiliary_loss_clip": 0.01202838, + "auxiliary_loss_mlp": 0.01024407, + "balance_loss_clip": 1.0599227, + "balance_loss_mlp": 1.01555538, + "epoch": 0.3089039860518247, + "flos": 17599245384960.0, + "grad_norm": 2.053802747620043, + "language_loss": 0.87533939, + "learning_rate": 3.238172605070388e-06, + "loss": 0.89761186, + "num_input_tokens_seen": 55225795, + "step": 2569, + "time_per_iteration": 2.574364185333252 + }, + { + "auxiliary_loss_clip": 0.01206013, + "auxiliary_loss_mlp": 0.01379615, + "balance_loss_clip": 1.05648661, + "balance_loss_mlp": 1.0001049, + "epoch": 0.3090242289424638, + "flos": 14383839611520.0, + "grad_norm": 2.861075097813381, + "language_loss": 0.78530198, + "learning_rate": 3.2375607668518745e-06, + "loss": 0.8111583, + "num_input_tokens_seen": 55238830, + "step": 2570, + "time_per_iteration": 2.6369731426239014 + }, + { + "auxiliary_loss_clip": 0.0120443, + "auxiliary_loss_mlp": 0.0103202, + "balance_loss_clip": 1.05218041, + "balance_loss_mlp": 1.02256083, + "epoch": 0.30914447183310284, + "flos": 16068625084800.0, + "grad_norm": 2.1541490681298923, + "language_loss": 0.89524412, + "learning_rate": 3.236948740896377e-06, + "loss": 0.91760862, + "num_input_tokens_seen": 55253630, + "step": 2571, + "time_per_iteration": 2.623579502105713 + }, + { + "auxiliary_loss_clip": 0.01208279, + "auxiliary_loss_mlp": 0.01034972, + "balance_loss_clip": 1.05728495, + "balance_loss_mlp": 1.02548337, + "epoch": 0.30926471472374195, + "flos": 32230221546240.0, + "grad_norm": 1.4373360358155516, + "language_loss": 0.84369123, + "learning_rate": 3.2363365272967384e-06, + "loss": 0.86612374, + "num_input_tokens_seen": 55276200, + "step": 2572, + "time_per_iteration": 2.7373008728027344 + }, + { + "auxiliary_loss_clip": 0.01208933, + "auxiliary_loss_mlp": 0.01029901, + "balance_loss_clip": 1.05871272, + "balance_loss_mlp": 1.02036428, + "epoch": 0.30938495761438106, + "flos": 20370722970240.0, + "grad_norm": 1.91829770374152, + "language_loss": 0.81674773, + "learning_rate": 3.235724126145832e-06, + "loss": 0.83913612, + "num_input_tokens_seen": 55292235, + "step": 2573, + "time_per_iteration": 2.5993049144744873 + }, + { + "auxiliary_loss_clip": 0.01196007, + "auxiliary_loss_mlp": 0.01034392, + "balance_loss_clip": 1.05390131, + "balance_loss_mlp": 1.02548146, + "epoch": 0.3095052005050201, + "flos": 24061155131520.0, + "grad_norm": 1.8257483054285564, + "language_loss": 0.77652395, + "learning_rate": 3.235111537536558e-06, + "loss": 0.79882801, + "num_input_tokens_seen": 55313050, + "step": 2574, + "time_per_iteration": 2.676602602005005 + }, + { + "auxiliary_loss_clip": 0.01210507, + "auxiliary_loss_mlp": 0.01027983, + "balance_loss_clip": 1.05747235, + "balance_loss_mlp": 1.01917958, + "epoch": 0.30962544339565923, + "flos": 23401547729280.0, + "grad_norm": 2.66424774302518, + "language_loss": 0.82470047, + "learning_rate": 3.2344987615618456e-06, + "loss": 0.84708542, + "num_input_tokens_seen": 55332885, + "step": 2575, + "time_per_iteration": 2.637422800064087 + }, + { + "auxiliary_loss_clip": 0.01225985, + "auxiliary_loss_mlp": 0.01034665, + "balance_loss_clip": 1.05722094, + "balance_loss_mlp": 1.0253613, + "epoch": 0.30974568628629834, + "flos": 33799984692480.0, + "grad_norm": 1.7515036487579938, + "language_loss": 0.78550982, + "learning_rate": 3.2338857983146533e-06, + "loss": 0.80811632, + "num_input_tokens_seen": 55354385, + "step": 2576, + "time_per_iteration": 2.855494499206543 + }, + { + "auxiliary_loss_clip": 0.01203814, + "auxiliary_loss_mlp": 0.01031546, + "balance_loss_clip": 1.05410409, + "balance_loss_mlp": 1.0220871, + "epoch": 0.3098659291769374, + "flos": 20229594433920.0, + "grad_norm": 2.5824011379560887, + "language_loss": 0.7656225, + "learning_rate": 3.233272647887966e-06, + "loss": 0.78797615, + "num_input_tokens_seen": 55373275, + "step": 2577, + "time_per_iteration": 2.6915788650512695 + }, + { + "auxiliary_loss_clip": 0.01205012, + "auxiliary_loss_mlp": 0.01034684, + "balance_loss_clip": 1.06166315, + "balance_loss_mlp": 1.02584505, + "epoch": 0.3099861720675765, + "flos": 24748556682240.0, + "grad_norm": 1.6812947887263214, + "language_loss": 0.90110165, + "learning_rate": 3.2326593103747985e-06, + "loss": 0.92349863, + "num_input_tokens_seen": 55392290, + "step": 2578, + "time_per_iteration": 2.657541036605835 + }, + { + "auxiliary_loss_clip": 0.01209354, + "auxiliary_loss_mlp": 0.01034312, + "balance_loss_clip": 1.05918384, + "balance_loss_mlp": 1.02503133, + "epoch": 0.3101064149582156, + "flos": 11765485704960.0, + "grad_norm": 2.4924872487155163, + "language_loss": 0.85130358, + "learning_rate": 3.2320457858681936e-06, + "loss": 0.8737402, + "num_input_tokens_seen": 55410680, + "step": 2579, + "time_per_iteration": 2.7020630836486816 + }, + { + "auxiliary_loss_clip": 0.01216068, + "auxiliary_loss_mlp": 0.0103068, + "balance_loss_clip": 1.05511165, + "balance_loss_mlp": 1.02209115, + "epoch": 0.31022665784885467, + "flos": 23033247626880.0, + "grad_norm": 2.552775222702522, + "language_loss": 0.85322922, + "learning_rate": 3.2314320744612228e-06, + "loss": 0.87569672, + "num_input_tokens_seen": 55425980, + "step": 2580, + "time_per_iteration": 2.7214009761810303 + }, + { + "auxiliary_loss_clip": 0.0120846, + "auxiliary_loss_mlp": 0.0103258, + "balance_loss_clip": 1.05986547, + "balance_loss_mlp": 1.02409816, + "epoch": 0.3103469007394938, + "flos": 16289188548480.0, + "grad_norm": 1.8566202862462935, + "language_loss": 0.76398945, + "learning_rate": 3.2308181762469854e-06, + "loss": 0.78639984, + "num_input_tokens_seen": 55443925, + "step": 2581, + "time_per_iteration": 2.5908315181732178 + }, + { + "auxiliary_loss_clip": 0.01206076, + "auxiliary_loss_mlp": 0.01031484, + "balance_loss_clip": 1.05981112, + "balance_loss_mlp": 1.02141726, + "epoch": 0.3104671436301329, + "flos": 30515271626880.0, + "grad_norm": 1.9328236462776263, + "language_loss": 0.77730137, + "learning_rate": 3.230204091318609e-06, + "loss": 0.79967701, + "num_input_tokens_seen": 55464465, + "step": 2582, + "time_per_iteration": 2.7420971393585205 + }, + { + "auxiliary_loss_clip": 0.01199771, + "auxiliary_loss_mlp": 0.01379674, + "balance_loss_clip": 1.05677772, + "balance_loss_mlp": 1.00010407, + "epoch": 0.31058738652077195, + "flos": 20047240062720.0, + "grad_norm": 1.8993287956871174, + "language_loss": 0.847157, + "learning_rate": 3.2295898197692503e-06, + "loss": 0.87295139, + "num_input_tokens_seen": 55483425, + "step": 2583, + "time_per_iteration": 2.5790090560913086 + }, + { + "auxiliary_loss_clip": 0.0120074, + "auxiliary_loss_mlp": 0.01033436, + "balance_loss_clip": 1.05864334, + "balance_loss_mlp": 1.02491236, + "epoch": 0.31070762941141106, + "flos": 28074639237120.0, + "grad_norm": 1.6444421964742673, + "language_loss": 0.79364681, + "learning_rate": 3.228975361692094e-06, + "loss": 0.81598866, + "num_input_tokens_seen": 55504445, + "step": 2584, + "time_per_iteration": 2.5941712856292725 + }, + { + "auxiliary_loss_clip": 0.01214086, + "auxiliary_loss_mlp": 0.01380211, + "balance_loss_clip": 1.05562603, + "balance_loss_mlp": 1.00022936, + "epoch": 0.31082787230205017, + "flos": 20521907314560.0, + "grad_norm": 2.140608724370487, + "language_loss": 0.80385947, + "learning_rate": 3.228360717180352e-06, + "loss": 0.82980239, + "num_input_tokens_seen": 55521970, + "step": 2585, + "time_per_iteration": 2.610684871673584 + }, + { + "auxiliary_loss_clip": 0.01102987, + "auxiliary_loss_mlp": 0.0137557, + "balance_loss_clip": 1.02991486, + "balance_loss_mlp": 1.00015533, + "epoch": 0.3109481151926892, + "flos": 62445928723200.0, + "grad_norm": 0.8215025336224717, + "language_loss": 0.59404659, + "learning_rate": 3.227745886327266e-06, + "loss": 0.61883217, + "num_input_tokens_seen": 55580665, + "step": 2586, + "time_per_iteration": 4.041039228439331 + }, + { + "auxiliary_loss_clip": 0.01101723, + "auxiliary_loss_mlp": 0.0100246, + "balance_loss_clip": 1.02922094, + "balance_loss_mlp": 1.00030196, + "epoch": 0.31106835808332833, + "flos": 44746744723200.0, + "grad_norm": 0.8047779732343593, + "language_loss": 0.55866957, + "learning_rate": 3.227130869226105e-06, + "loss": 0.57971132, + "num_input_tokens_seen": 55637825, + "step": 2587, + "time_per_iteration": 3.146965503692627 + }, + { + "auxiliary_loss_clip": 0.01213038, + "auxiliary_loss_mlp": 0.01027464, + "balance_loss_clip": 1.05789256, + "balance_loss_mlp": 1.01829135, + "epoch": 0.3111886009739674, + "flos": 23403056100480.0, + "grad_norm": 2.6482152560664227, + "language_loss": 0.82379019, + "learning_rate": 3.226515665970167e-06, + "loss": 0.84619522, + "num_input_tokens_seen": 55655365, + "step": 2588, + "time_per_iteration": 4.4080705642700195 + }, + { + "auxiliary_loss_clip": 0.01207639, + "auxiliary_loss_mlp": 0.01033264, + "balance_loss_clip": 1.05631328, + "balance_loss_mlp": 1.02312565, + "epoch": 0.3113088438646065, + "flos": 17530728192000.0, + "grad_norm": 2.5800358008541355, + "language_loss": 0.86075932, + "learning_rate": 3.225900276652777e-06, + "loss": 0.88316834, + "num_input_tokens_seen": 55672140, + "step": 2589, + "time_per_iteration": 2.57928729057312 + }, + { + "auxiliary_loss_clip": 0.01224809, + "auxiliary_loss_mlp": 0.01029978, + "balance_loss_clip": 1.05650258, + "balance_loss_mlp": 1.02143121, + "epoch": 0.3114290867552456, + "flos": 28365802882560.0, + "grad_norm": 1.5486629908832579, + "language_loss": 0.75634706, + "learning_rate": 3.2252847013672906e-06, + "loss": 0.7788949, + "num_input_tokens_seen": 55694800, + "step": 2590, + "time_per_iteration": 2.7492055892944336 + }, + { + "auxiliary_loss_clip": 0.0121777, + "auxiliary_loss_mlp": 0.0103074, + "balance_loss_clip": 1.05212069, + "balance_loss_mlp": 1.0216862, + "epoch": 0.31154932964588467, + "flos": 27379157126400.0, + "grad_norm": 2.2620812648716937, + "language_loss": 0.76312494, + "learning_rate": 3.224668940207089e-06, + "loss": 0.78561008, + "num_input_tokens_seen": 55713785, + "step": 2591, + "time_per_iteration": 2.7629168033599854 + }, + { + "auxiliary_loss_clip": 0.01223197, + "auxiliary_loss_mlp": 0.01026789, + "balance_loss_clip": 1.04964757, + "balance_loss_mlp": 1.01709127, + "epoch": 0.3116695725365238, + "flos": 26541864120960.0, + "grad_norm": 2.278460204213002, + "language_loss": 0.86480951, + "learning_rate": 3.2240529932655828e-06, + "loss": 0.88730937, + "num_input_tokens_seen": 55733050, + "step": 2592, + "time_per_iteration": 2.7364296913146973 + }, + { + "auxiliary_loss_clip": 0.01216241, + "auxiliary_loss_mlp": 0.01029801, + "balance_loss_clip": 1.05686212, + "balance_loss_mlp": 1.02021706, + "epoch": 0.3117898154271629, + "flos": 21177600134400.0, + "grad_norm": 2.49090018306171, + "language_loss": 0.88487554, + "learning_rate": 3.223436860636211e-06, + "loss": 0.90733594, + "num_input_tokens_seen": 55748685, + "step": 2593, + "time_per_iteration": 3.561450958251953 + }, + { + "auxiliary_loss_clip": 0.01201522, + "auxiliary_loss_mlp": 0.01033272, + "balance_loss_clip": 1.05824554, + "balance_loss_mlp": 1.02409303, + "epoch": 0.31191005831780194, + "flos": 27272430840960.0, + "grad_norm": 1.6072763706555002, + "language_loss": 0.73938203, + "learning_rate": 3.2228205424124403e-06, + "loss": 0.76172996, + "num_input_tokens_seen": 55771840, + "step": 2594, + "time_per_iteration": 2.6471364498138428 + }, + { + "auxiliary_loss_clip": 0.01202447, + "auxiliary_loss_mlp": 0.01027021, + "balance_loss_clip": 1.05270672, + "balance_loss_mlp": 1.01858056, + "epoch": 0.31203030120844105, + "flos": 12963501043200.0, + "grad_norm": 2.301937420712949, + "language_loss": 0.74716604, + "learning_rate": 3.222204038687765e-06, + "loss": 0.76946068, + "num_input_tokens_seen": 55784975, + "step": 2595, + "time_per_iteration": 2.6434342861175537 + }, + { + "auxiliary_loss_clip": 0.01205625, + "auxiliary_loss_mlp": 0.01034017, + "balance_loss_clip": 1.05671966, + "balance_loss_mlp": 1.02506495, + "epoch": 0.31215054409908016, + "flos": 27562014288000.0, + "grad_norm": 1.6044452500082458, + "language_loss": 0.87716854, + "learning_rate": 3.221587349555709e-06, + "loss": 0.89956498, + "num_input_tokens_seen": 55805235, + "step": 2596, + "time_per_iteration": 2.6637680530548096 + }, + { + "auxiliary_loss_clip": 0.01220165, + "auxiliary_loss_mlp": 0.01033322, + "balance_loss_clip": 1.05504918, + "balance_loss_mlp": 1.02442908, + "epoch": 0.3122707869897192, + "flos": 21506326427520.0, + "grad_norm": 1.7217010971991606, + "language_loss": 0.69456041, + "learning_rate": 3.2209704751098236e-06, + "loss": 0.71709526, + "num_input_tokens_seen": 55824265, + "step": 2597, + "time_per_iteration": 2.686795234680176 + }, + { + "auxiliary_loss_clip": 0.01223487, + "auxiliary_loss_mlp": 0.01034089, + "balance_loss_clip": 1.05770206, + "balance_loss_mlp": 1.02460027, + "epoch": 0.31239102988035833, + "flos": 15187017674880.0, + "grad_norm": 2.0659115140099384, + "language_loss": 0.82854718, + "learning_rate": 3.2203534154436875e-06, + "loss": 0.85112298, + "num_input_tokens_seen": 55838620, + "step": 2598, + "time_per_iteration": 2.626469612121582 + }, + { + "auxiliary_loss_clip": 0.01238747, + "auxiliary_loss_mlp": 0.01029678, + "balance_loss_clip": 1.04805589, + "balance_loss_mlp": 1.02032042, + "epoch": 0.31251127277099744, + "flos": 22053712763520.0, + "grad_norm": 1.9493858153123607, + "language_loss": 0.75812232, + "learning_rate": 3.2197361706509084e-06, + "loss": 0.7808066, + "num_input_tokens_seen": 55859375, + "step": 2599, + "time_per_iteration": 2.843463897705078 + }, + { + "auxiliary_loss_clip": 0.01202579, + "auxiliary_loss_mlp": 0.01032889, + "balance_loss_clip": 1.0578872, + "balance_loss_mlp": 1.02310801, + "epoch": 0.3126315156616365, + "flos": 15193984913280.0, + "grad_norm": 2.695537010363691, + "language_loss": 0.83232969, + "learning_rate": 3.2191187408251228e-06, + "loss": 0.85468435, + "num_input_tokens_seen": 55876535, + "step": 2600, + "time_per_iteration": 2.552968740463257 + }, + { + "auxiliary_loss_clip": 0.01217152, + "auxiliary_loss_mlp": 0.01036305, + "balance_loss_clip": 1.05712688, + "balance_loss_mlp": 1.02657175, + "epoch": 0.3127517585522756, + "flos": 18145338831360.0, + "grad_norm": 2.468941853856935, + "language_loss": 0.78717315, + "learning_rate": 3.218501126059993e-06, + "loss": 0.80970776, + "num_input_tokens_seen": 55891930, + "step": 2601, + "time_per_iteration": 2.5776922702789307 + }, + { + "auxiliary_loss_clip": 0.0120796, + "auxiliary_loss_mlp": 0.01033, + "balance_loss_clip": 1.05358326, + "balance_loss_mlp": 1.02377295, + "epoch": 0.31287200144291466, + "flos": 21908633731200.0, + "grad_norm": 1.8199759082850946, + "language_loss": 0.81289971, + "learning_rate": 3.2178833264492116e-06, + "loss": 0.83530933, + "num_input_tokens_seen": 55910635, + "step": 2602, + "time_per_iteration": 2.6831281185150146 + }, + { + "auxiliary_loss_clip": 0.01217256, + "auxiliary_loss_mlp": 0.0104204, + "balance_loss_clip": 1.05908477, + "balance_loss_mlp": 1.03234875, + "epoch": 0.31299224433355377, + "flos": 29896997800320.0, + "grad_norm": 3.786201839426547, + "language_loss": 0.76534688, + "learning_rate": 3.217265342086498e-06, + "loss": 0.78793979, + "num_input_tokens_seen": 55931125, + "step": 2603, + "time_per_iteration": 2.6939547061920166 + }, + { + "auxiliary_loss_clip": 0.01231246, + "auxiliary_loss_mlp": 0.01380411, + "balance_loss_clip": 1.05481541, + "balance_loss_mlp": 1.00029218, + "epoch": 0.3131124872241929, + "flos": 11655886331520.0, + "grad_norm": 2.0386061763346763, + "language_loss": 0.73199767, + "learning_rate": 3.216647173065599e-06, + "loss": 0.75811422, + "num_input_tokens_seen": 55946590, + "step": 2604, + "time_per_iteration": 2.7390291690826416 + }, + { + "auxiliary_loss_clip": 0.0121714, + "auxiliary_loss_mlp": 0.01030986, + "balance_loss_clip": 1.0574162, + "balance_loss_mlp": 1.02211666, + "epoch": 0.31323273011483194, + "flos": 49848785470080.0, + "grad_norm": 1.916379673844605, + "language_loss": 0.73757285, + "learning_rate": 3.216028819480292e-06, + "loss": 0.76005411, + "num_input_tokens_seen": 55967930, + "step": 2605, + "time_per_iteration": 2.9129061698913574 + }, + { + "auxiliary_loss_clip": 0.0120444, + "auxiliary_loss_mlp": 0.01026839, + "balance_loss_clip": 1.05436611, + "balance_loss_mlp": 1.01826751, + "epoch": 0.31335297300547105, + "flos": 22601278667520.0, + "grad_norm": 1.935249924767552, + "language_loss": 0.75919259, + "learning_rate": 3.2154102814243793e-06, + "loss": 0.78150535, + "num_input_tokens_seen": 55987070, + "step": 2606, + "time_per_iteration": 2.735849142074585 + }, + { + "auxiliary_loss_clip": 0.01230936, + "auxiliary_loss_mlp": 0.01035299, + "balance_loss_clip": 1.05449843, + "balance_loss_mlp": 1.02653718, + "epoch": 0.31347321589611016, + "flos": 34710858708480.0, + "grad_norm": 2.1855973836737057, + "language_loss": 0.67184979, + "learning_rate": 3.2147915589916937e-06, + "loss": 0.69451213, + "num_input_tokens_seen": 56008630, + "step": 2607, + "time_per_iteration": 2.8345518112182617 + }, + { + "auxiliary_loss_clip": 0.0120737, + "auxiliary_loss_mlp": 0.01034084, + "balance_loss_clip": 1.05253506, + "balance_loss_mlp": 1.02489293, + "epoch": 0.3135934587867492, + "flos": 19755789108480.0, + "grad_norm": 7.5469667419875845, + "language_loss": 0.82583565, + "learning_rate": 3.2141726522760938e-06, + "loss": 0.84825015, + "num_input_tokens_seen": 56026690, + "step": 2608, + "time_per_iteration": 2.6805386543273926 + }, + { + "auxiliary_loss_clip": 0.01109583, + "auxiliary_loss_mlp": 0.01002786, + "balance_loss_clip": 1.02581656, + "balance_loss_mlp": 1.00059271, + "epoch": 0.3137137016773883, + "flos": 65815535583360.0, + "grad_norm": 0.7001121695674196, + "language_loss": 0.5263527, + "learning_rate": 3.213553561371469e-06, + "loss": 0.54747641, + "num_input_tokens_seen": 56090425, + "step": 2609, + "time_per_iteration": 3.314582347869873 + }, + { + "auxiliary_loss_clip": 0.01238386, + "auxiliary_loss_mlp": 0.01035037, + "balance_loss_clip": 1.05675936, + "balance_loss_mlp": 1.02645421, + "epoch": 0.31383394456802743, + "flos": 16252739222400.0, + "grad_norm": 2.086069570261416, + "language_loss": 0.95932651, + "learning_rate": 3.212934286371733e-06, + "loss": 0.98206079, + "num_input_tokens_seen": 56107135, + "step": 2610, + "time_per_iteration": 2.7197866439819336 + }, + { + "auxiliary_loss_clip": 0.0120921, + "auxiliary_loss_mlp": 0.01032483, + "balance_loss_clip": 1.05771482, + "balance_loss_mlp": 1.0227735, + "epoch": 0.3139541874586665, + "flos": 38795517613440.0, + "grad_norm": 2.272787567879851, + "language_loss": 0.8295275, + "learning_rate": 3.2123148273708304e-06, + "loss": 0.85194445, + "num_input_tokens_seen": 56127325, + "step": 2611, + "time_per_iteration": 2.8066909313201904 + }, + { + "auxiliary_loss_clip": 0.01202706, + "auxiliary_loss_mlp": 0.01035126, + "balance_loss_clip": 1.06101894, + "balance_loss_mlp": 1.02578056, + "epoch": 0.3140744303493056, + "flos": 25046328430080.0, + "grad_norm": 2.053169019987239, + "language_loss": 0.76734024, + "learning_rate": 3.211695184462733e-06, + "loss": 0.78971857, + "num_input_tokens_seen": 56148500, + "step": 2612, + "time_per_iteration": 3.579976797103882 + }, + { + "auxiliary_loss_clip": 0.01131412, + "auxiliary_loss_mlp": 0.01002325, + "balance_loss_clip": 1.02256405, + "balance_loss_mlp": 1.00019169, + "epoch": 0.3141946732399447, + "flos": 72504254782080.0, + "grad_norm": 0.8842622011306873, + "language_loss": 0.60485369, + "learning_rate": 3.2110753577414383e-06, + "loss": 0.62619102, + "num_input_tokens_seen": 56210080, + "step": 2613, + "time_per_iteration": 4.170072317123413 + }, + { + "auxiliary_loss_clip": 0.01222833, + "auxiliary_loss_mlp": 0.01030079, + "balance_loss_clip": 1.05630422, + "balance_loss_mlp": 1.02060819, + "epoch": 0.31431491613058377, + "flos": 19239788280960.0, + "grad_norm": 1.7676710845779906, + "language_loss": 0.79077387, + "learning_rate": 3.2104553473009757e-06, + "loss": 0.81330299, + "num_input_tokens_seen": 56228200, + "step": 2614, + "time_per_iteration": 3.5293519496917725 + }, + { + "auxiliary_loss_clip": 0.01233008, + "auxiliary_loss_mlp": 0.01028054, + "balance_loss_clip": 1.05278182, + "balance_loss_mlp": 1.01840985, + "epoch": 0.3144351590212229, + "flos": 36210596290560.0, + "grad_norm": 1.82546710198772, + "language_loss": 0.67628306, + "learning_rate": 3.209835153235399e-06, + "loss": 0.69889367, + "num_input_tokens_seen": 56249755, + "step": 2615, + "time_per_iteration": 2.8105220794677734 + }, + { + "auxiliary_loss_clip": 0.01212969, + "auxiliary_loss_mlp": 0.01029463, + "balance_loss_clip": 1.05210638, + "balance_loss_mlp": 1.02000356, + "epoch": 0.314555401911862, + "flos": 18551740285440.0, + "grad_norm": 1.9270186700233978, + "language_loss": 0.6814943, + "learning_rate": 3.2092147756387916e-06, + "loss": 0.70391858, + "num_input_tokens_seen": 56270080, + "step": 2616, + "time_per_iteration": 2.7639989852905273 + }, + { + "auxiliary_loss_clip": 0.01209952, + "auxiliary_loss_mlp": 0.0103179, + "balance_loss_clip": 1.05311465, + "balance_loss_mlp": 1.02273035, + "epoch": 0.31467564480250104, + "flos": 16362877299840.0, + "grad_norm": 2.6367142992696406, + "language_loss": 0.83206928, + "learning_rate": 3.208594214605264e-06, + "loss": 0.8544867, + "num_input_tokens_seen": 56288625, + "step": 2617, + "time_per_iteration": 2.698126792907715 + }, + { + "auxiliary_loss_clip": 0.0120614, + "auxiliary_loss_mlp": 0.01027716, + "balance_loss_clip": 1.05384564, + "balance_loss_mlp": 1.01888847, + "epoch": 0.31479588769314015, + "flos": 21652375127040.0, + "grad_norm": 2.2686383162399846, + "language_loss": 0.77292013, + "learning_rate": 3.2079734702289553e-06, + "loss": 0.7952587, + "num_input_tokens_seen": 56307520, + "step": 2618, + "time_per_iteration": 2.6815884113311768 + }, + { + "auxiliary_loss_clip": 0.0110253, + "auxiliary_loss_mlp": 0.01375598, + "balance_loss_clip": 1.0222249, + "balance_loss_mlp": 1.00018013, + "epoch": 0.3149161305837792, + "flos": 66051072040320.0, + "grad_norm": 0.807845405496995, + "language_loss": 0.6046803, + "learning_rate": 3.207352542604031e-06, + "loss": 0.62946153, + "num_input_tokens_seen": 56369855, + "step": 2619, + "time_per_iteration": 4.162853002548218 + }, + { + "auxiliary_loss_clip": 0.01212569, + "auxiliary_loss_mlp": 0.0102918, + "balance_loss_clip": 1.05075276, + "balance_loss_mlp": 1.02060246, + "epoch": 0.3150363734744183, + "flos": 28987201192320.0, + "grad_norm": 1.5852335430660496, + "language_loss": 0.78495693, + "learning_rate": 3.2067314318246864e-06, + "loss": 0.80737436, + "num_input_tokens_seen": 56390570, + "step": 2620, + "time_per_iteration": 2.7606773376464844 + }, + { + "auxiliary_loss_clip": 0.0122635, + "auxiliary_loss_mlp": 0.01033229, + "balance_loss_clip": 1.05564785, + "balance_loss_mlp": 1.02391875, + "epoch": 0.31515661636505743, + "flos": 27636600879360.0, + "grad_norm": 1.8393626087479193, + "language_loss": 0.77394009, + "learning_rate": 3.206110137985143e-06, + "loss": 0.79653585, + "num_input_tokens_seen": 56410775, + "step": 2621, + "time_per_iteration": 2.7334351539611816 + }, + { + "auxiliary_loss_clip": 0.01211966, + "auxiliary_loss_mlp": 0.01035113, + "balance_loss_clip": 1.05099726, + "balance_loss_mlp": 1.0258919, + "epoch": 0.3152768592556965, + "flos": 24605632465920.0, + "grad_norm": 1.897511382311803, + "language_loss": 0.92431897, + "learning_rate": 3.2054886611796505e-06, + "loss": 0.94678974, + "num_input_tokens_seen": 56429770, + "step": 2622, + "time_per_iteration": 2.8340442180633545 + }, + { + "auxiliary_loss_clip": 0.01095209, + "auxiliary_loss_mlp": 0.01006973, + "balance_loss_clip": 1.02455056, + "balance_loss_mlp": 1.00482738, + "epoch": 0.3153971021463356, + "flos": 68476908026880.0, + "grad_norm": 0.8831720207858502, + "language_loss": 0.63589573, + "learning_rate": 3.204867001502487e-06, + "loss": 0.65691757, + "num_input_tokens_seen": 56488425, + "step": 2623, + "time_per_iteration": 3.1527199745178223 + }, + { + "auxiliary_loss_clip": 0.01201772, + "auxiliary_loss_mlp": 0.01031784, + "balance_loss_clip": 1.05962181, + "balance_loss_mlp": 1.02255082, + "epoch": 0.3155173450369747, + "flos": 25593714766080.0, + "grad_norm": 2.0197743073716303, + "language_loss": 0.80451709, + "learning_rate": 3.2042451590479567e-06, + "loss": 0.82685256, + "num_input_tokens_seen": 56508940, + "step": 2624, + "time_per_iteration": 2.7017157077789307 + }, + { + "auxiliary_loss_clip": 0.0120221, + "auxiliary_loss_mlp": 0.01030159, + "balance_loss_clip": 1.06123877, + "balance_loss_mlp": 1.0211345, + "epoch": 0.31563758792761376, + "flos": 24309333175680.0, + "grad_norm": 1.7465499328079095, + "language_loss": 0.86821669, + "learning_rate": 3.203623133910394e-06, + "loss": 0.89054042, + "num_input_tokens_seen": 56527245, + "step": 2625, + "time_per_iteration": 2.6447553634643555 + }, + { + "auxiliary_loss_clip": 0.01246417, + "auxiliary_loss_mlp": 0.0102702, + "balance_loss_clip": 1.05096674, + "balance_loss_mlp": 1.01766205, + "epoch": 0.31575783081825287, + "flos": 31903865550720.0, + "grad_norm": 2.598107793865589, + "language_loss": 0.77455449, + "learning_rate": 3.203000926184158e-06, + "loss": 0.79728889, + "num_input_tokens_seen": 56546170, + "step": 2626, + "time_per_iteration": 2.8976783752441406 + }, + { + "auxiliary_loss_clip": 0.01200282, + "auxiliary_loss_mlp": 0.01029174, + "balance_loss_clip": 1.05954504, + "balance_loss_mlp": 1.02095461, + "epoch": 0.315878073708892, + "flos": 30810960385920.0, + "grad_norm": 1.74054333200614, + "language_loss": 0.77521342, + "learning_rate": 3.202378535963639e-06, + "loss": 0.797508, + "num_input_tokens_seen": 56567085, + "step": 2627, + "time_per_iteration": 2.637019395828247 + }, + { + "auxiliary_loss_clip": 0.01205768, + "auxiliary_loss_mlp": 0.01380229, + "balance_loss_clip": 1.05286646, + "balance_loss_mlp": 1.00023603, + "epoch": 0.31599831659953104, + "flos": 22200264253440.0, + "grad_norm": 1.7881865087895568, + "language_loss": 0.83479047, + "learning_rate": 3.2017559633432516e-06, + "loss": 0.86065042, + "num_input_tokens_seen": 56586715, + "step": 2628, + "time_per_iteration": 2.7229204177856445 + }, + { + "auxiliary_loss_clip": 0.01226422, + "auxiliary_loss_mlp": 0.01033227, + "balance_loss_clip": 1.05715454, + "balance_loss_mlp": 1.02348733, + "epoch": 0.31611855949017015, + "flos": 25593463370880.0, + "grad_norm": 1.9428609151005563, + "language_loss": 0.66081244, + "learning_rate": 3.2011332084174398e-06, + "loss": 0.68340892, + "num_input_tokens_seen": 56607585, + "step": 2629, + "time_per_iteration": 2.6577529907226562 + }, + { + "auxiliary_loss_clip": 0.01210389, + "auxiliary_loss_mlp": 0.01034244, + "balance_loss_clip": 1.05985928, + "balance_loss_mlp": 1.02482605, + "epoch": 0.31623880238080926, + "flos": 20594087694720.0, + "grad_norm": 1.7734638485656633, + "language_loss": 0.8932054, + "learning_rate": 3.2005102712806756e-06, + "loss": 0.91565174, + "num_input_tokens_seen": 56626415, + "step": 2630, + "time_per_iteration": 2.6961119174957275 + }, + { + "auxiliary_loss_clip": 0.01217008, + "auxiliary_loss_mlp": 0.0103131, + "balance_loss_clip": 1.06029344, + "balance_loss_mlp": 1.02154124, + "epoch": 0.3163590452714483, + "flos": 12784917600000.0, + "grad_norm": 2.016804630195112, + "language_loss": 0.73072779, + "learning_rate": 3.1998871520274575e-06, + "loss": 0.75321102, + "num_input_tokens_seen": 56641750, + "step": 2631, + "time_per_iteration": 2.7197606563568115 + }, + { + "auxiliary_loss_clip": 0.01214297, + "auxiliary_loss_mlp": 0.01032585, + "balance_loss_clip": 1.05216944, + "balance_loss_mlp": 1.02329278, + "epoch": 0.3164792881620874, + "flos": 23041292273280.0, + "grad_norm": 1.658102025435322, + "language_loss": 0.85035694, + "learning_rate": 3.199263850752312e-06, + "loss": 0.87282574, + "num_input_tokens_seen": 56662585, + "step": 2632, + "time_per_iteration": 2.6777100563049316 + }, + { + "auxiliary_loss_clip": 0.01213204, + "auxiliary_loss_mlp": 0.01033578, + "balance_loss_clip": 1.05858409, + "balance_loss_mlp": 1.0248518, + "epoch": 0.31659953105272653, + "flos": 18296271780480.0, + "grad_norm": 2.443093821303859, + "language_loss": 0.85042566, + "learning_rate": 3.198640367549795e-06, + "loss": 0.87289345, + "num_input_tokens_seen": 56681480, + "step": 2633, + "time_per_iteration": 2.6756591796875 + }, + { + "auxiliary_loss_clip": 0.0121085, + "auxiliary_loss_mlp": 0.01379585, + "balance_loss_clip": 1.05645466, + "balance_loss_mlp": 1.00018287, + "epoch": 0.3167197739433656, + "flos": 25703421880320.0, + "grad_norm": 1.6815996061091838, + "language_loss": 0.85829943, + "learning_rate": 3.198016702514487e-06, + "loss": 0.88420379, + "num_input_tokens_seen": 56701760, + "step": 2634, + "time_per_iteration": 2.7108500003814697 + }, + { + "auxiliary_loss_clip": 0.01201321, + "auxiliary_loss_mlp": 0.0103299, + "balance_loss_clip": 1.05978978, + "balance_loss_mlp": 1.02429342, + "epoch": 0.3168400168340047, + "flos": 23546016230400.0, + "grad_norm": 1.656485086430437, + "language_loss": 0.84395397, + "learning_rate": 3.1973928557409972e-06, + "loss": 0.86629707, + "num_input_tokens_seen": 56719800, + "step": 2635, + "time_per_iteration": 2.66949725151062 + }, + { + "auxiliary_loss_clip": 0.0119911, + "auxiliary_loss_mlp": 0.0103347, + "balance_loss_clip": 1.05912852, + "balance_loss_mlp": 1.02508974, + "epoch": 0.31696025972464376, + "flos": 28366449327360.0, + "grad_norm": 2.10245841250099, + "language_loss": 0.71504223, + "learning_rate": 3.1967688273239636e-06, + "loss": 0.73736805, + "num_input_tokens_seen": 56739605, + "step": 2636, + "time_per_iteration": 2.645500659942627 + }, + { + "auxiliary_loss_clip": 0.01224961, + "auxiliary_loss_mlp": 0.01029104, + "balance_loss_clip": 1.05456388, + "balance_loss_mlp": 1.02015734, + "epoch": 0.31708050261528287, + "flos": 16399111144320.0, + "grad_norm": 1.7179115535211793, + "language_loss": 0.82274473, + "learning_rate": 3.1961446173580503e-06, + "loss": 0.84528542, + "num_input_tokens_seen": 56756545, + "step": 2637, + "time_per_iteration": 2.7697694301605225 + }, + { + "auxiliary_loss_clip": 0.01215514, + "auxiliary_loss_mlp": 0.01032143, + "balance_loss_clip": 1.05745506, + "balance_loss_mlp": 1.02295828, + "epoch": 0.317200745505922, + "flos": 26212347728640.0, + "grad_norm": 1.6425834300353908, + "language_loss": 0.77553231, + "learning_rate": 3.1955202259379502e-06, + "loss": 0.79800892, + "num_input_tokens_seen": 56778275, + "step": 2638, + "time_per_iteration": 2.7267417907714844 + }, + { + "auxiliary_loss_clip": 0.01204028, + "auxiliary_loss_mlp": 0.01032195, + "balance_loss_clip": 1.05502343, + "balance_loss_mlp": 1.02320623, + "epoch": 0.31732098839656103, + "flos": 31350876693120.0, + "grad_norm": 1.6932266829198575, + "language_loss": 0.83064055, + "learning_rate": 3.194895653158381e-06, + "loss": 0.85300279, + "num_input_tokens_seen": 56797215, + "step": 2639, + "time_per_iteration": 5.799081325531006 + }, + { + "auxiliary_loss_clip": 0.01090848, + "auxiliary_loss_mlp": 0.01003139, + "balance_loss_clip": 1.02125931, + "balance_loss_mlp": 1.00105274, + "epoch": 0.31744123128720014, + "flos": 58989024835200.0, + "grad_norm": 0.767031082318444, + "language_loss": 0.55536866, + "learning_rate": 3.194270899114093e-06, + "loss": 0.57630861, + "num_input_tokens_seen": 56863010, + "step": 2640, + "time_per_iteration": 3.2995851039886475 + }, + { + "auxiliary_loss_clip": 0.01213999, + "auxiliary_loss_mlp": 0.01034589, + "balance_loss_clip": 1.05779386, + "balance_loss_mlp": 1.02463555, + "epoch": 0.31756147417783925, + "flos": 17417573372160.0, + "grad_norm": 1.8093158075804179, + "language_loss": 0.82247925, + "learning_rate": 3.193645963899858e-06, + "loss": 0.8449651, + "num_input_tokens_seen": 56880625, + "step": 2641, + "time_per_iteration": 2.6724536418914795 + }, + { + "auxiliary_loss_clip": 0.01210104, + "auxiliary_loss_mlp": 0.01031754, + "balance_loss_clip": 1.05518794, + "balance_loss_mlp": 1.02247345, + "epoch": 0.3176817170684783, + "flos": 25481673267840.0, + "grad_norm": 1.823445003887561, + "language_loss": 0.84120291, + "learning_rate": 3.193020847610479e-06, + "loss": 0.86362147, + "num_input_tokens_seen": 56900945, + "step": 2642, + "time_per_iteration": 2.6846184730529785 + }, + { + "auxiliary_loss_clip": 0.01210727, + "auxiliary_loss_mlp": 0.01040263, + "balance_loss_clip": 1.05720377, + "balance_loss_mlp": 1.03011882, + "epoch": 0.3178019599591174, + "flos": 24972603765120.0, + "grad_norm": 2.730742309175087, + "language_loss": 0.71246403, + "learning_rate": 3.192395550340787e-06, + "loss": 0.73497391, + "num_input_tokens_seen": 56918895, + "step": 2643, + "time_per_iteration": 2.666456699371338 + }, + { + "auxiliary_loss_clip": 0.01206388, + "auxiliary_loss_mlp": 0.01033456, + "balance_loss_clip": 1.05717015, + "balance_loss_mlp": 1.02438974, + "epoch": 0.31792220284975653, + "flos": 12422220019200.0, + "grad_norm": 2.1175203284935566, + "language_loss": 0.77279115, + "learning_rate": 3.191770072185638e-06, + "loss": 0.79518962, + "num_input_tokens_seen": 56935890, + "step": 2644, + "time_per_iteration": 3.4645445346832275 + }, + { + "auxiliary_loss_clip": 0.01206062, + "auxiliary_loss_mlp": 0.0103083, + "balance_loss_clip": 1.05707693, + "balance_loss_mlp": 1.02242017, + "epoch": 0.3180424457403956, + "flos": 15485759089920.0, + "grad_norm": 2.9069581822306185, + "language_loss": 0.72586894, + "learning_rate": 3.191144413239916e-06, + "loss": 0.74823785, + "num_input_tokens_seen": 56952460, + "step": 2645, + "time_per_iteration": 2.6227097511291504 + }, + { + "auxiliary_loss_clip": 0.01218023, + "auxiliary_loss_mlp": 0.01034167, + "balance_loss_clip": 1.05668056, + "balance_loss_mlp": 1.02519083, + "epoch": 0.3181626886310347, + "flos": 26174964648960.0, + "grad_norm": 2.251501080056513, + "language_loss": 0.88178623, + "learning_rate": 3.190518573598534e-06, + "loss": 0.90430814, + "num_input_tokens_seen": 56969065, + "step": 2646, + "time_per_iteration": 2.7080023288726807 + }, + { + "auxiliary_loss_clip": 0.01229579, + "auxiliary_loss_mlp": 0.01037045, + "balance_loss_clip": 1.0529964, + "balance_loss_mlp": 1.02791953, + "epoch": 0.3182829315216738, + "flos": 25483109811840.0, + "grad_norm": 1.4314170134488031, + "language_loss": 0.7747224, + "learning_rate": 3.1898925533564308e-06, + "loss": 0.79738861, + "num_input_tokens_seen": 56990535, + "step": 2647, + "time_per_iteration": 2.7703826427459717 + }, + { + "auxiliary_loss_clip": 0.01212734, + "auxiliary_loss_mlp": 0.01035824, + "balance_loss_clip": 1.05217969, + "balance_loss_mlp": 1.0269196, + "epoch": 0.31840317441231286, + "flos": 18113701927680.0, + "grad_norm": 1.9435497007760945, + "language_loss": 0.63769782, + "learning_rate": 3.1892663526085733e-06, + "loss": 0.66018337, + "num_input_tokens_seen": 57008910, + "step": 2648, + "time_per_iteration": 2.770443916320801 + }, + { + "auxiliary_loss_clip": 0.01089107, + "auxiliary_loss_mlp": 0.00999046, + "balance_loss_clip": 1.01981068, + "balance_loss_mlp": 0.99704379, + "epoch": 0.31852341730295197, + "flos": 64741948957440.0, + "grad_norm": 0.750290127016887, + "language_loss": 0.56919992, + "learning_rate": 3.188639971449956e-06, + "loss": 0.59008145, + "num_input_tokens_seen": 57074960, + "step": 2649, + "time_per_iteration": 3.185398817062378 + }, + { + "auxiliary_loss_clip": 0.01203362, + "auxiliary_loss_mlp": 0.01029542, + "balance_loss_clip": 1.0606029, + "balance_loss_mlp": 1.02035689, + "epoch": 0.318643660193591, + "flos": 20668135582080.0, + "grad_norm": 2.2732695817296436, + "language_loss": 0.72481191, + "learning_rate": 3.1880134099756e-06, + "loss": 0.747141, + "num_input_tokens_seen": 57094595, + "step": 2650, + "time_per_iteration": 2.656658172607422 + }, + { + "auxiliary_loss_clip": 0.01204106, + "auxiliary_loss_mlp": 0.01035229, + "balance_loss_clip": 1.05411887, + "balance_loss_mlp": 1.02647352, + "epoch": 0.31876390308423014, + "flos": 26943345411840.0, + "grad_norm": 1.7345738188907802, + "language_loss": 0.69541323, + "learning_rate": 3.1873866682805535e-06, + "loss": 0.71780658, + "num_input_tokens_seen": 57115290, + "step": 2651, + "time_per_iteration": 2.6791272163391113 + }, + { + "auxiliary_loss_clip": 0.01220158, + "auxiliary_loss_mlp": 0.01031053, + "balance_loss_clip": 1.05472362, + "balance_loss_mlp": 1.02148676, + "epoch": 0.31888414597486925, + "flos": 18041916597120.0, + "grad_norm": 1.8309139712465359, + "language_loss": 0.88751811, + "learning_rate": 3.186759746459894e-06, + "loss": 0.91003025, + "num_input_tokens_seen": 57134400, + "step": 2652, + "time_per_iteration": 2.7693748474121094 + }, + { + "auxiliary_loss_clip": 0.01214246, + "auxiliary_loss_mlp": 0.01028588, + "balance_loss_clip": 1.05390251, + "balance_loss_mlp": 1.01993978, + "epoch": 0.3190043888655083, + "flos": 25149319701120.0, + "grad_norm": 1.7276389830993057, + "language_loss": 0.79253078, + "learning_rate": 3.1861326446087246e-06, + "loss": 0.81495917, + "num_input_tokens_seen": 57153140, + "step": 2653, + "time_per_iteration": 2.674635410308838 + }, + { + "auxiliary_loss_clip": 0.01213471, + "auxiliary_loss_mlp": 0.01036728, + "balance_loss_clip": 1.05788898, + "balance_loss_mlp": 1.02704835, + "epoch": 0.3191246317561474, + "flos": 22053892331520.0, + "grad_norm": 2.5592035949429692, + "language_loss": 0.71609169, + "learning_rate": 3.1855053628221763e-06, + "loss": 0.7385937, + "num_input_tokens_seen": 57172395, + "step": 2654, + "time_per_iteration": 2.6342642307281494 + }, + { + "auxiliary_loss_clip": 0.01212276, + "auxiliary_loss_mlp": 0.01035133, + "balance_loss_clip": 1.04987383, + "balance_loss_mlp": 1.02551901, + "epoch": 0.3192448746467865, + "flos": 14901815687040.0, + "grad_norm": 2.8831623105121014, + "language_loss": 0.90098548, + "learning_rate": 3.184877901195407e-06, + "loss": 0.92345953, + "num_input_tokens_seen": 57189090, + "step": 2655, + "time_per_iteration": 2.6442697048187256 + }, + { + "auxiliary_loss_clip": 0.01122523, + "auxiliary_loss_mlp": 0.0100535, + "balance_loss_clip": 1.03352523, + "balance_loss_mlp": 1.0037291, + "epoch": 0.3193651175374256, + "flos": 67234832657280.0, + "grad_norm": 0.8876539511067907, + "language_loss": 0.62773597, + "learning_rate": 3.184250259823602e-06, + "loss": 0.64901471, + "num_input_tokens_seen": 57251620, + "step": 2656, + "time_per_iteration": 3.323624610900879 + }, + { + "auxiliary_loss_clip": 0.01229536, + "auxiliary_loss_mlp": 0.01035984, + "balance_loss_clip": 1.05662024, + "balance_loss_mlp": 1.02607775, + "epoch": 0.3194853604280647, + "flos": 12233077977600.0, + "grad_norm": 2.3036548469891063, + "language_loss": 0.82336175, + "learning_rate": 3.183622438801974e-06, + "loss": 0.84601694, + "num_input_tokens_seen": 57266910, + "step": 2657, + "time_per_iteration": 2.7602076530456543 + }, + { + "auxiliary_loss_clip": 0.01204177, + "auxiliary_loss_mlp": 0.01032175, + "balance_loss_clip": 1.06345618, + "balance_loss_mlp": 1.02299023, + "epoch": 0.3196056033187038, + "flos": 14939917038720.0, + "grad_norm": 1.7166173689267423, + "language_loss": 0.75279123, + "learning_rate": 3.1829944382257637e-06, + "loss": 0.77515477, + "num_input_tokens_seen": 57285040, + "step": 2658, + "time_per_iteration": 2.5716419219970703 + }, + { + "auxiliary_loss_clip": 0.01207478, + "auxiliary_loss_mlp": 0.01031677, + "balance_loss_clip": 1.0592258, + "balance_loss_mlp": 1.02324891, + "epoch": 0.31972584620934286, + "flos": 23768878164480.0, + "grad_norm": 2.234588531213843, + "language_loss": 0.81506014, + "learning_rate": 3.1823662581902373e-06, + "loss": 0.8374517, + "num_input_tokens_seen": 57302725, + "step": 2659, + "time_per_iteration": 2.6786935329437256 + }, + { + "auxiliary_loss_clip": 0.01208355, + "auxiliary_loss_mlp": 0.01033869, + "balance_loss_clip": 1.04970503, + "balance_loss_mlp": 1.02486324, + "epoch": 0.31984608909998197, + "flos": 21251540280960.0, + "grad_norm": 4.583995479813596, + "language_loss": 0.7485714, + "learning_rate": 3.1817378987906896e-06, + "loss": 0.77099365, + "num_input_tokens_seen": 57322230, + "step": 2660, + "time_per_iteration": 2.7028541564941406 + }, + { + "auxiliary_loss_clip": 0.01227589, + "auxiliary_loss_mlp": 0.01031468, + "balance_loss_clip": 1.05420709, + "balance_loss_mlp": 1.02191329, + "epoch": 0.3199663319906211, + "flos": 18296235866880.0, + "grad_norm": 2.049079708621194, + "language_loss": 0.79619396, + "learning_rate": 3.181109360122442e-06, + "loss": 0.81878459, + "num_input_tokens_seen": 57339820, + "step": 2661, + "time_per_iteration": 2.7935941219329834 + }, + { + "auxiliary_loss_clip": 0.01220981, + "auxiliary_loss_mlp": 0.0103128, + "balance_loss_clip": 1.05123079, + "balance_loss_mlp": 1.02219021, + "epoch": 0.32008657488126013, + "flos": 18733627779840.0, + "grad_norm": 2.228445738772542, + "language_loss": 0.78198838, + "learning_rate": 3.1804806422808445e-06, + "loss": 0.80451095, + "num_input_tokens_seen": 57356955, + "step": 2662, + "time_per_iteration": 2.6412177085876465 + }, + { + "auxiliary_loss_clip": 0.01207004, + "auxiliary_loss_mlp": 0.01035784, + "balance_loss_clip": 1.0541358, + "balance_loss_mlp": 1.02649212, + "epoch": 0.32020681777189924, + "flos": 20595344670720.0, + "grad_norm": 1.5713846699962173, + "language_loss": 0.73036021, + "learning_rate": 3.1798517453612714e-06, + "loss": 0.75278807, + "num_input_tokens_seen": 57376760, + "step": 2663, + "time_per_iteration": 2.718282699584961 + }, + { + "auxiliary_loss_clip": 0.01209194, + "auxiliary_loss_mlp": 0.01038885, + "balance_loss_clip": 1.06058371, + "balance_loss_mlp": 1.03009927, + "epoch": 0.32032706066253835, + "flos": 35261692750080.0, + "grad_norm": 1.6352164097528306, + "language_loss": 0.7501545, + "learning_rate": 3.1792226694591265e-06, + "loss": 0.77263522, + "num_input_tokens_seen": 57398145, + "step": 2664, + "time_per_iteration": 3.742324113845825 + }, + { + "auxiliary_loss_clip": 0.01221741, + "auxiliary_loss_mlp": 0.01037531, + "balance_loss_clip": 1.05372143, + "balance_loss_mlp": 1.02813172, + "epoch": 0.3204473035531774, + "flos": 15304230731520.0, + "grad_norm": 1.8356078527986548, + "language_loss": 0.80061388, + "learning_rate": 3.178593414669841e-06, + "loss": 0.82320654, + "num_input_tokens_seen": 57416730, + "step": 2665, + "time_per_iteration": 4.528234243392944 + }, + { + "auxiliary_loss_clip": 0.01213608, + "auxiliary_loss_mlp": 0.01030924, + "balance_loss_clip": 1.0607059, + "balance_loss_mlp": 1.02181041, + "epoch": 0.3205675464438165, + "flos": 24462564595200.0, + "grad_norm": 2.1350411010945476, + "language_loss": 0.70966649, + "learning_rate": 3.1779639810888707e-06, + "loss": 0.73211181, + "num_input_tokens_seen": 57436325, + "step": 2666, + "time_per_iteration": 2.6200153827667236 + }, + { + "auxiliary_loss_clip": 0.01207863, + "auxiliary_loss_mlp": 0.01033853, + "balance_loss_clip": 1.06034756, + "balance_loss_mlp": 1.02494824, + "epoch": 0.3206877893344556, + "flos": 22456235548800.0, + "grad_norm": 1.9088754389624256, + "language_loss": 0.76428246, + "learning_rate": 3.1773343688117013e-06, + "loss": 0.78669965, + "num_input_tokens_seen": 57457235, + "step": 2667, + "time_per_iteration": 2.666962146759033 + }, + { + "auxiliary_loss_clip": 0.01217015, + "auxiliary_loss_mlp": 0.01379906, + "balance_loss_clip": 1.0548352, + "balance_loss_mlp": 1.00038302, + "epoch": 0.3208080322250947, + "flos": 20412236113920.0, + "grad_norm": 3.0168892276250165, + "language_loss": 0.8388623, + "learning_rate": 3.1767045779338445e-06, + "loss": 0.86483151, + "num_input_tokens_seen": 57474895, + "step": 2668, + "time_per_iteration": 2.675658941268921 + }, + { + "auxiliary_loss_clip": 0.01206714, + "auxiliary_loss_mlp": 0.01027253, + "balance_loss_clip": 1.05637264, + "balance_loss_mlp": 1.01889634, + "epoch": 0.3209282751157338, + "flos": 21762118154880.0, + "grad_norm": 4.362797181395253, + "language_loss": 0.91182959, + "learning_rate": 3.176074608550839e-06, + "loss": 0.93416929, + "num_input_tokens_seen": 57490715, + "step": 2669, + "time_per_iteration": 2.6522724628448486 + }, + { + "auxiliary_loss_clip": 0.01241573, + "auxiliary_loss_mlp": 0.01034312, + "balance_loss_clip": 1.05242658, + "balance_loss_mlp": 1.02504349, + "epoch": 0.32104851800637285, + "flos": 22055041566720.0, + "grad_norm": 2.3650125788626424, + "language_loss": 0.82139295, + "learning_rate": 3.17544446075825e-06, + "loss": 0.84415179, + "num_input_tokens_seen": 57509880, + "step": 2670, + "time_per_iteration": 3.662062168121338 + }, + { + "auxiliary_loss_clip": 0.01224336, + "auxiliary_loss_mlp": 0.01035162, + "balance_loss_clip": 1.05733776, + "balance_loss_mlp": 1.02690125, + "epoch": 0.32116876089701196, + "flos": 37012301896320.0, + "grad_norm": 1.5777867031442492, + "language_loss": 0.7090292, + "learning_rate": 3.174814134651671e-06, + "loss": 0.73162425, + "num_input_tokens_seen": 57532430, + "step": 2671, + "time_per_iteration": 2.774160623550415 + }, + { + "auxiliary_loss_clip": 0.01198556, + "auxiliary_loss_mlp": 0.0103027, + "balance_loss_clip": 1.05957639, + "balance_loss_mlp": 1.02134693, + "epoch": 0.3212890037876511, + "flos": 21979233912960.0, + "grad_norm": 1.7521158447058442, + "language_loss": 0.80381709, + "learning_rate": 3.1741836303267215e-06, + "loss": 0.82610536, + "num_input_tokens_seen": 57551965, + "step": 2672, + "time_per_iteration": 2.652170181274414 + }, + { + "auxiliary_loss_clip": 0.01198124, + "auxiliary_loss_mlp": 0.0103331, + "balance_loss_clip": 1.0599823, + "balance_loss_mlp": 1.0240947, + "epoch": 0.32140924667829013, + "flos": 10342345875840.0, + "grad_norm": 1.9028267017627796, + "language_loss": 0.75355905, + "learning_rate": 3.1735529478790496e-06, + "loss": 0.77587336, + "num_input_tokens_seen": 57569955, + "step": 2673, + "time_per_iteration": 2.5874037742614746 + }, + { + "auxiliary_loss_clip": 0.0121283, + "auxiliary_loss_mlp": 0.0103297, + "balance_loss_clip": 1.06093407, + "balance_loss_mlp": 1.02385688, + "epoch": 0.32152948956892924, + "flos": 50798910072960.0, + "grad_norm": 1.949648172801611, + "language_loss": 0.79936403, + "learning_rate": 3.172922087404328e-06, + "loss": 0.82182205, + "num_input_tokens_seen": 57592215, + "step": 2674, + "time_per_iteration": 2.971597194671631 + }, + { + "auxiliary_loss_clip": 0.01092635, + "auxiliary_loss_mlp": 0.01006274, + "balance_loss_clip": 1.02334428, + "balance_loss_mlp": 1.00447416, + "epoch": 0.32164973245956835, + "flos": 63863250549120.0, + "grad_norm": 0.76701660997881, + "language_loss": 0.55241692, + "learning_rate": 3.1722910489982586e-06, + "loss": 0.57340598, + "num_input_tokens_seen": 57652575, + "step": 2675, + "time_per_iteration": 3.2499163150787354 + }, + { + "auxiliary_loss_clip": 0.01209446, + "auxiliary_loss_mlp": 0.01034893, + "balance_loss_clip": 1.05565941, + "balance_loss_mlp": 1.02587509, + "epoch": 0.3217699753502074, + "flos": 23513948363520.0, + "grad_norm": 1.5575577917286076, + "language_loss": 0.80073047, + "learning_rate": 3.1716598327565694e-06, + "loss": 0.82317388, + "num_input_tokens_seen": 57672215, + "step": 2676, + "time_per_iteration": 2.7638208866119385 + }, + { + "auxiliary_loss_clip": 0.01200327, + "auxiliary_loss_mlp": 0.01028284, + "balance_loss_clip": 1.06037974, + "balance_loss_mlp": 1.01989806, + "epoch": 0.3218902182408465, + "flos": 19062533640960.0, + "grad_norm": 1.5454066765349839, + "language_loss": 0.8383159, + "learning_rate": 3.171028438775015e-06, + "loss": 0.86060202, + "num_input_tokens_seen": 57691410, + "step": 2677, + "time_per_iteration": 2.650782585144043 + }, + { + "auxiliary_loss_clip": 0.01197484, + "auxiliary_loss_mlp": 0.01032296, + "balance_loss_clip": 1.05809164, + "balance_loss_mlp": 1.02352202, + "epoch": 0.3220104611314856, + "flos": 20375571306240.0, + "grad_norm": 1.9097213240560482, + "language_loss": 0.83812261, + "learning_rate": 3.170396867149377e-06, + "loss": 0.86042041, + "num_input_tokens_seen": 57709415, + "step": 2678, + "time_per_iteration": 2.6275620460510254 + }, + { + "auxiliary_loss_clip": 0.01226366, + "auxiliary_loss_mlp": 0.01032421, + "balance_loss_clip": 1.05254471, + "balance_loss_mlp": 1.02334917, + "epoch": 0.3221307040221247, + "flos": 20117014231680.0, + "grad_norm": 1.8349248086907717, + "language_loss": 0.86511171, + "learning_rate": 3.1697651179754653e-06, + "loss": 0.8876996, + "num_input_tokens_seen": 57728075, + "step": 2679, + "time_per_iteration": 2.7016539573669434 + }, + { + "auxiliary_loss_clip": 0.01232525, + "auxiliary_loss_mlp": 0.01033259, + "balance_loss_clip": 1.06077313, + "balance_loss_mlp": 1.02378178, + "epoch": 0.3222509469127638, + "flos": 23987789602560.0, + "grad_norm": 1.7586010178054454, + "language_loss": 0.72841156, + "learning_rate": 3.1691331913491153e-06, + "loss": 0.75106943, + "num_input_tokens_seen": 57750645, + "step": 2680, + "time_per_iteration": 2.859776020050049 + }, + { + "auxiliary_loss_clip": 0.01203149, + "auxiliary_loss_mlp": 0.01029941, + "balance_loss_clip": 1.06045508, + "balance_loss_mlp": 1.02160847, + "epoch": 0.32237118980340285, + "flos": 17675735397120.0, + "grad_norm": 1.9620219766538598, + "language_loss": 0.84519434, + "learning_rate": 3.1685010873661898e-06, + "loss": 0.86752522, + "num_input_tokens_seen": 57769820, + "step": 2681, + "time_per_iteration": 2.5790603160858154 + }, + { + "auxiliary_loss_clip": 0.01204948, + "auxiliary_loss_mlp": 0.01027579, + "balance_loss_clip": 1.05782735, + "balance_loss_mlp": 1.01782775, + "epoch": 0.32249143269404196, + "flos": 23147982645120.0, + "grad_norm": 2.26570186176679, + "language_loss": 0.79428434, + "learning_rate": 3.167868806122578e-06, + "loss": 0.81660962, + "num_input_tokens_seen": 57788870, + "step": 2682, + "time_per_iteration": 2.707307815551758 + }, + { + "auxiliary_loss_clip": 0.01225687, + "auxiliary_loss_mlp": 0.01038063, + "balance_loss_clip": 1.06097031, + "balance_loss_mlp": 1.02896118, + "epoch": 0.32261167558468107, + "flos": 24422308427520.0, + "grad_norm": 1.944872426243691, + "language_loss": 0.65974724, + "learning_rate": 3.1672363477141968e-06, + "loss": 0.68238473, + "num_input_tokens_seen": 57808165, + "step": 2683, + "time_per_iteration": 2.7196366786956787 + }, + { + "auxiliary_loss_clip": 0.01223277, + "auxiliary_loss_mlp": 0.01031965, + "balance_loss_clip": 1.05829287, + "balance_loss_mlp": 1.02244592, + "epoch": 0.3227319184753201, + "flos": 30367175852160.0, + "grad_norm": 1.9493172010621196, + "language_loss": 0.85005319, + "learning_rate": 3.1666037122369903e-06, + "loss": 0.87260556, + "num_input_tokens_seen": 57828825, + "step": 2684, + "time_per_iteration": 2.703307867050171 + }, + { + "auxiliary_loss_clip": 0.01204607, + "auxiliary_loss_mlp": 0.01029514, + "balance_loss_clip": 1.05541539, + "balance_loss_mlp": 1.02010846, + "epoch": 0.32285216136595923, + "flos": 16946174257920.0, + "grad_norm": 2.402173398488981, + "language_loss": 0.86993712, + "learning_rate": 3.165970899786928e-06, + "loss": 0.89227831, + "num_input_tokens_seen": 57846740, + "step": 2685, + "time_per_iteration": 2.6289618015289307 + }, + { + "auxiliary_loss_clip": 0.01235238, + "auxiliary_loss_mlp": 0.01027172, + "balance_loss_clip": 1.05833459, + "balance_loss_mlp": 1.0180409, + "epoch": 0.32297240425659834, + "flos": 21981532383360.0, + "grad_norm": 1.910962346581801, + "language_loss": 0.75479633, + "learning_rate": 3.1653379104600067e-06, + "loss": 0.7774204, + "num_input_tokens_seen": 57866885, + "step": 2686, + "time_per_iteration": 2.7883570194244385 + }, + { + "auxiliary_loss_clip": 0.01208299, + "auxiliary_loss_mlp": 0.01034708, + "balance_loss_clip": 1.05725873, + "balance_loss_mlp": 1.02613091, + "epoch": 0.3230926471472374, + "flos": 22748045639040.0, + "grad_norm": 1.7275781939522408, + "language_loss": 0.69644409, + "learning_rate": 3.164704744352251e-06, + "loss": 0.7188741, + "num_input_tokens_seen": 57887690, + "step": 2687, + "time_per_iteration": 2.690736770629883 + }, + { + "auxiliary_loss_clip": 0.01206303, + "auxiliary_loss_mlp": 0.01030211, + "balance_loss_clip": 1.0561471, + "balance_loss_mlp": 1.0213542, + "epoch": 0.3232128900378765, + "flos": 16942977947520.0, + "grad_norm": 1.7546574650105442, + "language_loss": 0.81099856, + "learning_rate": 3.164071401559713e-06, + "loss": 0.83336365, + "num_input_tokens_seen": 57905090, + "step": 2688, + "time_per_iteration": 2.598273515701294 + }, + { + "auxiliary_loss_clip": 0.01219585, + "auxiliary_loss_mlp": 0.01034218, + "balance_loss_clip": 1.05787671, + "balance_loss_mlp": 1.02531934, + "epoch": 0.3233331329285156, + "flos": 24023736138240.0, + "grad_norm": 2.1251819318226746, + "language_loss": 0.71380448, + "learning_rate": 3.1634378821784674e-06, + "loss": 0.73634249, + "num_input_tokens_seen": 57925305, + "step": 2689, + "time_per_iteration": 2.745732307434082 + }, + { + "auxiliary_loss_clip": 0.012288, + "auxiliary_loss_mlp": 0.01032724, + "balance_loss_clip": 1.05617881, + "balance_loss_mlp": 1.02319908, + "epoch": 0.3234533758191547, + "flos": 18113845582080.0, + "grad_norm": 2.3657067344954847, + "language_loss": 0.74091101, + "learning_rate": 3.1628041863046208e-06, + "loss": 0.76352626, + "num_input_tokens_seen": 57942720, + "step": 2690, + "time_per_iteration": 3.6363625526428223 + }, + { + "auxiliary_loss_clip": 0.01203322, + "auxiliary_loss_mlp": 0.0103066, + "balance_loss_clip": 1.05792189, + "balance_loss_mlp": 1.02128398, + "epoch": 0.3235736187097938, + "flos": 16946138344320.0, + "grad_norm": 2.70298099095619, + "language_loss": 0.91184545, + "learning_rate": 3.162170314034304e-06, + "loss": 0.93418527, + "num_input_tokens_seen": 57960135, + "step": 2691, + "time_per_iteration": 4.31362509727478 + }, + { + "auxiliary_loss_clip": 0.01203408, + "auxiliary_loss_mlp": 0.01038416, + "balance_loss_clip": 1.05938315, + "balance_loss_mlp": 1.02826548, + "epoch": 0.3236938616004329, + "flos": 22127150119680.0, + "grad_norm": 1.516434311872386, + "language_loss": 0.80813378, + "learning_rate": 3.1615362654636738e-06, + "loss": 0.83055204, + "num_input_tokens_seen": 57980875, + "step": 2692, + "time_per_iteration": 2.6981866359710693 + }, + { + "auxiliary_loss_clip": 0.01223924, + "auxiliary_loss_mlp": 0.01033023, + "balance_loss_clip": 1.06013596, + "balance_loss_mlp": 1.02445173, + "epoch": 0.32381410449107195, + "flos": 17164618819200.0, + "grad_norm": 1.9255065708517032, + "language_loss": 0.87091041, + "learning_rate": 3.1609020406889163e-06, + "loss": 0.89347994, + "num_input_tokens_seen": 57998310, + "step": 2693, + "time_per_iteration": 2.691805362701416 + }, + { + "auxiliary_loss_clip": 0.01215325, + "auxiliary_loss_mlp": 0.01034323, + "balance_loss_clip": 1.05463576, + "balance_loss_mlp": 1.02421403, + "epoch": 0.32393434738171106, + "flos": 16578125550720.0, + "grad_norm": 1.7074549296633288, + "language_loss": 0.85014671, + "learning_rate": 3.1602676398062416e-06, + "loss": 0.87264323, + "num_input_tokens_seen": 58017220, + "step": 2694, + "time_per_iteration": 2.680070638656616 + }, + { + "auxiliary_loss_clip": 0.01209562, + "auxiliary_loss_mlp": 0.01031791, + "balance_loss_clip": 1.05968237, + "balance_loss_mlp": 1.02258193, + "epoch": 0.3240545902723502, + "flos": 25483612602240.0, + "grad_norm": 2.127709133673669, + "language_loss": 0.61776316, + "learning_rate": 3.1596330629118886e-06, + "loss": 0.64017671, + "num_input_tokens_seen": 58037190, + "step": 2695, + "time_per_iteration": 2.6943063735961914 + }, + { + "auxiliary_loss_clip": 0.0122693, + "auxiliary_loss_mlp": 0.01037944, + "balance_loss_clip": 1.05256045, + "balance_loss_mlp": 1.02833021, + "epoch": 0.32417483316298923, + "flos": 35845851634560.0, + "grad_norm": 2.0926441833255, + "language_loss": 0.73109496, + "learning_rate": 3.1589983101021223e-06, + "loss": 0.75374371, + "num_input_tokens_seen": 58055820, + "step": 2696, + "time_per_iteration": 2.905855655670166 + }, + { + "auxiliary_loss_clip": 0.01222933, + "auxiliary_loss_mlp": 0.01032958, + "balance_loss_clip": 1.05858517, + "balance_loss_mlp": 1.02428269, + "epoch": 0.32429507605362834, + "flos": 30080501406720.0, + "grad_norm": 2.096802981335645, + "language_loss": 0.84501338, + "learning_rate": 3.1583633814732337e-06, + "loss": 0.86757231, + "num_input_tokens_seen": 58075340, + "step": 2697, + "time_per_iteration": 3.63531756401062 + }, + { + "auxiliary_loss_clip": 0.01200669, + "auxiliary_loss_mlp": 0.01029429, + "balance_loss_clip": 1.05686092, + "balance_loss_mlp": 1.02014899, + "epoch": 0.3244153189442674, + "flos": 18223265387520.0, + "grad_norm": 2.8391263401752505, + "language_loss": 0.71889943, + "learning_rate": 3.157728277121541e-06, + "loss": 0.74120045, + "num_input_tokens_seen": 58093515, + "step": 2698, + "time_per_iteration": 2.6074490547180176 + }, + { + "auxiliary_loss_clip": 0.01200631, + "auxiliary_loss_mlp": 0.01032386, + "balance_loss_clip": 1.05706394, + "balance_loss_mlp": 1.02318263, + "epoch": 0.3245355618349065, + "flos": 17710317216000.0, + "grad_norm": 2.359876683346954, + "language_loss": 0.78737497, + "learning_rate": 3.1570929971433897e-06, + "loss": 0.80970514, + "num_input_tokens_seen": 58109300, + "step": 2699, + "time_per_iteration": 2.5909390449523926 + }, + { + "auxiliary_loss_clip": 0.01211484, + "auxiliary_loss_mlp": 0.01038806, + "balance_loss_clip": 1.06069708, + "balance_loss_mlp": 1.03005552, + "epoch": 0.3246558047255456, + "flos": 23440798316160.0, + "grad_norm": 1.8142388543908625, + "language_loss": 0.83690071, + "learning_rate": 3.1564575416351504e-06, + "loss": 0.85940361, + "num_input_tokens_seen": 58128000, + "step": 2700, + "time_per_iteration": 2.635230779647827 + }, + { + "auxiliary_loss_clip": 0.01202577, + "auxiliary_loss_mlp": 0.01033542, + "balance_loss_clip": 1.06135082, + "balance_loss_mlp": 1.02446389, + "epoch": 0.32477604761618467, + "flos": 21760861178880.0, + "grad_norm": 1.7271821459727166, + "language_loss": 0.74312723, + "learning_rate": 3.155821910693221e-06, + "loss": 0.76548845, + "num_input_tokens_seen": 58147415, + "step": 2701, + "time_per_iteration": 2.602280378341675 + }, + { + "auxiliary_loss_clip": 0.01214717, + "auxiliary_loss_mlp": 0.01034156, + "balance_loss_clip": 1.05451083, + "balance_loss_mlp": 1.0245533, + "epoch": 0.3248962905068238, + "flos": 19828328624640.0, + "grad_norm": 1.6628551082775496, + "language_loss": 0.85932195, + "learning_rate": 3.1551861044140275e-06, + "loss": 0.88181067, + "num_input_tokens_seen": 58167050, + "step": 2702, + "time_per_iteration": 2.677027463912964 + }, + { + "auxiliary_loss_clip": 0.01228106, + "auxiliary_loss_mlp": 0.0103673, + "balance_loss_clip": 1.05713677, + "balance_loss_mlp": 1.02752721, + "epoch": 0.3250165333974629, + "flos": 23948215793280.0, + "grad_norm": 1.6678779812131352, + "language_loss": 0.77506602, + "learning_rate": 3.15455012289402e-06, + "loss": 0.79771447, + "num_input_tokens_seen": 58186695, + "step": 2703, + "time_per_iteration": 2.808863401412964 + }, + { + "auxiliary_loss_clip": 0.01211711, + "auxiliary_loss_mlp": 0.01030964, + "balance_loss_clip": 1.05991721, + "balance_loss_mlp": 1.02140355, + "epoch": 0.32513677628810195, + "flos": 23989333887360.0, + "grad_norm": 1.7917341762391288, + "language_loss": 0.83881986, + "learning_rate": 3.153913966229677e-06, + "loss": 0.86124659, + "num_input_tokens_seen": 58205815, + "step": 2704, + "time_per_iteration": 2.727860689163208 + }, + { + "auxiliary_loss_clip": 0.01100309, + "auxiliary_loss_mlp": 0.01000556, + "balance_loss_clip": 1.01781821, + "balance_loss_mlp": 0.99879134, + "epoch": 0.32525701917874106, + "flos": 70655790009600.0, + "grad_norm": 0.6797710722341073, + "language_loss": 0.50271779, + "learning_rate": 3.1532776345175027e-06, + "loss": 0.52372646, + "num_input_tokens_seen": 58270960, + "step": 2705, + "time_per_iteration": 3.1929333209991455 + }, + { + "auxiliary_loss_clip": 0.01198007, + "auxiliary_loss_mlp": 0.01034606, + "balance_loss_clip": 1.05776191, + "balance_loss_mlp": 1.0253849, + "epoch": 0.32537726206938017, + "flos": 19682639061120.0, + "grad_norm": 1.9277147881969603, + "language_loss": 0.78749537, + "learning_rate": 3.1526411278540285e-06, + "loss": 0.80982149, + "num_input_tokens_seen": 58289390, + "step": 2706, + "time_per_iteration": 2.7264490127563477 + }, + { + "auxiliary_loss_clip": 0.01222305, + "auxiliary_loss_mlp": 0.01033823, + "balance_loss_clip": 1.05538726, + "balance_loss_mlp": 1.02486408, + "epoch": 0.3254975049600192, + "flos": 28760999293440.0, + "grad_norm": 2.709032185760867, + "language_loss": 0.81468546, + "learning_rate": 3.1520044463358116e-06, + "loss": 0.83724678, + "num_input_tokens_seen": 58306120, + "step": 2707, + "time_per_iteration": 2.748154878616333 + }, + { + "auxiliary_loss_clip": 0.01211064, + "auxiliary_loss_mlp": 0.01028675, + "balance_loss_clip": 1.0596323, + "balance_loss_mlp": 1.02001488, + "epoch": 0.32561774785065833, + "flos": 18877378008960.0, + "grad_norm": 1.4715978856890088, + "language_loss": 0.79965192, + "learning_rate": 3.151367590059436e-06, + "loss": 0.82204932, + "num_input_tokens_seen": 58324545, + "step": 2708, + "time_per_iteration": 2.667694330215454 + }, + { + "auxiliary_loss_clip": 0.01204008, + "auxiliary_loss_mlp": 0.01380407, + "balance_loss_clip": 1.06045365, + "balance_loss_mlp": 1.00048196, + "epoch": 0.32573799074129745, + "flos": 23112107936640.0, + "grad_norm": 2.063362878743879, + "language_loss": 0.8672663, + "learning_rate": 3.1507305591215117e-06, + "loss": 0.89311051, + "num_input_tokens_seen": 58342455, + "step": 2709, + "time_per_iteration": 2.6575026512145996 + }, + { + "auxiliary_loss_clip": 0.01099904, + "auxiliary_loss_mlp": 0.01004176, + "balance_loss_clip": 1.01831818, + "balance_loss_mlp": 1.00232816, + "epoch": 0.3258582336319365, + "flos": 71237650423680.0, + "grad_norm": 0.6721860578437736, + "language_loss": 0.55746591, + "learning_rate": 3.150093353618677e-06, + "loss": 0.57850671, + "num_input_tokens_seen": 58407185, + "step": 2710, + "time_per_iteration": 3.2393405437469482 + }, + { + "auxiliary_loss_clip": 0.01214157, + "auxiliary_loss_mlp": 0.01034222, + "balance_loss_clip": 1.05774808, + "balance_loss_mlp": 1.02479303, + "epoch": 0.3259784765225756, + "flos": 22456020067200.0, + "grad_norm": 2.6541483817951455, + "language_loss": 0.88298959, + "learning_rate": 3.149455973647596e-06, + "loss": 0.90547341, + "num_input_tokens_seen": 58425245, + "step": 2711, + "time_per_iteration": 2.639453887939453 + }, + { + "auxiliary_loss_clip": 0.0121642, + "auxiliary_loss_mlp": 0.01032588, + "balance_loss_clip": 1.04991317, + "balance_loss_mlp": 1.02262759, + "epoch": 0.32609871941321467, + "flos": 20484811543680.0, + "grad_norm": 1.941656803353848, + "language_loss": 0.76873469, + "learning_rate": 3.1488184193049563e-06, + "loss": 0.79122478, + "num_input_tokens_seen": 58444780, + "step": 2712, + "time_per_iteration": 2.6895668506622314 + }, + { + "auxiliary_loss_clip": 0.01201624, + "auxiliary_loss_mlp": 0.01029444, + "balance_loss_clip": 1.06047046, + "balance_loss_mlp": 1.02078319, + "epoch": 0.3262189623038538, + "flos": 22416805393920.0, + "grad_norm": 1.5367594371598132, + "language_loss": 0.72212565, + "learning_rate": 3.1481806906874767e-06, + "loss": 0.74443626, + "num_input_tokens_seen": 58466090, + "step": 2713, + "time_per_iteration": 2.650221586227417 + }, + { + "auxiliary_loss_clip": 0.01201056, + "auxiliary_loss_mlp": 0.0103596, + "balance_loss_clip": 1.06028461, + "balance_loss_mlp": 1.02679265, + "epoch": 0.3263392051944929, + "flos": 20923496346240.0, + "grad_norm": 1.5807006878861518, + "language_loss": 0.87390649, + "learning_rate": 3.147542787891899e-06, + "loss": 0.89627665, + "num_input_tokens_seen": 58485435, + "step": 2714, + "time_per_iteration": 2.6016170978546143 + }, + { + "auxiliary_loss_clip": 0.01220803, + "auxiliary_loss_mlp": 0.01031992, + "balance_loss_clip": 1.0584147, + "balance_loss_mlp": 1.0230093, + "epoch": 0.32645944808513194, + "flos": 24025172682240.0, + "grad_norm": 4.929728460730992, + "language_loss": 0.75233251, + "learning_rate": 3.1469047110149926e-06, + "loss": 0.7748605, + "num_input_tokens_seen": 58504175, + "step": 2715, + "time_per_iteration": 2.735809326171875 + }, + { + "auxiliary_loss_clip": 0.01230971, + "auxiliary_loss_mlp": 0.01034622, + "balance_loss_clip": 1.05595827, + "balance_loss_mlp": 1.02459645, + "epoch": 0.32657969097577105, + "flos": 21032413361280.0, + "grad_norm": 2.1594877357560427, + "language_loss": 0.85460889, + "learning_rate": 3.146266460153554e-06, + "loss": 0.8772648, + "num_input_tokens_seen": 58523885, + "step": 2716, + "time_per_iteration": 4.607028245925903 + }, + { + "auxiliary_loss_clip": 0.0121586, + "auxiliary_loss_mlp": 0.01380494, + "balance_loss_clip": 1.05664158, + "balance_loss_mlp": 1.00053835, + "epoch": 0.32669993386641016, + "flos": 22710267509760.0, + "grad_norm": 4.546305293211612, + "language_loss": 0.79991567, + "learning_rate": 3.145628035404404e-06, + "loss": 0.82587922, + "num_input_tokens_seen": 58543085, + "step": 2717, + "time_per_iteration": 3.682678699493408 + }, + { + "auxiliary_loss_clip": 0.01101919, + "auxiliary_loss_mlp": 0.01004026, + "balance_loss_clip": 1.0220902, + "balance_loss_mlp": 1.00210714, + "epoch": 0.3268201767570492, + "flos": 72105718406400.0, + "grad_norm": 0.9188517064682178, + "language_loss": 0.5745796, + "learning_rate": 3.1449894368643922e-06, + "loss": 0.59563899, + "num_input_tokens_seen": 58605400, + "step": 2718, + "time_per_iteration": 3.2877914905548096 + }, + { + "auxiliary_loss_clip": 0.01231675, + "auxiliary_loss_mlp": 0.01034333, + "balance_loss_clip": 1.06171072, + "balance_loss_mlp": 1.02588677, + "epoch": 0.32694041964768833, + "flos": 24535175938560.0, + "grad_norm": 1.576344038766578, + "language_loss": 0.71467352, + "learning_rate": 3.1443506646303934e-06, + "loss": 0.73733354, + "num_input_tokens_seen": 58626700, + "step": 2719, + "time_per_iteration": 2.7871735095977783 + }, + { + "auxiliary_loss_clip": 0.01212481, + "auxiliary_loss_mlp": 0.01032185, + "balance_loss_clip": 1.05812538, + "balance_loss_mlp": 1.02226663, + "epoch": 0.32706066253832744, + "flos": 33183003755520.0, + "grad_norm": 2.1911088695378385, + "language_loss": 0.67143047, + "learning_rate": 3.1437117187993086e-06, + "loss": 0.6938771, + "num_input_tokens_seen": 58649020, + "step": 2720, + "time_per_iteration": 2.7522776126861572 + }, + { + "auxiliary_loss_clip": 0.01218555, + "auxiliary_loss_mlp": 0.01041053, + "balance_loss_clip": 1.05244136, + "balance_loss_mlp": 1.03108704, + "epoch": 0.3271809054289665, + "flos": 24061622008320.0, + "grad_norm": 1.5441423919882196, + "language_loss": 0.7988053, + "learning_rate": 3.143072599468065e-06, + "loss": 0.82140136, + "num_input_tokens_seen": 58668845, + "step": 2721, + "time_per_iteration": 2.7973766326904297 + }, + { + "auxiliary_loss_clip": 0.01219213, + "auxiliary_loss_mlp": 0.01031636, + "balance_loss_clip": 1.05910659, + "balance_loss_mlp": 1.02208173, + "epoch": 0.3273011483196056, + "flos": 38253769712640.0, + "grad_norm": 1.5606096286185607, + "language_loss": 0.7568565, + "learning_rate": 3.1424333067336174e-06, + "loss": 0.779365, + "num_input_tokens_seen": 58691610, + "step": 2722, + "time_per_iteration": 2.8443288803100586 + }, + { + "auxiliary_loss_clip": 0.0121606, + "auxiliary_loss_mlp": 0.01033672, + "balance_loss_clip": 1.0589242, + "balance_loss_mlp": 1.02437401, + "epoch": 0.3274213912102447, + "flos": 29054389582080.0, + "grad_norm": 1.7378008076866975, + "language_loss": 0.7811625, + "learning_rate": 3.141793840692945e-06, + "loss": 0.80365986, + "num_input_tokens_seen": 58712360, + "step": 2723, + "time_per_iteration": 3.6050336360931396 + }, + { + "auxiliary_loss_clip": 0.01208606, + "auxiliary_loss_mlp": 0.0103173, + "balance_loss_clip": 1.05693507, + "balance_loss_mlp": 1.02259231, + "epoch": 0.32754163410088377, + "flos": 29133249891840.0, + "grad_norm": 2.2677455605727586, + "language_loss": 0.61048138, + "learning_rate": 3.1411542014430553e-06, + "loss": 0.63288474, + "num_input_tokens_seen": 58733440, + "step": 2724, + "time_per_iteration": 2.772723913192749 + }, + { + "auxiliary_loss_clip": 0.01221507, + "auxiliary_loss_mlp": 0.01027292, + "balance_loss_clip": 1.04999256, + "balance_loss_mlp": 1.0189234, + "epoch": 0.3276618769915229, + "flos": 20631075724800.0, + "grad_norm": 1.6521493691268418, + "language_loss": 0.82014757, + "learning_rate": 3.1405143890809804e-06, + "loss": 0.84263557, + "num_input_tokens_seen": 58752735, + "step": 2725, + "time_per_iteration": 2.6824657917022705 + }, + { + "auxiliary_loss_clip": 0.0121509, + "auxiliary_loss_mlp": 0.01027754, + "balance_loss_clip": 1.05768895, + "balance_loss_mlp": 1.01954031, + "epoch": 0.327782119882162, + "flos": 18657425076480.0, + "grad_norm": 1.8440541609571615, + "language_loss": 0.69742602, + "learning_rate": 3.1398744037037796e-06, + "loss": 0.71985447, + "num_input_tokens_seen": 58772070, + "step": 2726, + "time_per_iteration": 2.7143921852111816 + }, + { + "auxiliary_loss_clip": 0.01220865, + "auxiliary_loss_mlp": 0.0103057, + "balance_loss_clip": 1.06001019, + "balance_loss_mlp": 1.0217247, + "epoch": 0.32790236277280105, + "flos": 21795802133760.0, + "grad_norm": 2.1609762170302753, + "language_loss": 0.84434021, + "learning_rate": 3.139234245408538e-06, + "loss": 0.86685461, + "num_input_tokens_seen": 58790950, + "step": 2727, + "time_per_iteration": 2.6614110469818115 + }, + { + "auxiliary_loss_clip": 0.01225743, + "auxiliary_loss_mlp": 0.01380128, + "balance_loss_clip": 1.05592132, + "balance_loss_mlp": 1.00049257, + "epoch": 0.32802260566344016, + "flos": 23331414424320.0, + "grad_norm": 1.916142487484874, + "language_loss": 0.7625953, + "learning_rate": 3.1385939142923666e-06, + "loss": 0.78865409, + "num_input_tokens_seen": 58813340, + "step": 2728, + "time_per_iteration": 2.751753330230713 + }, + { + "auxiliary_loss_clip": 0.01218483, + "auxiliary_loss_mlp": 0.01028306, + "balance_loss_clip": 1.05671263, + "balance_loss_mlp": 1.01861989, + "epoch": 0.3281428485540792, + "flos": 24206988349440.0, + "grad_norm": 2.279338277958731, + "language_loss": 0.77692986, + "learning_rate": 3.137953410452405e-06, + "loss": 0.79939777, + "num_input_tokens_seen": 58833610, + "step": 2729, + "time_per_iteration": 2.6992669105529785 + }, + { + "auxiliary_loss_clip": 0.01214982, + "auxiliary_loss_mlp": 0.01037263, + "balance_loss_clip": 1.05582833, + "balance_loss_mlp": 1.02742267, + "epoch": 0.3282630914447183, + "flos": 34128962380800.0, + "grad_norm": 1.7978727863959285, + "language_loss": 0.74589938, + "learning_rate": 3.1373127339858146e-06, + "loss": 0.76842177, + "num_input_tokens_seen": 58856210, + "step": 2730, + "time_per_iteration": 2.8558542728424072 + }, + { + "auxiliary_loss_clip": 0.01220761, + "auxiliary_loss_mlp": 0.01030093, + "balance_loss_clip": 1.05405283, + "balance_loss_mlp": 1.021981, + "epoch": 0.32838333433535744, + "flos": 27600726170880.0, + "grad_norm": 1.7295927017501254, + "language_loss": 0.74638736, + "learning_rate": 3.136671884989787e-06, + "loss": 0.76889586, + "num_input_tokens_seen": 58876120, + "step": 2731, + "time_per_iteration": 2.7670536041259766 + }, + { + "auxiliary_loss_clip": 0.01245528, + "auxiliary_loss_mlp": 0.01039792, + "balance_loss_clip": 1.05288529, + "balance_loss_mlp": 1.0307498, + "epoch": 0.3285035772259965, + "flos": 12349500935040.0, + "grad_norm": 2.437520283382197, + "language_loss": 0.87041509, + "learning_rate": 3.1360308635615383e-06, + "loss": 0.89326829, + "num_input_tokens_seen": 58894660, + "step": 2732, + "time_per_iteration": 2.900879383087158 + }, + { + "auxiliary_loss_clip": 0.01231428, + "auxiliary_loss_mlp": 0.01039975, + "balance_loss_clip": 1.06318653, + "balance_loss_mlp": 1.03044426, + "epoch": 0.3286238201166356, + "flos": 24316084932480.0, + "grad_norm": 1.8344758878577854, + "language_loss": 0.78396791, + "learning_rate": 3.135389669798311e-06, + "loss": 0.80668193, + "num_input_tokens_seen": 58912720, + "step": 2733, + "time_per_iteration": 2.8507537841796875 + }, + { + "auxiliary_loss_clip": 0.01210033, + "auxiliary_loss_mlp": 0.01380362, + "balance_loss_clip": 1.05859184, + "balance_loss_mlp": 1.00049651, + "epoch": 0.3287440630072747, + "flos": 21392812471680.0, + "grad_norm": 2.250248539006235, + "language_loss": 0.79867351, + "learning_rate": 3.134748303797373e-06, + "loss": 0.82457745, + "num_input_tokens_seen": 58930090, + "step": 2734, + "time_per_iteration": 2.6858503818511963 + }, + { + "auxiliary_loss_clip": 0.01235526, + "auxiliary_loss_mlp": 0.01032596, + "balance_loss_clip": 1.0546062, + "balance_loss_mlp": 1.02249873, + "epoch": 0.32886430589791377, + "flos": 23732536579200.0, + "grad_norm": 2.628087426633902, + "language_loss": 0.811795, + "learning_rate": 3.1341067656560203e-06, + "loss": 0.83447623, + "num_input_tokens_seen": 58947935, + "step": 2735, + "time_per_iteration": 2.825038194656372 + }, + { + "auxiliary_loss_clip": 0.01227539, + "auxiliary_loss_mlp": 0.01037431, + "balance_loss_clip": 1.05805182, + "balance_loss_mlp": 1.02829313, + "epoch": 0.3289845487885529, + "flos": 22418708814720.0, + "grad_norm": 2.145156672667493, + "language_loss": 0.86371362, + "learning_rate": 3.133465055471572e-06, + "loss": 0.88636327, + "num_input_tokens_seen": 58967720, + "step": 2736, + "time_per_iteration": 2.7159857749938965 + }, + { + "auxiliary_loss_clip": 0.01220882, + "auxiliary_loss_mlp": 0.01026064, + "balance_loss_clip": 1.05483651, + "balance_loss_mlp": 1.01689696, + "epoch": 0.329104791679192, + "flos": 19682603147520.0, + "grad_norm": 2.363185729445902, + "language_loss": 0.66282511, + "learning_rate": 3.1328231733413767e-06, + "loss": 0.68529451, + "num_input_tokens_seen": 58984360, + "step": 2737, + "time_per_iteration": 2.722405433654785 + }, + { + "auxiliary_loss_clip": 0.01204067, + "auxiliary_loss_mlp": 0.01035563, + "balance_loss_clip": 1.05834401, + "balance_loss_mlp": 1.02603233, + "epoch": 0.32922503456983104, + "flos": 15997234803840.0, + "grad_norm": 2.165296246544036, + "language_loss": 0.90612781, + "learning_rate": 3.1321811193628067e-06, + "loss": 0.92852402, + "num_input_tokens_seen": 59002505, + "step": 2738, + "time_per_iteration": 2.710050582885742 + }, + { + "auxiliary_loss_clip": 0.01211582, + "auxiliary_loss_mlp": 0.01380786, + "balance_loss_clip": 1.06075084, + "balance_loss_mlp": 1.000597, + "epoch": 0.32934527746047015, + "flos": 26834069260800.0, + "grad_norm": 1.9955811520412958, + "language_loss": 0.70053446, + "learning_rate": 3.131538893633261e-06, + "loss": 0.72645819, + "num_input_tokens_seen": 59022065, + "step": 2739, + "time_per_iteration": 2.718442916870117 + }, + { + "auxiliary_loss_clip": 0.01203155, + "auxiliary_loss_mlp": 0.01033226, + "balance_loss_clip": 1.06154323, + "balance_loss_mlp": 1.02372539, + "epoch": 0.32946552035110926, + "flos": 23403774372480.0, + "grad_norm": 2.0901910165945523, + "language_loss": 0.78471339, + "learning_rate": 3.130896496250165e-06, + "loss": 0.80707717, + "num_input_tokens_seen": 59041890, + "step": 2740, + "time_per_iteration": 2.6197702884674072 + }, + { + "auxiliary_loss_clip": 0.01204349, + "auxiliary_loss_mlp": 0.01031775, + "balance_loss_clip": 1.0606401, + "balance_loss_mlp": 1.02280474, + "epoch": 0.3295857632417483, + "flos": 14172470029440.0, + "grad_norm": 2.157531493650716, + "language_loss": 0.86474097, + "learning_rate": 3.1302539273109693e-06, + "loss": 0.88710219, + "num_input_tokens_seen": 59058715, + "step": 2741, + "time_per_iteration": 2.687347888946533 + }, + { + "auxiliary_loss_clip": 0.01209167, + "auxiliary_loss_mlp": 0.01037462, + "balance_loss_clip": 1.05776548, + "balance_loss_mlp": 1.02781248, + "epoch": 0.32970600613238743, + "flos": 22196708807040.0, + "grad_norm": 1.6863589040621578, + "language_loss": 0.80610049, + "learning_rate": 3.1296111869131513e-06, + "loss": 0.82856679, + "num_input_tokens_seen": 59076140, + "step": 2742, + "time_per_iteration": 3.6359782218933105 + }, + { + "auxiliary_loss_clip": 0.01201941, + "auxiliary_loss_mlp": 0.01027726, + "balance_loss_clip": 1.06023228, + "balance_loss_mlp": 1.01868367, + "epoch": 0.32982624902302654, + "flos": 22053784590720.0, + "grad_norm": 1.8129924020880808, + "language_loss": 0.86124218, + "learning_rate": 3.1289682751542153e-06, + "loss": 0.88353884, + "num_input_tokens_seen": 59095700, + "step": 2743, + "time_per_iteration": 3.5249381065368652 + }, + { + "auxiliary_loss_clip": 0.01205776, + "auxiliary_loss_mlp": 0.01032384, + "balance_loss_clip": 1.0578115, + "balance_loss_mlp": 1.02381301, + "epoch": 0.3299464919136656, + "flos": 18661626967680.0, + "grad_norm": 2.1921577230423677, + "language_loss": 0.71160507, + "learning_rate": 3.1283251921316883e-06, + "loss": 0.73398668, + "num_input_tokens_seen": 59113445, + "step": 2744, + "time_per_iteration": 2.672847032546997 + }, + { + "auxiliary_loss_clip": 0.01234806, + "auxiliary_loss_mlp": 0.01039598, + "balance_loss_clip": 1.05646813, + "balance_loss_mlp": 1.03066921, + "epoch": 0.3300667348043047, + "flos": 13407357404160.0, + "grad_norm": 2.098903030749304, + "language_loss": 0.80846351, + "learning_rate": 3.1276819379431277e-06, + "loss": 0.83120757, + "num_input_tokens_seen": 59131535, + "step": 2745, + "time_per_iteration": 2.7309072017669678 + }, + { + "auxiliary_loss_clip": 0.01227504, + "auxiliary_loss_mlp": 0.01380384, + "balance_loss_clip": 1.05866051, + "balance_loss_mlp": 1.00036252, + "epoch": 0.33018697769494376, + "flos": 15742556398080.0, + "grad_norm": 1.9237048547184132, + "language_loss": 0.7510289, + "learning_rate": 3.1270385126861134e-06, + "loss": 0.77710772, + "num_input_tokens_seen": 59149520, + "step": 2746, + "time_per_iteration": 2.696871519088745 + }, + { + "auxiliary_loss_clip": 0.01208427, + "auxiliary_loss_mlp": 0.01029618, + "balance_loss_clip": 1.06447363, + "balance_loss_mlp": 1.02025461, + "epoch": 0.3303072205855829, + "flos": 18258601392000.0, + "grad_norm": 1.9010624368417623, + "language_loss": 0.81870586, + "learning_rate": 3.1263949164582533e-06, + "loss": 0.84108627, + "num_input_tokens_seen": 59169170, + "step": 2747, + "time_per_iteration": 2.6279356479644775 + }, + { + "auxiliary_loss_clip": 0.01204127, + "auxiliary_loss_mlp": 0.01030989, + "balance_loss_clip": 1.06032205, + "balance_loss_mlp": 1.02166724, + "epoch": 0.330427463476222, + "flos": 17749424148480.0, + "grad_norm": 9.271291848743413, + "language_loss": 0.78435874, + "learning_rate": 3.1257511493571797e-06, + "loss": 0.80670989, + "num_input_tokens_seen": 59187675, + "step": 2748, + "time_per_iteration": 2.6926281452178955 + }, + { + "auxiliary_loss_clip": 0.0122752, + "auxiliary_loss_mlp": 0.01026933, + "balance_loss_clip": 1.05504704, + "balance_loss_mlp": 1.01845741, + "epoch": 0.33054770636686104, + "flos": 27162580072320.0, + "grad_norm": 1.7787681097566177, + "language_loss": 0.78550768, + "learning_rate": 3.125107211480552e-06, + "loss": 0.80805224, + "num_input_tokens_seen": 59207610, + "step": 2749, + "time_per_iteration": 3.679558038711548 + }, + { + "auxiliary_loss_clip": 0.0124111, + "auxiliary_loss_mlp": 0.01034516, + "balance_loss_clip": 1.05113721, + "balance_loss_mlp": 1.02560496, + "epoch": 0.33066794925750015, + "flos": 20117193799680.0, + "grad_norm": 1.5751272931663585, + "language_loss": 0.79641306, + "learning_rate": 3.124463102926054e-06, + "loss": 0.81916928, + "num_input_tokens_seen": 59226945, + "step": 2750, + "time_per_iteration": 2.775938034057617 + }, + { + "auxiliary_loss_clip": 0.01106312, + "auxiliary_loss_mlp": 0.01004039, + "balance_loss_clip": 1.03090787, + "balance_loss_mlp": 1.00233448, + "epoch": 0.33078819214813926, + "flos": 70642609718400.0, + "grad_norm": 0.7627139715115135, + "language_loss": 0.61592615, + "learning_rate": 3.1238188237913984e-06, + "loss": 0.63702965, + "num_input_tokens_seen": 59291485, + "step": 2751, + "time_per_iteration": 3.4008257389068604 + }, + { + "auxiliary_loss_clip": 0.01211932, + "auxiliary_loss_mlp": 0.01033623, + "balance_loss_clip": 1.06540847, + "balance_loss_mlp": 1.02409196, + "epoch": 0.3309084350387783, + "flos": 21141940907520.0, + "grad_norm": 2.439265668328458, + "language_loss": 0.76679701, + "learning_rate": 3.1231743741743202e-06, + "loss": 0.78925258, + "num_input_tokens_seen": 59310990, + "step": 2752, + "time_per_iteration": 2.5255537033081055 + }, + { + "auxiliary_loss_clip": 0.01202356, + "auxiliary_loss_mlp": 0.01027781, + "balance_loss_clip": 1.05481887, + "balance_loss_mlp": 1.01796401, + "epoch": 0.3310286779294174, + "flos": 14209350318720.0, + "grad_norm": 2.2273910461951214, + "language_loss": 0.83732402, + "learning_rate": 3.122529754172582e-06, + "loss": 0.8596254, + "num_input_tokens_seen": 59327875, + "step": 2753, + "time_per_iteration": 2.621816396713257 + }, + { + "auxiliary_loss_clip": 0.0121074, + "auxiliary_loss_mlp": 0.01036822, + "balance_loss_clip": 1.0611043, + "balance_loss_mlp": 1.02773833, + "epoch": 0.33114892082005654, + "flos": 20778130005120.0, + "grad_norm": 1.9406254042813074, + "language_loss": 0.72443509, + "learning_rate": 3.1218849638839736e-06, + "loss": 0.74691069, + "num_input_tokens_seen": 59347135, + "step": 2754, + "time_per_iteration": 2.676518678665161 + }, + { + "auxiliary_loss_clip": 0.01209561, + "auxiliary_loss_mlp": 0.01032427, + "balance_loss_clip": 1.05081487, + "balance_loss_mlp": 1.02277672, + "epoch": 0.3312691637106956, + "flos": 17090750499840.0, + "grad_norm": 1.9523476983794548, + "language_loss": 0.78574896, + "learning_rate": 3.121240003406307e-06, + "loss": 0.80816877, + "num_input_tokens_seen": 59365985, + "step": 2755, + "time_per_iteration": 2.6883299350738525 + }, + { + "auxiliary_loss_clip": 0.01231457, + "auxiliary_loss_mlp": 0.01032769, + "balance_loss_clip": 1.05834877, + "balance_loss_mlp": 1.02282059, + "epoch": 0.3313894066013347, + "flos": 29456230008960.0, + "grad_norm": 2.3354103060298077, + "language_loss": 0.72619671, + "learning_rate": 3.120594872837425e-06, + "loss": 0.74883896, + "num_input_tokens_seen": 59384655, + "step": 2756, + "time_per_iteration": 2.802762508392334 + }, + { + "auxiliary_loss_clip": 0.01104083, + "auxiliary_loss_mlp": 0.01376348, + "balance_loss_clip": 1.02335846, + "balance_loss_mlp": 1.00084877, + "epoch": 0.3315096494919738, + "flos": 61419242280960.0, + "grad_norm": 0.8289771117082071, + "language_loss": 0.62356663, + "learning_rate": 3.1199495722751906e-06, + "loss": 0.64837086, + "num_input_tokens_seen": 59444185, + "step": 2757, + "time_per_iteration": 3.2169015407562256 + }, + { + "auxiliary_loss_clip": 0.01236939, + "auxiliary_loss_mlp": 0.0102727, + "balance_loss_clip": 1.05424452, + "balance_loss_mlp": 1.01793551, + "epoch": 0.33162989238261287, + "flos": 21653057485440.0, + "grad_norm": 2.311195471249982, + "language_loss": 0.83949536, + "learning_rate": 3.1193041018174972e-06, + "loss": 0.86213744, + "num_input_tokens_seen": 59464900, + "step": 2758, + "time_per_iteration": 2.7725863456726074 + }, + { + "auxiliary_loss_clip": 0.01216937, + "auxiliary_loss_mlp": 0.01036481, + "balance_loss_clip": 1.06116104, + "balance_loss_mlp": 1.02714741, + "epoch": 0.331750135273252, + "flos": 22674787850880.0, + "grad_norm": 2.108401127670268, + "language_loss": 0.94456911, + "learning_rate": 3.118658461562261e-06, + "loss": 0.96710324, + "num_input_tokens_seen": 59481000, + "step": 2759, + "time_per_iteration": 2.6332600116729736 + }, + { + "auxiliary_loss_clip": 0.01220155, + "auxiliary_loss_mlp": 0.01035444, + "balance_loss_clip": 1.05973601, + "balance_loss_mlp": 1.02708137, + "epoch": 0.33187037816389103, + "flos": 22746896403840.0, + "grad_norm": 1.400645532172676, + "language_loss": 0.85022271, + "learning_rate": 3.118012651607426e-06, + "loss": 0.87277871, + "num_input_tokens_seen": 59502605, + "step": 2760, + "time_per_iteration": 2.7824482917785645 + }, + { + "auxiliary_loss_clip": 0.01207768, + "auxiliary_loss_mlp": 0.01033143, + "balance_loss_clip": 1.06559038, + "balance_loss_mlp": 1.02427399, + "epoch": 0.33199062105453014, + "flos": 19203769918080.0, + "grad_norm": 2.361765362229825, + "language_loss": 0.83147687, + "learning_rate": 3.1173666720509603e-06, + "loss": 0.85388595, + "num_input_tokens_seen": 59519540, + "step": 2761, + "time_per_iteration": 2.6694679260253906 + }, + { + "auxiliary_loss_clip": 0.01225188, + "auxiliary_loss_mlp": 0.01028612, + "balance_loss_clip": 1.05944109, + "balance_loss_mlp": 1.01937914, + "epoch": 0.33211086394516925, + "flos": 31577006764800.0, + "grad_norm": 2.5053527552443295, + "language_loss": 0.67854238, + "learning_rate": 3.116720522990859e-06, + "loss": 0.70108032, + "num_input_tokens_seen": 59540415, + "step": 2762, + "time_per_iteration": 2.816758155822754 + }, + { + "auxiliary_loss_clip": 0.0124728, + "auxiliary_loss_mlp": 0.01035744, + "balance_loss_clip": 1.05483913, + "balance_loss_mlp": 1.0266664, + "epoch": 0.3322311068358083, + "flos": 17932496791680.0, + "grad_norm": 1.9308647274865252, + "language_loss": 0.62201607, + "learning_rate": 3.116074204525142e-06, + "loss": 0.64484632, + "num_input_tokens_seen": 59558590, + "step": 2763, + "time_per_iteration": 2.882687568664551 + }, + { + "auxiliary_loss_clip": 0.01199357, + "auxiliary_loss_mlp": 0.01029267, + "balance_loss_clip": 1.05803919, + "balance_loss_mlp": 1.02031446, + "epoch": 0.3323513497264474, + "flos": 32269831269120.0, + "grad_norm": 1.495312558894397, + "language_loss": 0.83532625, + "learning_rate": 3.1154277167518553e-06, + "loss": 0.85761249, + "num_input_tokens_seen": 59580205, + "step": 2764, + "time_per_iteration": 2.76686954498291 + }, + { + "auxiliary_loss_clip": 0.01109392, + "auxiliary_loss_mlp": 0.01001134, + "balance_loss_clip": 1.02006376, + "balance_loss_mlp": 0.99929804, + "epoch": 0.33247159261708653, + "flos": 52668674588160.0, + "grad_norm": 0.778637237423169, + "language_loss": 0.59470713, + "learning_rate": 3.114781059769072e-06, + "loss": 0.6158123, + "num_input_tokens_seen": 59631530, + "step": 2765, + "time_per_iteration": 3.0966005325317383 + }, + { + "auxiliary_loss_clip": 0.0121634, + "auxiliary_loss_mlp": 0.01029282, + "balance_loss_clip": 1.05570686, + "balance_loss_mlp": 1.02022755, + "epoch": 0.3325918355077256, + "flos": 27125232906240.0, + "grad_norm": 2.7761478368482826, + "language_loss": 0.67541516, + "learning_rate": 3.1141342336748874e-06, + "loss": 0.69787139, + "num_input_tokens_seen": 59651090, + "step": 2766, + "time_per_iteration": 2.720442056655884 + }, + { + "auxiliary_loss_clip": 0.01210106, + "auxiliary_loss_mlp": 0.01030668, + "balance_loss_clip": 1.06243324, + "balance_loss_mlp": 1.02237701, + "epoch": 0.3327120783983647, + "flos": 23664414435840.0, + "grad_norm": 1.3990912422085604, + "language_loss": 0.82126594, + "learning_rate": 3.1134872385674253e-06, + "loss": 0.84367365, + "num_input_tokens_seen": 59675245, + "step": 2767, + "time_per_iteration": 2.7241177558898926 + }, + { + "auxiliary_loss_clip": 0.01223084, + "auxiliary_loss_mlp": 0.01031457, + "balance_loss_clip": 1.05537605, + "balance_loss_mlp": 1.02233744, + "epoch": 0.3328323212890038, + "flos": 19171378828800.0, + "grad_norm": 1.6700207172272092, + "language_loss": 0.8542527, + "learning_rate": 3.1128400745448353e-06, + "loss": 0.87679809, + "num_input_tokens_seen": 59694625, + "step": 2768, + "time_per_iteration": 3.6058506965637207 + }, + { + "auxiliary_loss_clip": 0.0121672, + "auxiliary_loss_mlp": 0.01038599, + "balance_loss_clip": 1.06244028, + "balance_loss_mlp": 1.02955103, + "epoch": 0.33295256417964286, + "flos": 37706347463040.0, + "grad_norm": 1.934565686087957, + "language_loss": 0.63006365, + "learning_rate": 3.11219274170529e-06, + "loss": 0.65261686, + "num_input_tokens_seen": 59716435, + "step": 2769, + "time_per_iteration": 4.573383808135986 + }, + { + "auxiliary_loss_clip": 0.01211737, + "auxiliary_loss_mlp": 0.01030257, + "balance_loss_clip": 1.05558503, + "balance_loss_mlp": 1.02131593, + "epoch": 0.333072807070282, + "flos": 26505989412480.0, + "grad_norm": 1.8270048779062342, + "language_loss": 0.81627041, + "learning_rate": 3.1115452401469903e-06, + "loss": 0.83869034, + "num_input_tokens_seen": 59736835, + "step": 2770, + "time_per_iteration": 2.7712576389312744 + }, + { + "auxiliary_loss_clip": 0.01223121, + "auxiliary_loss_mlp": 0.01032331, + "balance_loss_clip": 1.05092239, + "balance_loss_mlp": 1.02378941, + "epoch": 0.3331930499609211, + "flos": 21430913823360.0, + "grad_norm": 1.881154673124408, + "language_loss": 0.86770535, + "learning_rate": 3.1108975699681613e-06, + "loss": 0.89025986, + "num_input_tokens_seen": 59754230, + "step": 2771, + "time_per_iteration": 2.733732223510742 + }, + { + "auxiliary_loss_clip": 0.01227655, + "auxiliary_loss_mlp": 0.01033999, + "balance_loss_clip": 1.05744672, + "balance_loss_mlp": 1.02563095, + "epoch": 0.33331329285156014, + "flos": 20659947281280.0, + "grad_norm": 1.8044940053708904, + "language_loss": 0.71747822, + "learning_rate": 3.1102497312670542e-06, + "loss": 0.74009478, + "num_input_tokens_seen": 59772235, + "step": 2772, + "time_per_iteration": 2.7585973739624023 + }, + { + "auxiliary_loss_clip": 0.01206541, + "auxiliary_loss_mlp": 0.01034033, + "balance_loss_clip": 1.05591977, + "balance_loss_mlp": 1.02463937, + "epoch": 0.33343353574219925, + "flos": 28001596930560.0, + "grad_norm": 2.95375546331026, + "language_loss": 0.80231202, + "learning_rate": 3.109601724141946e-06, + "loss": 0.82471782, + "num_input_tokens_seen": 59791230, + "step": 2773, + "time_per_iteration": 2.8011386394500732 + }, + { + "auxiliary_loss_clip": 0.0121308, + "auxiliary_loss_mlp": 0.01029445, + "balance_loss_clip": 1.05714345, + "balance_loss_mlp": 1.02024806, + "epoch": 0.33355377863283836, + "flos": 23764963582080.0, + "grad_norm": 1.9952621628255602, + "language_loss": 0.68166262, + "learning_rate": 3.108953548691138e-06, + "loss": 0.70408785, + "num_input_tokens_seen": 59811315, + "step": 2774, + "time_per_iteration": 2.863060474395752 + }, + { + "auxiliary_loss_clip": 0.01201249, + "auxiliary_loss_mlp": 0.01032058, + "balance_loss_clip": 1.05900764, + "balance_loss_mlp": 1.02302206, + "epoch": 0.3336740215234774, + "flos": 37779677078400.0, + "grad_norm": 4.137056765663017, + "language_loss": 0.73032236, + "learning_rate": 3.108305205012959e-06, + "loss": 0.75265539, + "num_input_tokens_seen": 59832010, + "step": 2775, + "time_per_iteration": 3.675976514816284 + }, + { + "auxiliary_loss_clip": 0.01215771, + "auxiliary_loss_mlp": 0.01036157, + "balance_loss_clip": 1.05593419, + "balance_loss_mlp": 1.02698946, + "epoch": 0.3337942644141165, + "flos": 25519056347520.0, + "grad_norm": 2.32274215721062, + "language_loss": 0.86995292, + "learning_rate": 3.107656693205761e-06, + "loss": 0.89247227, + "num_input_tokens_seen": 59851450, + "step": 2776, + "time_per_iteration": 2.7344484329223633 + }, + { + "auxiliary_loss_clip": 0.01207309, + "auxiliary_loss_mlp": 0.01033575, + "balance_loss_clip": 1.06052065, + "balance_loss_mlp": 1.02251863, + "epoch": 0.3339145073047556, + "flos": 25989844930560.0, + "grad_norm": 2.605519257272306, + "language_loss": 0.70296341, + "learning_rate": 3.107008013367924e-06, + "loss": 0.72537225, + "num_input_tokens_seen": 59870245, + "step": 2777, + "time_per_iteration": 2.668959856033325 + }, + { + "auxiliary_loss_clip": 0.01221144, + "auxiliary_loss_mlp": 0.0103174, + "balance_loss_clip": 1.05430365, + "balance_loss_mlp": 1.0224359, + "epoch": 0.3340347501953947, + "flos": 19062569554560.0, + "grad_norm": 2.0701621565051935, + "language_loss": 0.86836338, + "learning_rate": 3.1063591655978507e-06, + "loss": 0.89089227, + "num_input_tokens_seen": 59886195, + "step": 2778, + "time_per_iteration": 2.694918155670166 + }, + { + "auxiliary_loss_clip": 0.01216767, + "auxiliary_loss_mlp": 0.0103491, + "balance_loss_clip": 1.04906726, + "balance_loss_mlp": 1.02594519, + "epoch": 0.3341549930860338, + "flos": 18109715518080.0, + "grad_norm": 1.858098693186922, + "language_loss": 0.79823399, + "learning_rate": 3.105710149993972e-06, + "loss": 0.82075077, + "num_input_tokens_seen": 59905525, + "step": 2779, + "time_per_iteration": 2.7474021911621094 + }, + { + "auxiliary_loss_clip": 0.01203398, + "auxiliary_loss_mlp": 0.01035096, + "balance_loss_clip": 1.06004953, + "balance_loss_mlp": 1.02560067, + "epoch": 0.33427523597667286, + "flos": 22674967418880.0, + "grad_norm": 1.8655612102643933, + "language_loss": 0.85433686, + "learning_rate": 3.1050609666547427e-06, + "loss": 0.8767218, + "num_input_tokens_seen": 59925085, + "step": 2780, + "time_per_iteration": 2.611330986022949 + }, + { + "auxiliary_loss_clip": 0.01232937, + "auxiliary_loss_mlp": 0.01033666, + "balance_loss_clip": 1.05645108, + "balance_loss_mlp": 1.02504086, + "epoch": 0.33439547886731197, + "flos": 22638338524800.0, + "grad_norm": 3.541162887224132, + "language_loss": 0.77037084, + "learning_rate": 3.104411615678644e-06, + "loss": 0.79303682, + "num_input_tokens_seen": 59943935, + "step": 2781, + "time_per_iteration": 2.7429678440093994 + }, + { + "auxiliary_loss_clip": 0.0121178, + "auxiliary_loss_mlp": 0.01032167, + "balance_loss_clip": 1.05628395, + "balance_loss_mlp": 1.0221653, + "epoch": 0.3345157217579511, + "flos": 24096383395200.0, + "grad_norm": 2.1644551943263495, + "language_loss": 0.74039853, + "learning_rate": 3.1037620971641803e-06, + "loss": 0.76283801, + "num_input_tokens_seen": 59963725, + "step": 2782, + "time_per_iteration": 2.6945552825927734 + }, + { + "auxiliary_loss_clip": 0.01202241, + "auxiliary_loss_mlp": 0.01032173, + "balance_loss_clip": 1.06002975, + "balance_loss_mlp": 1.02271366, + "epoch": 0.33463596464859013, + "flos": 18989491334400.0, + "grad_norm": 2.9168633350436175, + "language_loss": 0.64886498, + "learning_rate": 3.1031124112098844e-06, + "loss": 0.6712091, + "num_input_tokens_seen": 59981935, + "step": 2783, + "time_per_iteration": 2.568804979324341 + }, + { + "auxiliary_loss_clip": 0.01221199, + "auxiliary_loss_mlp": 0.01036754, + "balance_loss_clip": 1.05831552, + "balance_loss_mlp": 1.02796841, + "epoch": 0.33475620753922924, + "flos": 20375607219840.0, + "grad_norm": 1.8348267015004485, + "language_loss": 0.72466803, + "learning_rate": 3.1024625579143127e-06, + "loss": 0.74724758, + "num_input_tokens_seen": 59999455, + "step": 2784, + "time_per_iteration": 2.7728681564331055 + }, + { + "auxiliary_loss_clip": 0.01203376, + "auxiliary_loss_mlp": 0.01031763, + "balance_loss_clip": 1.06121862, + "balance_loss_mlp": 1.02216101, + "epoch": 0.33487645042986836, + "flos": 18182578256640.0, + "grad_norm": 2.133140038379456, + "language_loss": 0.7276212, + "learning_rate": 3.101812537376048e-06, + "loss": 0.74997258, + "num_input_tokens_seen": 60018475, + "step": 2785, + "time_per_iteration": 2.6171674728393555 + }, + { + "auxiliary_loss_clip": 0.01207942, + "auxiliary_loss_mlp": 0.01380471, + "balance_loss_clip": 1.05384207, + "balance_loss_mlp": 1.00069916, + "epoch": 0.3349966933205074, + "flos": 25848824135040.0, + "grad_norm": 2.3314870905669878, + "language_loss": 0.83972025, + "learning_rate": 3.1011623496936973e-06, + "loss": 0.86560434, + "num_input_tokens_seen": 60036770, + "step": 2786, + "time_per_iteration": 2.73211669921875 + }, + { + "auxiliary_loss_clip": 0.01203039, + "auxiliary_loss_mlp": 0.01028904, + "balance_loss_clip": 1.06223452, + "balance_loss_mlp": 1.01926613, + "epoch": 0.3351169362111465, + "flos": 28111447699200.0, + "grad_norm": 1.7461764949704206, + "language_loss": 0.69191635, + "learning_rate": 3.100511994965893e-06, + "loss": 0.71423578, + "num_input_tokens_seen": 60056725, + "step": 2787, + "time_per_iteration": 2.698880195617676 + }, + { + "auxiliary_loss_clip": 0.01202877, + "auxiliary_loss_mlp": 0.01031134, + "balance_loss_clip": 1.05798829, + "balance_loss_mlp": 1.0217998, + "epoch": 0.33523717910178563, + "flos": 22673315393280.0, + "grad_norm": 1.5907998294783912, + "language_loss": 0.84338242, + "learning_rate": 3.0998614732912947e-06, + "loss": 0.86572254, + "num_input_tokens_seen": 60076100, + "step": 2788, + "time_per_iteration": 2.702061891555786 + }, + { + "auxiliary_loss_clip": 0.01210427, + "auxiliary_loss_mlp": 0.01031536, + "balance_loss_clip": 1.06131768, + "balance_loss_mlp": 1.02230906, + "epoch": 0.3353574219924247, + "flos": 15669801400320.0, + "grad_norm": 1.9064195563413286, + "language_loss": 0.67964125, + "learning_rate": 3.0992107847685855e-06, + "loss": 0.70206088, + "num_input_tokens_seen": 60093815, + "step": 2789, + "time_per_iteration": 2.6655378341674805 + }, + { + "auxiliary_loss_clip": 0.01219295, + "auxiliary_loss_mlp": 0.01037076, + "balance_loss_clip": 1.05986941, + "balance_loss_mlp": 1.02780199, + "epoch": 0.3354776648830638, + "flos": 24790644443520.0, + "grad_norm": 1.7122878906266439, + "language_loss": 0.78946871, + "learning_rate": 3.0985599294964736e-06, + "loss": 0.81203246, + "num_input_tokens_seen": 60113370, + "step": 2790, + "time_per_iteration": 2.675440549850464 + }, + { + "auxiliary_loss_clip": 0.01238172, + "auxiliary_loss_mlp": 0.01039445, + "balance_loss_clip": 1.05633998, + "balance_loss_mlp": 1.03015852, + "epoch": 0.33559790777370285, + "flos": 28694852398080.0, + "grad_norm": 2.209820567136946, + "language_loss": 0.69714898, + "learning_rate": 3.097908907573695e-06, + "loss": 0.71992517, + "num_input_tokens_seen": 60131350, + "step": 2791, + "time_per_iteration": 2.7715747356414795 + }, + { + "auxiliary_loss_clip": 0.01247678, + "auxiliary_loss_mlp": 0.01030508, + "balance_loss_clip": 1.05800486, + "balance_loss_mlp": 1.02190673, + "epoch": 0.33571815066434196, + "flos": 22235779825920.0, + "grad_norm": 1.9575279993342132, + "language_loss": 0.8921411, + "learning_rate": 3.0972577190990067e-06, + "loss": 0.91492295, + "num_input_tokens_seen": 60149830, + "step": 2792, + "time_per_iteration": 2.7597644329071045 + }, + { + "auxiliary_loss_clip": 0.01232778, + "auxiliary_loss_mlp": 0.01029147, + "balance_loss_clip": 1.05691028, + "balance_loss_mlp": 1.02011108, + "epoch": 0.3358383935549811, + "flos": 23842279607040.0, + "grad_norm": 1.8111626304631634, + "language_loss": 0.80092716, + "learning_rate": 3.096606364171196e-06, + "loss": 0.82354641, + "num_input_tokens_seen": 60169620, + "step": 2793, + "time_per_iteration": 2.7237801551818848 + }, + { + "auxiliary_loss_clip": 0.01207864, + "auxiliary_loss_mlp": 0.01033902, + "balance_loss_clip": 1.05129647, + "balance_loss_mlp": 1.02462125, + "epoch": 0.33595863644562013, + "flos": 22267308988800.0, + "grad_norm": 1.9776982169084258, + "language_loss": 0.85288256, + "learning_rate": 3.0959548428890703e-06, + "loss": 0.87530029, + "num_input_tokens_seen": 60188490, + "step": 2794, + "time_per_iteration": 3.545255184173584 + }, + { + "auxiliary_loss_clip": 0.01207046, + "auxiliary_loss_mlp": 0.01033896, + "balance_loss_clip": 1.06011677, + "balance_loss_mlp": 1.02519345, + "epoch": 0.33607887933625924, + "flos": 20119779578880.0, + "grad_norm": 4.746149305439764, + "language_loss": 0.83999944, + "learning_rate": 3.095303155351468e-06, + "loss": 0.86240888, + "num_input_tokens_seen": 60208695, + "step": 2795, + "time_per_iteration": 4.5328452587127686 + }, + { + "auxiliary_loss_clip": 0.01225318, + "auxiliary_loss_mlp": 0.01034466, + "balance_loss_clip": 1.05266261, + "balance_loss_mlp": 1.02568603, + "epoch": 0.33619912222689835, + "flos": 19318109886720.0, + "grad_norm": 2.2906606860892365, + "language_loss": 0.79546368, + "learning_rate": 3.0946513016572464e-06, + "loss": 0.81806153, + "num_input_tokens_seen": 60227600, + "step": 2796, + "time_per_iteration": 2.8185696601867676 + }, + { + "auxiliary_loss_clip": 0.0121194, + "auxiliary_loss_mlp": 0.01033794, + "balance_loss_clip": 1.05859566, + "balance_loss_mlp": 1.02403069, + "epoch": 0.3363193651175374, + "flos": 16800664262400.0, + "grad_norm": 1.9794457110188641, + "language_loss": 0.76986492, + "learning_rate": 3.0939992819052938e-06, + "loss": 0.79232228, + "num_input_tokens_seen": 60245110, + "step": 2797, + "time_per_iteration": 2.607442617416382 + }, + { + "auxiliary_loss_clip": 0.01220677, + "auxiliary_loss_mlp": 0.01031836, + "balance_loss_clip": 1.05967426, + "balance_loss_mlp": 1.02291942, + "epoch": 0.3364396080081765, + "flos": 23550289948800.0, + "grad_norm": 2.3206832236340995, + "language_loss": 0.81104052, + "learning_rate": 3.0933470961945193e-06, + "loss": 0.83356565, + "num_input_tokens_seen": 60263405, + "step": 2798, + "time_per_iteration": 2.752535343170166 + }, + { + "auxiliary_loss_clip": 0.01214663, + "auxiliary_loss_mlp": 0.01033824, + "balance_loss_clip": 1.05852175, + "balance_loss_mlp": 1.02494931, + "epoch": 0.3365598508988156, + "flos": 28037902602240.0, + "grad_norm": 2.691320746363802, + "language_loss": 0.68532741, + "learning_rate": 3.0926947446238597e-06, + "loss": 0.70781225, + "num_input_tokens_seen": 60282975, + "step": 2799, + "time_per_iteration": 2.7561769485473633 + }, + { + "auxiliary_loss_clip": 0.01210763, + "auxiliary_loss_mlp": 0.01028961, + "balance_loss_clip": 1.05489647, + "balance_loss_mlp": 1.0191443, + "epoch": 0.3366800937894547, + "flos": 16982767238400.0, + "grad_norm": 48.931702753660886, + "language_loss": 0.82666779, + "learning_rate": 3.092042227292276e-06, + "loss": 0.84906501, + "num_input_tokens_seen": 60299810, + "step": 2800, + "time_per_iteration": 2.5850372314453125 + }, + { + "auxiliary_loss_clip": 0.01198259, + "auxiliary_loss_mlp": 0.01030802, + "balance_loss_clip": 1.06024384, + "balance_loss_mlp": 1.02196836, + "epoch": 0.3368003366800938, + "flos": 23915321913600.0, + "grad_norm": 1.5727591178346845, + "language_loss": 0.88006777, + "learning_rate": 3.0913895442987557e-06, + "loss": 0.90235841, + "num_input_tokens_seen": 60320775, + "step": 2801, + "time_per_iteration": 2.6872053146362305 + }, + { + "auxiliary_loss_clip": 0.0122424, + "auxiliary_loss_mlp": 0.01380401, + "balance_loss_clip": 1.05677032, + "balance_loss_mlp": 1.00065279, + "epoch": 0.3369205795707329, + "flos": 24791219061120.0, + "grad_norm": 1.6392985531523931, + "language_loss": 0.85775805, + "learning_rate": 3.090736695742308e-06, + "loss": 0.88380444, + "num_input_tokens_seen": 60341905, + "step": 2802, + "time_per_iteration": 3.6395318508148193 + }, + { + "auxiliary_loss_clip": 0.01225469, + "auxiliary_loss_mlp": 0.01028215, + "balance_loss_clip": 1.05207849, + "balance_loss_mlp": 1.01899433, + "epoch": 0.33704082246137196, + "flos": 17931096161280.0, + "grad_norm": 3.484855342770534, + "language_loss": 0.52401495, + "learning_rate": 3.0900836817219713e-06, + "loss": 0.54655176, + "num_input_tokens_seen": 60358335, + "step": 2803, + "time_per_iteration": 2.7782363891601562 + }, + { + "auxiliary_loss_clip": 0.01199375, + "auxiliary_loss_mlp": 0.01026943, + "balance_loss_clip": 1.05915225, + "balance_loss_mlp": 1.01790738, + "epoch": 0.33716106535201107, + "flos": 21286517149440.0, + "grad_norm": 1.7211173084147433, + "language_loss": 0.83430552, + "learning_rate": 3.089430502336807e-06, + "loss": 0.85656875, + "num_input_tokens_seen": 60378305, + "step": 2804, + "time_per_iteration": 2.681867837905884 + }, + { + "auxiliary_loss_clip": 0.012116, + "auxiliary_loss_mlp": 0.0102556, + "balance_loss_clip": 1.05854285, + "balance_loss_mlp": 1.01591575, + "epoch": 0.3372813082426502, + "flos": 18402962152320.0, + "grad_norm": 3.9078676201689055, + "language_loss": 0.90124607, + "learning_rate": 3.088777157685902e-06, + "loss": 0.9236176, + "num_input_tokens_seen": 60393895, + "step": 2805, + "time_per_iteration": 2.602215051651001 + }, + { + "auxiliary_loss_clip": 0.01215497, + "auxiliary_loss_mlp": 0.0102851, + "balance_loss_clip": 1.05859387, + "balance_loss_mlp": 1.01976585, + "epoch": 0.33740155113328923, + "flos": 17201391367680.0, + "grad_norm": 2.070825550671347, + "language_loss": 0.86204147, + "learning_rate": 3.088123647868367e-06, + "loss": 0.88448155, + "num_input_tokens_seen": 60410445, + "step": 2806, + "time_per_iteration": 2.639455795288086 + }, + { + "auxiliary_loss_clip": 0.01208996, + "auxiliary_loss_mlp": 0.01031205, + "balance_loss_clip": 1.05708337, + "balance_loss_mlp": 1.02187085, + "epoch": 0.33752179402392835, + "flos": 29058950609280.0, + "grad_norm": 2.0367245607130946, + "language_loss": 0.81393319, + "learning_rate": 3.0874699729833405e-06, + "loss": 0.83633518, + "num_input_tokens_seen": 60431815, + "step": 2807, + "time_per_iteration": 2.750744342803955 + }, + { + "auxiliary_loss_clip": 0.01208865, + "auxiliary_loss_mlp": 0.01030583, + "balance_loss_clip": 1.05414641, + "balance_loss_mlp": 1.02198195, + "epoch": 0.3376420369145674, + "flos": 25080730680960.0, + "grad_norm": 1.811048346377354, + "language_loss": 0.7943396, + "learning_rate": 3.086816133129983e-06, + "loss": 0.81673408, + "num_input_tokens_seen": 60452075, + "step": 2808, + "time_per_iteration": 2.6557304859161377 + }, + { + "auxiliary_loss_clip": 0.01204692, + "auxiliary_loss_mlp": 0.01031894, + "balance_loss_clip": 1.06408119, + "balance_loss_mlp": 1.02338231, + "epoch": 0.3377622798052065, + "flos": 27490624007040.0, + "grad_norm": 2.271647973329316, + "language_loss": 0.76065695, + "learning_rate": 3.0861621284074826e-06, + "loss": 0.78302276, + "num_input_tokens_seen": 60472600, + "step": 2809, + "time_per_iteration": 2.8796279430389404 + }, + { + "auxiliary_loss_clip": 0.01225113, + "auxiliary_loss_mlp": 0.01034263, + "balance_loss_clip": 1.05985391, + "balance_loss_mlp": 1.02616286, + "epoch": 0.3378825226958456, + "flos": 21975211589760.0, + "grad_norm": 1.4864699606618277, + "language_loss": 0.73207468, + "learning_rate": 3.085507958915051e-06, + "loss": 0.75466847, + "num_input_tokens_seen": 60491030, + "step": 2810, + "time_per_iteration": 2.7229909896850586 + }, + { + "auxiliary_loss_clip": 0.01214296, + "auxiliary_loss_mlp": 0.01030035, + "balance_loss_clip": 1.05669308, + "balance_loss_mlp": 1.02065933, + "epoch": 0.3380027655864847, + "flos": 42523189200000.0, + "grad_norm": 1.8101164606882134, + "language_loss": 0.71384555, + "learning_rate": 3.084853624751925e-06, + "loss": 0.73628891, + "num_input_tokens_seen": 60512615, + "step": 2811, + "time_per_iteration": 2.8370797634124756 + }, + { + "auxiliary_loss_clip": 0.01227444, + "auxiliary_loss_mlp": 0.01027574, + "balance_loss_clip": 1.05681074, + "balance_loss_mlp": 1.01848459, + "epoch": 0.3381230084771238, + "flos": 26725080418560.0, + "grad_norm": 1.9097403504272203, + "language_loss": 0.85903299, + "learning_rate": 3.0841991260173668e-06, + "loss": 0.88158315, + "num_input_tokens_seen": 60532520, + "step": 2812, + "time_per_iteration": 2.7847611904144287 + }, + { + "auxiliary_loss_clip": 0.01201769, + "auxiliary_loss_mlp": 0.01033058, + "balance_loss_clip": 1.06086922, + "balance_loss_mlp": 1.02433228, + "epoch": 0.3382432513677629, + "flos": 22710375250560.0, + "grad_norm": 1.837595465642499, + "language_loss": 0.80556977, + "learning_rate": 3.0835444628106634e-06, + "loss": 0.82791805, + "num_input_tokens_seen": 60551500, + "step": 2813, + "time_per_iteration": 2.553905725479126 + }, + { + "auxiliary_loss_clip": 0.01202343, + "auxiliary_loss_mlp": 0.01380825, + "balance_loss_clip": 1.06213009, + "balance_loss_mlp": 1.00089383, + "epoch": 0.33836349425840195, + "flos": 22122409524480.0, + "grad_norm": 4.238202742579538, + "language_loss": 0.82972944, + "learning_rate": 3.082889635231126e-06, + "loss": 0.85556108, + "num_input_tokens_seen": 60570160, + "step": 2814, + "time_per_iteration": 2.646627902984619 + }, + { + "auxiliary_loss_clip": 0.01217765, + "auxiliary_loss_mlp": 0.01035571, + "balance_loss_clip": 1.05587029, + "balance_loss_mlp": 1.02574778, + "epoch": 0.33848373714904106, + "flos": 27308090067840.0, + "grad_norm": 2.851295110765566, + "language_loss": 0.76280499, + "learning_rate": 3.0822346433780925e-06, + "loss": 0.78533828, + "num_input_tokens_seen": 60590885, + "step": 2815, + "time_per_iteration": 2.6815123558044434 + }, + { + "auxiliary_loss_clip": 0.01208987, + "auxiliary_loss_mlp": 0.01032128, + "balance_loss_clip": 1.05648267, + "balance_loss_mlp": 1.02306771, + "epoch": 0.3386039800396802, + "flos": 25848716394240.0, + "grad_norm": 2.0198981787270047, + "language_loss": 0.86992854, + "learning_rate": 3.0815794873509237e-06, + "loss": 0.89233971, + "num_input_tokens_seen": 60609170, + "step": 2816, + "time_per_iteration": 2.703941822052002 + }, + { + "auxiliary_loss_clip": 0.01201289, + "auxiliary_loss_mlp": 0.01029741, + "balance_loss_clip": 1.05914497, + "balance_loss_mlp": 1.02089536, + "epoch": 0.33872422293031923, + "flos": 18880646146560.0, + "grad_norm": 1.9399938008717854, + "language_loss": 0.72756124, + "learning_rate": 3.0809241672490066e-06, + "loss": 0.74987149, + "num_input_tokens_seen": 60627340, + "step": 2817, + "time_per_iteration": 2.616055965423584 + }, + { + "auxiliary_loss_clip": 0.01219324, + "auxiliary_loss_mlp": 0.01034215, + "balance_loss_clip": 1.0575397, + "balance_loss_mlp": 1.02575707, + "epoch": 0.33884446582095834, + "flos": 23146977064320.0, + "grad_norm": 2.000196588894164, + "language_loss": 0.84864384, + "learning_rate": 3.080268683171753e-06, + "loss": 0.87117922, + "num_input_tokens_seen": 60647630, + "step": 2818, + "time_per_iteration": 2.747812032699585 + }, + { + "auxiliary_loss_clip": 0.01208606, + "auxiliary_loss_mlp": 0.0102837, + "balance_loss_clip": 1.05678737, + "balance_loss_mlp": 1.01994848, + "epoch": 0.33896470871159745, + "flos": 15997342544640.0, + "grad_norm": 2.512946808371963, + "language_loss": 0.89321309, + "learning_rate": 3.0796130352185985e-06, + "loss": 0.91558284, + "num_input_tokens_seen": 60664485, + "step": 2819, + "time_per_iteration": 2.6303858757019043 + }, + { + "auxiliary_loss_clip": 0.01200616, + "auxiliary_loss_mlp": 0.01381241, + "balance_loss_clip": 1.05195761, + "balance_loss_mlp": 1.00109792, + "epoch": 0.3390849516022365, + "flos": 34495754112000.0, + "grad_norm": 1.9387690601918306, + "language_loss": 0.66428185, + "learning_rate": 3.0789572234890057e-06, + "loss": 0.69010043, + "num_input_tokens_seen": 60686125, + "step": 2820, + "time_per_iteration": 3.604227066040039 + }, + { + "auxiliary_loss_clip": 0.01220263, + "auxiliary_loss_mlp": 0.01030505, + "balance_loss_clip": 1.06028461, + "balance_loss_mlp": 1.02152824, + "epoch": 0.3392051944928756, + "flos": 16180307447040.0, + "grad_norm": 1.6132361571955254, + "language_loss": 0.77379084, + "learning_rate": 3.0783012480824596e-06, + "loss": 0.79629844, + "num_input_tokens_seen": 60705270, + "step": 2821, + "time_per_iteration": 4.534826040267944 + }, + { + "auxiliary_loss_clip": 0.01201301, + "auxiliary_loss_mlp": 0.01035692, + "balance_loss_clip": 1.06017089, + "balance_loss_mlp": 1.02587509, + "epoch": 0.33932543738351467, + "flos": 17086656349440.0, + "grad_norm": 2.104071620219608, + "language_loss": 0.74019271, + "learning_rate": 3.077645109098471e-06, + "loss": 0.76256263, + "num_input_tokens_seen": 60721540, + "step": 2822, + "time_per_iteration": 2.6257359981536865 + }, + { + "auxiliary_loss_clip": 0.01208407, + "auxiliary_loss_mlp": 0.01030688, + "balance_loss_clip": 1.05207324, + "balance_loss_mlp": 1.02148545, + "epoch": 0.3394456802741538, + "flos": 22126970551680.0, + "grad_norm": 1.8624399794336717, + "language_loss": 0.71979308, + "learning_rate": 3.076988806636577e-06, + "loss": 0.74218404, + "num_input_tokens_seen": 60739300, + "step": 2823, + "time_per_iteration": 2.6629438400268555 + }, + { + "auxiliary_loss_clip": 0.01221815, + "auxiliary_loss_mlp": 0.01381169, + "balance_loss_clip": 1.0579412, + "balance_loss_mlp": 1.00093055, + "epoch": 0.3395659231647929, + "flos": 25226887121280.0, + "grad_norm": 1.833290166544764, + "language_loss": 0.8872667, + "learning_rate": 3.0763323407963377e-06, + "loss": 0.91329658, + "num_input_tokens_seen": 60758910, + "step": 2824, + "time_per_iteration": 2.729215383529663 + }, + { + "auxiliary_loss_clip": 0.01206283, + "auxiliary_loss_mlp": 0.01033231, + "balance_loss_clip": 1.05552125, + "balance_loss_mlp": 1.02395678, + "epoch": 0.33968616605543195, + "flos": 29096477343360.0, + "grad_norm": 1.8375428602546633, + "language_loss": 0.79838306, + "learning_rate": 3.075675711677337e-06, + "loss": 0.82077819, + "num_input_tokens_seen": 60779005, + "step": 2825, + "time_per_iteration": 2.7154741287231445 + }, + { + "auxiliary_loss_clip": 0.01210392, + "auxiliary_loss_mlp": 0.0104387, + "balance_loss_clip": 1.05695176, + "balance_loss_mlp": 1.03461361, + "epoch": 0.33980640894607106, + "flos": 21433966479360.0, + "grad_norm": 1.838244282882707, + "language_loss": 0.78108639, + "learning_rate": 3.0750189193791865e-06, + "loss": 0.80362904, + "num_input_tokens_seen": 60798590, + "step": 2826, + "time_per_iteration": 2.683284282684326 + }, + { + "auxiliary_loss_clip": 0.01202851, + "auxiliary_loss_mlp": 0.01030207, + "balance_loss_clip": 1.0558598, + "balance_loss_mlp": 1.02121305, + "epoch": 0.33992665183671017, + "flos": 32490035596800.0, + "grad_norm": 1.9585331449922287, + "language_loss": 0.70226926, + "learning_rate": 3.0743619640015203e-06, + "loss": 0.72459984, + "num_input_tokens_seen": 60818840, + "step": 2827, + "time_per_iteration": 3.6357150077819824 + }, + { + "auxiliary_loss_clip": 0.01220291, + "auxiliary_loss_mlp": 0.01035487, + "balance_loss_clip": 1.05435991, + "balance_loss_mlp": 1.02642155, + "epoch": 0.3400468947273492, + "flos": 17055414495360.0, + "grad_norm": 1.9679812316887315, + "language_loss": 0.92604184, + "learning_rate": 3.073704845643999e-06, + "loss": 0.94859958, + "num_input_tokens_seen": 60835965, + "step": 2828, + "time_per_iteration": 2.6581544876098633 + }, + { + "auxiliary_loss_clip": 0.01212152, + "auxiliary_loss_mlp": 0.01032738, + "balance_loss_clip": 1.05784547, + "balance_loss_mlp": 1.02372599, + "epoch": 0.34016713761798834, + "flos": 16872988296960.0, + "grad_norm": 3.9111917392764344, + "language_loss": 0.77536952, + "learning_rate": 3.0730475644063063e-06, + "loss": 0.79781842, + "num_input_tokens_seen": 60851065, + "step": 2829, + "time_per_iteration": 2.696131944656372 + }, + { + "auxiliary_loss_clip": 0.01206839, + "auxiliary_loss_mlp": 0.01380447, + "balance_loss_clip": 1.05390835, + "balance_loss_mlp": 1.00087452, + "epoch": 0.34028738050862745, + "flos": 21907161273600.0, + "grad_norm": 1.6953724560271861, + "language_loss": 0.64803195, + "learning_rate": 3.072390120388151e-06, + "loss": 0.6739049, + "num_input_tokens_seen": 60869390, + "step": 2830, + "time_per_iteration": 2.7391886711120605 + }, + { + "auxiliary_loss_clip": 0.01210109, + "auxiliary_loss_mlp": 0.01035262, + "balance_loss_clip": 1.05902576, + "balance_loss_mlp": 1.02572536, + "epoch": 0.3404076233992665, + "flos": 22746034477440.0, + "grad_norm": 3.0760360607272164, + "language_loss": 0.71209854, + "learning_rate": 3.071732513689267e-06, + "loss": 0.73455226, + "num_input_tokens_seen": 60887925, + "step": 2831, + "time_per_iteration": 2.6834840774536133 + }, + { + "auxiliary_loss_clip": 0.0121301, + "auxiliary_loss_mlp": 0.01034762, + "balance_loss_clip": 1.0616107, + "balance_loss_mlp": 1.0255115, + "epoch": 0.3405278662899056, + "flos": 17052361839360.0, + "grad_norm": 2.5240864276044253, + "language_loss": 0.6764552, + "learning_rate": 3.0710747444094134e-06, + "loss": 0.69893289, + "num_input_tokens_seen": 60905955, + "step": 2832, + "time_per_iteration": 2.652759313583374 + }, + { + "auxiliary_loss_clip": 0.01223958, + "auxiliary_loss_mlp": 0.01033393, + "balance_loss_clip": 1.06164765, + "balance_loss_mlp": 1.02366567, + "epoch": 0.3406481091805447, + "flos": 42813131783040.0, + "grad_norm": 2.1810178544433803, + "language_loss": 0.64956105, + "learning_rate": 3.070416812648372e-06, + "loss": 0.67213458, + "num_input_tokens_seen": 60929405, + "step": 2833, + "time_per_iteration": 2.9355831146240234 + }, + { + "auxiliary_loss_clip": 0.01219005, + "auxiliary_loss_mlp": 0.01031797, + "balance_loss_clip": 1.05090642, + "balance_loss_mlp": 1.02268958, + "epoch": 0.3407683520711838, + "flos": 26761457917440.0, + "grad_norm": 2.4020472039288316, + "language_loss": 0.6530025, + "learning_rate": 3.069758718505951e-06, + "loss": 0.67551053, + "num_input_tokens_seen": 60951145, + "step": 2834, + "time_per_iteration": 2.7768821716308594 + }, + { + "auxiliary_loss_clip": 0.01204083, + "auxiliary_loss_mlp": 0.01035355, + "balance_loss_clip": 1.06327176, + "balance_loss_mlp": 1.02606845, + "epoch": 0.3408885949618229, + "flos": 28767643309440.0, + "grad_norm": 1.7922696844696775, + "language_loss": 0.80142331, + "learning_rate": 3.0691004620819836e-06, + "loss": 0.82381773, + "num_input_tokens_seen": 60971275, + "step": 2835, + "time_per_iteration": 2.6620898246765137 + }, + { + "auxiliary_loss_clip": 0.01137835, + "auxiliary_loss_mlp": 0.01006489, + "balance_loss_clip": 1.02324092, + "balance_loss_mlp": 1.00478423, + "epoch": 0.341008837852462, + "flos": 63576252881280.0, + "grad_norm": 0.8151842082874787, + "language_loss": 0.60215521, + "learning_rate": 3.0684420434763254e-06, + "loss": 0.62359846, + "num_input_tokens_seen": 61037460, + "step": 2836, + "time_per_iteration": 3.3493101596832275 + }, + { + "auxiliary_loss_clip": 0.01216134, + "auxiliary_loss_mlp": 0.01034415, + "balance_loss_clip": 1.05562496, + "balance_loss_mlp": 1.02528369, + "epoch": 0.34112908074310105, + "flos": 20812173120000.0, + "grad_norm": 2.142824033318834, + "language_loss": 0.76836866, + "learning_rate": 3.06778346278886e-06, + "loss": 0.79087418, + "num_input_tokens_seen": 61056295, + "step": 2837, + "time_per_iteration": 2.6648740768432617 + }, + { + "auxiliary_loss_clip": 0.01203449, + "auxiliary_loss_mlp": 0.01033882, + "balance_loss_clip": 1.0633744, + "balance_loss_mlp": 1.02377284, + "epoch": 0.34124932363374016, + "flos": 24976446520320.0, + "grad_norm": 1.8268368997488806, + "language_loss": 0.78928328, + "learning_rate": 3.0671247201194906e-06, + "loss": 0.81165659, + "num_input_tokens_seen": 61078430, + "step": 2838, + "time_per_iteration": 2.684248447418213 + }, + { + "auxiliary_loss_clip": 0.01225887, + "auxiliary_loss_mlp": 0.01036422, + "balance_loss_clip": 1.052912, + "balance_loss_mlp": 1.02634919, + "epoch": 0.3413695665243792, + "flos": 28402970480640.0, + "grad_norm": 1.6875188774600407, + "language_loss": 0.75866777, + "learning_rate": 3.066465815568151e-06, + "loss": 0.78129083, + "num_input_tokens_seen": 61099260, + "step": 2839, + "time_per_iteration": 2.732173442840576 + }, + { + "auxiliary_loss_clip": 0.01208285, + "auxiliary_loss_mlp": 0.01030021, + "balance_loss_clip": 1.0557996, + "balance_loss_mlp": 1.02075839, + "epoch": 0.34148980941501833, + "flos": 25302012416640.0, + "grad_norm": 1.745942391143093, + "language_loss": 0.68889886, + "learning_rate": 3.0658067492347947e-06, + "loss": 0.7112819, + "num_input_tokens_seen": 61121900, + "step": 2840, + "time_per_iteration": 2.77624249458313 + }, + { + "auxiliary_loss_clip": 0.01243778, + "auxiliary_loss_mlp": 0.01032041, + "balance_loss_clip": 1.05018997, + "balance_loss_mlp": 1.02302909, + "epoch": 0.34161005230565744, + "flos": 17530081747200.0, + "grad_norm": 1.9133155060408147, + "language_loss": 0.66843045, + "learning_rate": 3.065147521219402e-06, + "loss": 0.69118869, + "num_input_tokens_seen": 61141155, + "step": 2841, + "time_per_iteration": 3.083209276199341 + }, + { + "auxiliary_loss_clip": 0.01210433, + "auxiliary_loss_mlp": 0.01029929, + "balance_loss_clip": 1.05867255, + "balance_loss_mlp": 1.02048194, + "epoch": 0.3417302951962965, + "flos": 43650101566080.0, + "grad_norm": 1.5264558268149189, + "language_loss": 0.74437737, + "learning_rate": 3.064488131621977e-06, + "loss": 0.76678103, + "num_input_tokens_seen": 61164480, + "step": 2842, + "time_per_iteration": 3.0671515464782715 + }, + { + "auxiliary_loss_clip": 0.01201409, + "auxiliary_loss_mlp": 0.01032302, + "balance_loss_clip": 1.05692041, + "balance_loss_mlp": 1.02358139, + "epoch": 0.3418505380869356, + "flos": 30882207012480.0, + "grad_norm": 1.7337838929783982, + "language_loss": 0.73901349, + "learning_rate": 3.063828580542549e-06, + "loss": 0.76135063, + "num_input_tokens_seen": 61185675, + "step": 2843, + "time_per_iteration": 2.6601874828338623 + }, + { + "auxiliary_loss_clip": 0.01216384, + "auxiliary_loss_mlp": 0.01030559, + "balance_loss_clip": 1.05573773, + "balance_loss_mlp": 1.02193975, + "epoch": 0.3419707809775747, + "flos": 19463871277440.0, + "grad_norm": 1.88712866041291, + "language_loss": 0.73341036, + "learning_rate": 3.0631688680811706e-06, + "loss": 0.75587976, + "num_input_tokens_seen": 61205300, + "step": 2844, + "time_per_iteration": 2.696699619293213 + }, + { + "auxiliary_loss_clip": 0.01202311, + "auxiliary_loss_mlp": 0.01032838, + "balance_loss_clip": 1.05982792, + "balance_loss_mlp": 1.02361131, + "epoch": 0.3420910238682138, + "flos": 28727818104960.0, + "grad_norm": 1.8745053351960692, + "language_loss": 0.75482452, + "learning_rate": 3.062508994337921e-06, + "loss": 0.77717602, + "num_input_tokens_seen": 61224905, + "step": 2845, + "time_per_iteration": 2.692234992980957 + }, + { + "auxiliary_loss_clip": 0.01207525, + "auxiliary_loss_mlp": 0.01031716, + "balance_loss_clip": 1.05519557, + "balance_loss_mlp": 1.02099895, + "epoch": 0.3422112667588529, + "flos": 21397265758080.0, + "grad_norm": 1.957258798666713, + "language_loss": 0.78988087, + "learning_rate": 3.0618489594129013e-06, + "loss": 0.81227332, + "num_input_tokens_seen": 61243045, + "step": 2846, + "time_per_iteration": 3.5836386680603027 + }, + { + "auxiliary_loss_clip": 0.0123113, + "auxiliary_loss_mlp": 0.01031236, + "balance_loss_clip": 1.05609989, + "balance_loss_mlp": 1.02129436, + "epoch": 0.342331509649492, + "flos": 13881450038400.0, + "grad_norm": 2.1230398351697133, + "language_loss": 0.71304512, + "learning_rate": 3.061188763406239e-06, + "loss": 0.73566878, + "num_input_tokens_seen": 61259190, + "step": 2847, + "time_per_iteration": 3.7608678340911865 + }, + { + "auxiliary_loss_clip": 0.01213466, + "auxiliary_loss_mlp": 0.01031227, + "balance_loss_clip": 1.05622363, + "balance_loss_mlp": 1.02177966, + "epoch": 0.34245175254013105, + "flos": 28621450955520.0, + "grad_norm": 2.2245542707566948, + "language_loss": 0.82281649, + "learning_rate": 3.060528406418085e-06, + "loss": 0.84526342, + "num_input_tokens_seen": 61279040, + "step": 2848, + "time_per_iteration": 2.7732222080230713 + }, + { + "auxiliary_loss_clip": 0.01209099, + "auxiliary_loss_mlp": 0.01031188, + "balance_loss_clip": 1.05575979, + "balance_loss_mlp": 1.02245557, + "epoch": 0.34257199543077016, + "flos": 34127058960000.0, + "grad_norm": 1.624495398764279, + "language_loss": 0.6197924, + "learning_rate": 3.0598678885486145e-06, + "loss": 0.64219528, + "num_input_tokens_seen": 61301580, + "step": 2849, + "time_per_iteration": 2.7623627185821533 + }, + { + "auxiliary_loss_clip": 0.01223382, + "auxiliary_loss_mlp": 0.01380427, + "balance_loss_clip": 1.05063176, + "balance_loss_mlp": 1.00055134, + "epoch": 0.34269223832140927, + "flos": 19974018188160.0, + "grad_norm": 1.836551757366374, + "language_loss": 0.74283391, + "learning_rate": 3.0592072098980282e-06, + "loss": 0.76887202, + "num_input_tokens_seen": 61321240, + "step": 2850, + "time_per_iteration": 2.752863883972168 + }, + { + "auxiliary_loss_clip": 0.01213316, + "auxiliary_loss_mlp": 0.01035602, + "balance_loss_clip": 1.05682111, + "balance_loss_mlp": 1.0261786, + "epoch": 0.3428124812120483, + "flos": 27235658292480.0, + "grad_norm": 2.0539687364898427, + "language_loss": 0.73037291, + "learning_rate": 3.0585463705665514e-06, + "loss": 0.7528621, + "num_input_tokens_seen": 61341615, + "step": 2851, + "time_per_iteration": 2.773071765899658 + }, + { + "auxiliary_loss_clip": 0.01222257, + "auxiliary_loss_mlp": 0.01031873, + "balance_loss_clip": 1.05278492, + "balance_loss_mlp": 1.02242529, + "epoch": 0.34293272410268744, + "flos": 24570871079040.0, + "grad_norm": 2.1243950763194634, + "language_loss": 0.70820606, + "learning_rate": 3.0578853706544304e-06, + "loss": 0.73074734, + "num_input_tokens_seen": 61359005, + "step": 2852, + "time_per_iteration": 2.805356502532959 + }, + { + "auxiliary_loss_clip": 0.01225998, + "auxiliary_loss_mlp": 0.01380849, + "balance_loss_clip": 1.05428815, + "balance_loss_mlp": 1.00069332, + "epoch": 0.34305296699332655, + "flos": 21506865131520.0, + "grad_norm": 1.9448493083280518, + "language_loss": 0.6544044, + "learning_rate": 3.0572242102619404e-06, + "loss": 0.68047291, + "num_input_tokens_seen": 61376160, + "step": 2853, + "time_per_iteration": 2.6994411945343018 + }, + { + "auxiliary_loss_clip": 0.01218509, + "auxiliary_loss_mlp": 0.01038366, + "balance_loss_clip": 1.05917764, + "balance_loss_mlp": 1.02992547, + "epoch": 0.3431732098839656, + "flos": 24056665931520.0, + "grad_norm": 1.931192014600705, + "language_loss": 0.8022368, + "learning_rate": 3.0565628894893784e-06, + "loss": 0.8248055, + "num_input_tokens_seen": 61396795, + "step": 2854, + "time_per_iteration": 4.012827634811401 + }, + { + "auxiliary_loss_clip": 0.01199465, + "auxiliary_loss_mlp": 0.01037878, + "balance_loss_clip": 1.05818772, + "balance_loss_mlp": 1.02911592, + "epoch": 0.3432934527746047, + "flos": 16800879744000.0, + "grad_norm": 1.819241844876398, + "language_loss": 0.74878782, + "learning_rate": 3.0559014084370655e-06, + "loss": 0.7711612, + "num_input_tokens_seen": 61415320, + "step": 2855, + "time_per_iteration": 2.672077178955078 + }, + { + "auxiliary_loss_clip": 0.01225111, + "auxiliary_loss_mlp": 0.01033558, + "balance_loss_clip": 1.05745029, + "balance_loss_mlp": 1.02329993, + "epoch": 0.34341369566524377, + "flos": 23439720908160.0, + "grad_norm": 1.7602047627983517, + "language_loss": 0.79069668, + "learning_rate": 3.055239767205349e-06, + "loss": 0.81328332, + "num_input_tokens_seen": 61437070, + "step": 2856, + "time_per_iteration": 2.779927968978882 + }, + { + "auxiliary_loss_clip": 0.01207772, + "auxiliary_loss_mlp": 0.0103696, + "balance_loss_clip": 1.05997062, + "balance_loss_mlp": 1.0278163, + "epoch": 0.3435339385558829, + "flos": 17267466435840.0, + "grad_norm": 1.7983037874974659, + "language_loss": 0.77953815, + "learning_rate": 3.054577965894599e-06, + "loss": 0.8019855, + "num_input_tokens_seen": 61453215, + "step": 2857, + "time_per_iteration": 2.6166257858276367 + }, + { + "auxiliary_loss_clip": 0.01223787, + "auxiliary_loss_mlp": 0.01028871, + "balance_loss_clip": 1.05860782, + "balance_loss_mlp": 1.01978087, + "epoch": 0.343654181446522, + "flos": 22199366413440.0, + "grad_norm": 2.4663100548149988, + "language_loss": 0.70414472, + "learning_rate": 3.0539160046052094e-06, + "loss": 0.72667134, + "num_input_tokens_seen": 61472915, + "step": 2858, + "time_per_iteration": 2.6930108070373535 + }, + { + "auxiliary_loss_clip": 0.01206635, + "auxiliary_loss_mlp": 0.01033317, + "balance_loss_clip": 1.05427456, + "balance_loss_mlp": 1.02304745, + "epoch": 0.34377442433716104, + "flos": 19901801894400.0, + "grad_norm": 5.29501335192862, + "language_loss": 0.70789665, + "learning_rate": 3.0532538834376003e-06, + "loss": 0.73029613, + "num_input_tokens_seen": 61492475, + "step": 2859, + "time_per_iteration": 2.7386057376861572 + }, + { + "auxiliary_loss_clip": 0.01213522, + "auxiliary_loss_mlp": 0.01035216, + "balance_loss_clip": 1.05801272, + "balance_loss_mlp": 1.02584577, + "epoch": 0.34389466722780015, + "flos": 22197678474240.0, + "grad_norm": 1.786786771596825, + "language_loss": 0.78044331, + "learning_rate": 3.0525916024922143e-06, + "loss": 0.80293065, + "num_input_tokens_seen": 61511660, + "step": 2860, + "time_per_iteration": 2.653886556625366 + }, + { + "auxiliary_loss_clip": 0.01213646, + "auxiliary_loss_mlp": 0.01033889, + "balance_loss_clip": 1.05498958, + "balance_loss_mlp": 1.02462626, + "epoch": 0.34401491011843927, + "flos": 18624567110400.0, + "grad_norm": 2.752642306361512, + "language_loss": 0.84162474, + "learning_rate": 3.0519291618695193e-06, + "loss": 0.8641001, + "num_input_tokens_seen": 61529060, + "step": 2861, + "time_per_iteration": 2.72263240814209 + }, + { + "auxiliary_loss_clip": 0.01214133, + "auxiliary_loss_mlp": 0.01031549, + "balance_loss_clip": 1.05017614, + "balance_loss_mlp": 1.02206016, + "epoch": 0.3441351530090783, + "flos": 17858197509120.0, + "grad_norm": 1.645118785336507, + "language_loss": 0.76175451, + "learning_rate": 3.0512665616700065e-06, + "loss": 0.78421134, + "num_input_tokens_seen": 61548125, + "step": 2862, + "time_per_iteration": 2.842089891433716 + }, + { + "auxiliary_loss_clip": 0.01221088, + "auxiliary_loss_mlp": 0.0103749, + "balance_loss_clip": 1.04856861, + "balance_loss_mlp": 1.02816176, + "epoch": 0.34425539589971743, + "flos": 23112754381440.0, + "grad_norm": 1.9700858197231637, + "language_loss": 0.89197117, + "learning_rate": 3.0506038019941933e-06, + "loss": 0.91455698, + "num_input_tokens_seen": 61568135, + "step": 2863, + "time_per_iteration": 2.7425289154052734 + }, + { + "auxiliary_loss_clip": 0.01226238, + "auxiliary_loss_mlp": 0.01037983, + "balance_loss_clip": 1.05616295, + "balance_loss_mlp": 1.02895904, + "epoch": 0.34437563879035654, + "flos": 21907699977600.0, + "grad_norm": 2.554541329613842, + "language_loss": 0.68019581, + "learning_rate": 3.049940882942617e-06, + "loss": 0.70283806, + "num_input_tokens_seen": 61586920, + "step": 2864, + "time_per_iteration": 2.6961886882781982 + }, + { + "auxiliary_loss_clip": 0.01201492, + "auxiliary_loss_mlp": 0.01031788, + "balance_loss_clip": 1.05967402, + "balance_loss_mlp": 1.02237666, + "epoch": 0.3444958816809956, + "flos": 23076915586560.0, + "grad_norm": 2.9521039180006525, + "language_loss": 0.80363721, + "learning_rate": 3.0492778046158448e-06, + "loss": 0.82597005, + "num_input_tokens_seen": 61608340, + "step": 2865, + "time_per_iteration": 2.6128525733947754 + }, + { + "auxiliary_loss_clip": 0.01210044, + "auxiliary_loss_mlp": 0.01041462, + "balance_loss_clip": 1.06312907, + "balance_loss_mlp": 1.03202057, + "epoch": 0.3446161245716347, + "flos": 21908633731200.0, + "grad_norm": 4.906994115207194, + "language_loss": 0.76970792, + "learning_rate": 3.0486145671144633e-06, + "loss": 0.79222298, + "num_input_tokens_seen": 61628130, + "step": 2866, + "time_per_iteration": 2.6448745727539062 + }, + { + "auxiliary_loss_clip": 0.01219406, + "auxiliary_loss_mlp": 0.01036132, + "balance_loss_clip": 1.0479182, + "balance_loss_mlp": 1.02662504, + "epoch": 0.3447363674622738, + "flos": 25112834461440.0, + "grad_norm": 2.6558857139376935, + "language_loss": 0.76604247, + "learning_rate": 3.047951170539086e-06, + "loss": 0.78859782, + "num_input_tokens_seen": 61647755, + "step": 2867, + "time_per_iteration": 2.850127935409546 + }, + { + "auxiliary_loss_clip": 0.01228064, + "auxiliary_loss_mlp": 0.01034889, + "balance_loss_clip": 1.05911839, + "balance_loss_mlp": 1.02644312, + "epoch": 0.3448566103529129, + "flos": 11984684451840.0, + "grad_norm": 2.116227428695121, + "language_loss": 0.84125811, + "learning_rate": 3.047287614990349e-06, + "loss": 0.86388767, + "num_input_tokens_seen": 61665675, + "step": 2868, + "time_per_iteration": 2.86226487159729 + }, + { + "auxiliary_loss_clip": 0.01207211, + "auxiliary_loss_mlp": 0.01028465, + "balance_loss_clip": 1.0565207, + "balance_loss_mlp": 1.01904202, + "epoch": 0.344976853243552, + "flos": 40187882465280.0, + "grad_norm": 2.9356713464704365, + "language_loss": 0.62009025, + "learning_rate": 3.046623900568914e-06, + "loss": 0.64244699, + "num_input_tokens_seen": 61688240, + "step": 2869, + "time_per_iteration": 2.7639758586883545 + }, + { + "auxiliary_loss_clip": 0.01213964, + "auxiliary_loss_mlp": 0.01034896, + "balance_loss_clip": 1.05574656, + "balance_loss_mlp": 1.02460206, + "epoch": 0.34509709613419104, + "flos": 28723652127360.0, + "grad_norm": 2.470427915710714, + "language_loss": 0.69993103, + "learning_rate": 3.045960027375465e-06, + "loss": 0.72241962, + "num_input_tokens_seen": 61706075, + "step": 2870, + "time_per_iteration": 2.71817684173584 + }, + { + "auxiliary_loss_clip": 0.01211746, + "auxiliary_loss_mlp": 0.01030915, + "balance_loss_clip": 1.05613422, + "balance_loss_mlp": 1.02053237, + "epoch": 0.34521733902483015, + "flos": 29967597982080.0, + "grad_norm": 2.688062526060621, + "language_loss": 0.8279438, + "learning_rate": 3.045295995510711e-06, + "loss": 0.85037041, + "num_input_tokens_seen": 61723045, + "step": 2871, + "time_per_iteration": 2.69704532623291 + }, + { + "auxiliary_loss_clip": 0.01215517, + "auxiliary_loss_mlp": 0.01028844, + "balance_loss_clip": 1.0565685, + "balance_loss_mlp": 1.02051711, + "epoch": 0.34533758191546926, + "flos": 27923059843200.0, + "grad_norm": 1.7582650349618136, + "language_loss": 0.73383439, + "learning_rate": 3.0446318050753865e-06, + "loss": 0.75627798, + "num_input_tokens_seen": 61743525, + "step": 2872, + "time_per_iteration": 3.6325278282165527 + }, + { + "auxiliary_loss_clip": 0.01201084, + "auxiliary_loss_mlp": 0.01038033, + "balance_loss_clip": 1.0567652, + "balance_loss_mlp": 1.02867544, + "epoch": 0.3454578248061083, + "flos": 27125879351040.0, + "grad_norm": 1.9887103270546278, + "language_loss": 0.77649188, + "learning_rate": 3.0439674561702474e-06, + "loss": 0.79888308, + "num_input_tokens_seen": 61763025, + "step": 2873, + "time_per_iteration": 4.586120128631592 + }, + { + "auxiliary_loss_clip": 0.01210495, + "auxiliary_loss_mlp": 0.01032165, + "balance_loss_clip": 1.06289124, + "balance_loss_mlp": 1.02304506, + "epoch": 0.3455780676967474, + "flos": 19024899166080.0, + "grad_norm": 2.15782829035991, + "language_loss": 0.88157606, + "learning_rate": 3.043302948896076e-06, + "loss": 0.90400261, + "num_input_tokens_seen": 61781630, + "step": 2874, + "time_per_iteration": 2.7373697757720947 + }, + { + "auxiliary_loss_clip": 0.01222777, + "auxiliary_loss_mlp": 0.01036759, + "balance_loss_clip": 1.05418968, + "balance_loss_mlp": 1.02709675, + "epoch": 0.34569831058738654, + "flos": 34496005507200.0, + "grad_norm": 2.208671099088299, + "language_loss": 0.6093328, + "learning_rate": 3.0426382833536756e-06, + "loss": 0.63192815, + "num_input_tokens_seen": 61804985, + "step": 2875, + "time_per_iteration": 2.809079647064209 + }, + { + "auxiliary_loss_clip": 0.01219583, + "auxiliary_loss_mlp": 0.01030779, + "balance_loss_clip": 1.05368054, + "balance_loss_mlp": 1.02139688, + "epoch": 0.3458185534780256, + "flos": 31138681098240.0, + "grad_norm": 2.4083160965877357, + "language_loss": 0.77810794, + "learning_rate": 3.041973459643877e-06, + "loss": 0.80061162, + "num_input_tokens_seen": 61824440, + "step": 2876, + "time_per_iteration": 2.8006982803344727 + }, + { + "auxiliary_loss_clip": 0.01224435, + "auxiliary_loss_mlp": 0.01030407, + "balance_loss_clip": 1.05063963, + "balance_loss_mlp": 1.02127552, + "epoch": 0.3459387963686647, + "flos": 32452508862720.0, + "grad_norm": 2.8092601843528753, + "language_loss": 0.66942823, + "learning_rate": 3.0413084778675334e-06, + "loss": 0.69197667, + "num_input_tokens_seen": 61845690, + "step": 2877, + "time_per_iteration": 2.8060853481292725 + }, + { + "auxiliary_loss_clip": 0.01206304, + "auxiliary_loss_mlp": 0.01380059, + "balance_loss_clip": 1.05459118, + "balance_loss_mlp": 1.00052047, + "epoch": 0.3460590392593038, + "flos": 24675658030080.0, + "grad_norm": 1.8450547731285702, + "language_loss": 0.83796418, + "learning_rate": 3.0406433381255214e-06, + "loss": 0.86382782, + "num_input_tokens_seen": 61863725, + "step": 2878, + "time_per_iteration": 2.757349967956543 + }, + { + "auxiliary_loss_clip": 0.01208367, + "auxiliary_loss_mlp": 0.01033735, + "balance_loss_clip": 1.06103015, + "balance_loss_mlp": 1.02475286, + "epoch": 0.34617928214994287, + "flos": 18807316531200.0, + "grad_norm": 2.235894437849546, + "language_loss": 0.81889272, + "learning_rate": 3.0399780405187425e-06, + "loss": 0.84131372, + "num_input_tokens_seen": 61882720, + "step": 2879, + "time_per_iteration": 3.5390844345092773 + }, + { + "auxiliary_loss_clip": 0.01205115, + "auxiliary_loss_mlp": 0.01031756, + "balance_loss_clip": 1.05990422, + "balance_loss_mlp": 1.02333426, + "epoch": 0.346299525040582, + "flos": 24857653265280.0, + "grad_norm": 1.9126083970879975, + "language_loss": 0.78506529, + "learning_rate": 3.0393125851481216e-06, + "loss": 0.80743396, + "num_input_tokens_seen": 61902595, + "step": 2880, + "time_per_iteration": 2.7266123294830322 + }, + { + "auxiliary_loss_clip": 0.01219444, + "auxiliary_loss_mlp": 0.01027748, + "balance_loss_clip": 1.05589139, + "balance_loss_mlp": 1.01908767, + "epoch": 0.3464197679312211, + "flos": 16434914025600.0, + "grad_norm": 2.1839118454246473, + "language_loss": 0.86476666, + "learning_rate": 3.038646972114608e-06, + "loss": 0.88723862, + "num_input_tokens_seen": 61918920, + "step": 2881, + "time_per_iteration": 2.693939685821533 + }, + { + "auxiliary_loss_clip": 0.0122214, + "auxiliary_loss_mlp": 0.0103648, + "balance_loss_clip": 1.05765867, + "balance_loss_mlp": 1.02784324, + "epoch": 0.34654001082186014, + "flos": 22382474970240.0, + "grad_norm": 1.9358210491497696, + "language_loss": 0.67760313, + "learning_rate": 3.037981201519174e-06, + "loss": 0.70018935, + "num_input_tokens_seen": 61939520, + "step": 2882, + "time_per_iteration": 2.685556173324585 + }, + { + "auxiliary_loss_clip": 0.01212948, + "auxiliary_loss_mlp": 0.01032048, + "balance_loss_clip": 1.06396139, + "balance_loss_mlp": 1.0232321, + "epoch": 0.34666025371249926, + "flos": 19573901614080.0, + "grad_norm": 2.335772056132602, + "language_loss": 0.71623683, + "learning_rate": 3.0373152734628175e-06, + "loss": 0.7386868, + "num_input_tokens_seen": 61957800, + "step": 2883, + "time_per_iteration": 2.687575578689575 + }, + { + "auxiliary_loss_clip": 0.01202185, + "auxiliary_loss_mlp": 0.01031636, + "balance_loss_clip": 1.05686665, + "balance_loss_mlp": 1.02278495, + "epoch": 0.34678049660313837, + "flos": 15267637751040.0, + "grad_norm": 2.051793602671928, + "language_loss": 0.76143658, + "learning_rate": 3.0366491880465584e-06, + "loss": 0.78377485, + "num_input_tokens_seen": 61975820, + "step": 2884, + "time_per_iteration": 2.593820810317993 + }, + { + "auxiliary_loss_clip": 0.01207207, + "auxiliary_loss_mlp": 0.01032977, + "balance_loss_clip": 1.06546319, + "balance_loss_mlp": 1.02338624, + "epoch": 0.3469007394937774, + "flos": 21181550630400.0, + "grad_norm": 1.5762563390687268, + "language_loss": 0.81969088, + "learning_rate": 3.035982945371443e-06, + "loss": 0.84209275, + "num_input_tokens_seen": 61997515, + "step": 2885, + "time_per_iteration": 2.677539825439453 + }, + { + "auxiliary_loss_clip": 0.01227518, + "auxiliary_loss_mlp": 0.01033696, + "balance_loss_clip": 1.06071973, + "balance_loss_mlp": 1.02426028, + "epoch": 0.34702098238441653, + "flos": 22375471818240.0, + "grad_norm": 2.166848229681769, + "language_loss": 0.8493576, + "learning_rate": 3.035316545538537e-06, + "loss": 0.87196976, + "num_input_tokens_seen": 62016310, + "step": 2886, + "time_per_iteration": 2.6826791763305664 + }, + { + "auxiliary_loss_clip": 0.01218673, + "auxiliary_loss_mlp": 0.01035889, + "balance_loss_clip": 1.06259537, + "balance_loss_mlp": 1.0271095, + "epoch": 0.3471412252750556, + "flos": 22929430343040.0, + "grad_norm": 2.280306310765405, + "language_loss": 0.79375285, + "learning_rate": 3.034649988648935e-06, + "loss": 0.81629848, + "num_input_tokens_seen": 62036075, + "step": 2887, + "time_per_iteration": 2.753791093826294 + }, + { + "auxiliary_loss_clip": 0.01219196, + "auxiliary_loss_mlp": 0.01027382, + "balance_loss_clip": 1.05838645, + "balance_loss_mlp": 1.01895428, + "epoch": 0.3472614681656947, + "flos": 21324259365120.0, + "grad_norm": 1.7242725364007356, + "language_loss": 0.80590618, + "learning_rate": 3.033983274803752e-06, + "loss": 0.828372, + "num_input_tokens_seen": 62055865, + "step": 2888, + "time_per_iteration": 2.6646907329559326 + }, + { + "auxiliary_loss_clip": 0.01213477, + "auxiliary_loss_mlp": 0.01033764, + "balance_loss_clip": 1.05687976, + "balance_loss_mlp": 1.02484751, + "epoch": 0.3473817110563338, + "flos": 23475739271040.0, + "grad_norm": 2.468061635653163, + "language_loss": 0.72783828, + "learning_rate": 3.0333164041041283e-06, + "loss": 0.75031066, + "num_input_tokens_seen": 62072180, + "step": 2889, + "time_per_iteration": 2.752019166946411 + }, + { + "auxiliary_loss_clip": 0.01247904, + "auxiliary_loss_mlp": 0.01027214, + "balance_loss_clip": 1.05585432, + "balance_loss_mlp": 1.01802921, + "epoch": 0.34750195394697286, + "flos": 22346025644160.0, + "grad_norm": 2.241000366561078, + "language_loss": 0.72163564, + "learning_rate": 3.032649376651228e-06, + "loss": 0.74438679, + "num_input_tokens_seen": 62091600, + "step": 2890, + "time_per_iteration": 2.91975736618042 + }, + { + "auxiliary_loss_clip": 0.01226728, + "auxiliary_loss_mlp": 0.01032607, + "balance_loss_clip": 1.05749953, + "balance_loss_mlp": 1.023422, + "epoch": 0.347622196837612, + "flos": 29095004885760.0, + "grad_norm": 1.9398002734361113, + "language_loss": 0.75705588, + "learning_rate": 3.031982192546238e-06, + "loss": 0.7796492, + "num_input_tokens_seen": 62114695, + "step": 2891, + "time_per_iteration": 3.0426156520843506 + }, + { + "auxiliary_loss_clip": 0.01210074, + "auxiliary_loss_mlp": 0.01032369, + "balance_loss_clip": 1.06033325, + "balance_loss_mlp": 1.02376246, + "epoch": 0.3477424397282511, + "flos": 22455732758400.0, + "grad_norm": 2.8683733470862864, + "language_loss": 0.94987553, + "learning_rate": 3.0313148518903696e-06, + "loss": 0.97229993, + "num_input_tokens_seen": 62134520, + "step": 2892, + "time_per_iteration": 2.6711254119873047 + }, + { + "auxiliary_loss_clip": 0.01219471, + "auxiliary_loss_mlp": 0.01031901, + "balance_loss_clip": 1.05855989, + "balance_loss_mlp": 1.02256083, + "epoch": 0.34786268261889014, + "flos": 15778790242560.0, + "grad_norm": 2.005291906982554, + "language_loss": 0.81066418, + "learning_rate": 3.030647354784859e-06, + "loss": 0.83317792, + "num_input_tokens_seen": 62151560, + "step": 2893, + "time_per_iteration": 2.5943715572357178 + }, + { + "auxiliary_loss_clip": 0.01224318, + "auxiliary_loss_mlp": 0.01030476, + "balance_loss_clip": 1.05774856, + "balance_loss_mlp": 1.02187538, + "epoch": 0.34798292550952925, + "flos": 20777627214720.0, + "grad_norm": 2.0522735499007267, + "language_loss": 0.77173269, + "learning_rate": 3.029979701330964e-06, + "loss": 0.79428071, + "num_input_tokens_seen": 62170985, + "step": 2894, + "time_per_iteration": 2.7110671997070312 + }, + { + "auxiliary_loss_clip": 0.01221077, + "auxiliary_loss_mlp": 0.01031458, + "balance_loss_clip": 1.05728137, + "balance_loss_mlp": 1.02232647, + "epoch": 0.34810316840016836, + "flos": 19937820257280.0, + "grad_norm": 2.1187005503938825, + "language_loss": 0.80158496, + "learning_rate": 3.029311891629966e-06, + "loss": 0.82411021, + "num_input_tokens_seen": 62189440, + "step": 2895, + "time_per_iteration": 2.675396203994751 + }, + { + "auxiliary_loss_clip": 0.01218111, + "auxiliary_loss_mlp": 0.01034333, + "balance_loss_clip": 1.06012702, + "balance_loss_mlp": 1.02521944, + "epoch": 0.3482234112908074, + "flos": 23623296341760.0, + "grad_norm": 1.9314005828295704, + "language_loss": 0.74661797, + "learning_rate": 3.0286439257831744e-06, + "loss": 0.76914245, + "num_input_tokens_seen": 62208910, + "step": 2896, + "time_per_iteration": 2.70467472076416 + }, + { + "auxiliary_loss_clip": 0.01205982, + "auxiliary_loss_mlp": 0.01032599, + "balance_loss_clip": 1.06250596, + "balance_loss_mlp": 1.02283597, + "epoch": 0.3483436541814465, + "flos": 23986712194560.0, + "grad_norm": 1.8588578653203887, + "language_loss": 0.71173477, + "learning_rate": 3.0279758038919156e-06, + "loss": 0.73412055, + "num_input_tokens_seen": 62227135, + "step": 2897, + "time_per_iteration": 2.5964086055755615 + }, + { + "auxiliary_loss_clip": 0.01212344, + "auxiliary_loss_mlp": 0.0103264, + "balance_loss_clip": 1.06150317, + "balance_loss_mlp": 1.0231514, + "epoch": 0.34846389707208564, + "flos": 22638338524800.0, + "grad_norm": 1.8160487859318324, + "language_loss": 0.78204739, + "learning_rate": 3.0273075260575455e-06, + "loss": 0.80449718, + "num_input_tokens_seen": 62246035, + "step": 2898, + "time_per_iteration": 3.6484007835388184 + }, + { + "auxiliary_loss_clip": 0.01218449, + "auxiliary_loss_mlp": 0.01029772, + "balance_loss_clip": 1.05749345, + "balance_loss_mlp": 1.02049124, + "epoch": 0.3485841399627247, + "flos": 21792857218560.0, + "grad_norm": 1.8767622701133717, + "language_loss": 0.80832696, + "learning_rate": 3.0266390923814396e-06, + "loss": 0.83080924, + "num_input_tokens_seen": 62264095, + "step": 2899, + "time_per_iteration": 3.5977540016174316 + }, + { + "auxiliary_loss_clip": 0.01220468, + "auxiliary_loss_mlp": 0.01036955, + "balance_loss_clip": 1.06058347, + "balance_loss_mlp": 1.02742457, + "epoch": 0.3487043828533638, + "flos": 17019036996480.0, + "grad_norm": 3.302950597121684, + "language_loss": 0.82011938, + "learning_rate": 3.0259705029650008e-06, + "loss": 0.84269357, + "num_input_tokens_seen": 62282025, + "step": 2900, + "time_per_iteration": 2.6347038745880127 + }, + { + "auxiliary_loss_clip": 0.01209412, + "auxiliary_loss_mlp": 0.01031535, + "balance_loss_clip": 1.05807543, + "balance_loss_mlp": 1.02282691, + "epoch": 0.34882462574400286, + "flos": 22601135013120.0, + "grad_norm": 1.7460145689810849, + "language_loss": 0.72600794, + "learning_rate": 3.025301757909652e-06, + "loss": 0.74841744, + "num_input_tokens_seen": 62302220, + "step": 2901, + "time_per_iteration": 2.723954439163208 + }, + { + "auxiliary_loss_clip": 0.01230965, + "auxiliary_loss_mlp": 0.01380626, + "balance_loss_clip": 1.05841911, + "balance_loss_mlp": 1.00058198, + "epoch": 0.34894486863464197, + "flos": 29861518141440.0, + "grad_norm": 1.7754018163393082, + "language_loss": 0.80894482, + "learning_rate": 3.024632857316842e-06, + "loss": 0.83506072, + "num_input_tokens_seen": 62323535, + "step": 2902, + "time_per_iteration": 2.7410969734191895 + }, + { + "auxiliary_loss_clip": 0.01212699, + "auxiliary_loss_mlp": 0.01031389, + "balance_loss_clip": 1.06262016, + "balance_loss_mlp": 1.02278197, + "epoch": 0.3490651115252811, + "flos": 22122265870080.0, + "grad_norm": 3.236025773294599, + "language_loss": 0.77367353, + "learning_rate": 3.0239638012880412e-06, + "loss": 0.79611444, + "num_input_tokens_seen": 62343430, + "step": 2903, + "time_per_iteration": 2.7076714038848877 + }, + { + "auxiliary_loss_clip": 0.01225652, + "auxiliary_loss_mlp": 0.01030067, + "balance_loss_clip": 1.05165911, + "balance_loss_mlp": 1.02023268, + "epoch": 0.34918535441592014, + "flos": 12676682943360.0, + "grad_norm": 2.2895401922286878, + "language_loss": 0.81195152, + "learning_rate": 3.0232945899247466e-06, + "loss": 0.83450866, + "num_input_tokens_seen": 62360365, + "step": 2904, + "time_per_iteration": 2.6864640712738037 + }, + { + "auxiliary_loss_clip": 0.01211637, + "auxiliary_loss_mlp": 0.01038759, + "balance_loss_clip": 1.06043601, + "balance_loss_mlp": 1.02901983, + "epoch": 0.34930559730655925, + "flos": 23185617120000.0, + "grad_norm": 2.165603319814757, + "language_loss": 0.77394986, + "learning_rate": 3.022625223328476e-06, + "loss": 0.79645383, + "num_input_tokens_seen": 62382105, + "step": 2905, + "time_per_iteration": 2.7037081718444824 + }, + { + "auxiliary_loss_clip": 0.0121806, + "auxiliary_loss_mlp": 0.01030925, + "balance_loss_clip": 1.06158233, + "balance_loss_mlp": 1.02107787, + "epoch": 0.34942584019719836, + "flos": 22855023319680.0, + "grad_norm": 2.138767619844283, + "language_loss": 0.69283235, + "learning_rate": 3.0219557016007723e-06, + "loss": 0.7153222, + "num_input_tokens_seen": 62402235, + "step": 2906, + "time_per_iteration": 3.5912604331970215 + }, + { + "auxiliary_loss_clip": 0.01207783, + "auxiliary_loss_mlp": 0.01032179, + "balance_loss_clip": 1.0611968, + "balance_loss_mlp": 1.02292264, + "epoch": 0.3495460830878374, + "flos": 24426043441920.0, + "grad_norm": 1.8928674169080906, + "language_loss": 0.69612116, + "learning_rate": 3.021286024843202e-06, + "loss": 0.71852076, + "num_input_tokens_seen": 62420430, + "step": 2907, + "time_per_iteration": 2.6359410285949707 + }, + { + "auxiliary_loss_clip": 0.01100656, + "auxiliary_loss_mlp": 0.01002051, + "balance_loss_clip": 1.03210998, + "balance_loss_mlp": 1.00047767, + "epoch": 0.3496663259784765, + "flos": 70008749389440.0, + "grad_norm": 1.0610947556520716, + "language_loss": 0.64747751, + "learning_rate": 3.0206161931573526e-06, + "loss": 0.66850448, + "num_input_tokens_seen": 62472980, + "step": 2908, + "time_per_iteration": 3.1275389194488525 + }, + { + "auxiliary_loss_clip": 0.01214533, + "auxiliary_loss_mlp": 0.01034124, + "balance_loss_clip": 1.05588055, + "balance_loss_mlp": 1.02499831, + "epoch": 0.34978656886911563, + "flos": 28692805322880.0, + "grad_norm": 1.5334586515549087, + "language_loss": 0.92764235, + "learning_rate": 3.0199462066448388e-06, + "loss": 0.95012891, + "num_input_tokens_seen": 62495175, + "step": 2909, + "time_per_iteration": 2.7191216945648193 + }, + { + "auxiliary_loss_clip": 0.01215483, + "auxiliary_loss_mlp": 0.01028107, + "balance_loss_clip": 1.06351018, + "balance_loss_mlp": 1.01873684, + "epoch": 0.3499068117597547, + "flos": 21142156389120.0, + "grad_norm": 1.860962605100214, + "language_loss": 0.6926825, + "learning_rate": 3.019276065407296e-06, + "loss": 0.71511841, + "num_input_tokens_seen": 62514295, + "step": 2910, + "time_per_iteration": 2.655341386795044 + }, + { + "auxiliary_loss_clip": 0.0123553, + "auxiliary_loss_mlp": 0.0103471, + "balance_loss_clip": 1.055372, + "balance_loss_mlp": 1.02524447, + "epoch": 0.3500270546503938, + "flos": 22782699285120.0, + "grad_norm": 1.7386490046744307, + "language_loss": 0.80562454, + "learning_rate": 3.018605769546385e-06, + "loss": 0.82832694, + "num_input_tokens_seen": 62534850, + "step": 2911, + "time_per_iteration": 2.7344858646392822 + }, + { + "auxiliary_loss_clip": 0.01206529, + "auxiliary_loss_mlp": 0.01031635, + "balance_loss_clip": 1.05791497, + "balance_loss_mlp": 1.02189612, + "epoch": 0.3501472975410329, + "flos": 22894058424960.0, + "grad_norm": 1.9894016223291748, + "language_loss": 0.79670715, + "learning_rate": 3.017935319163788e-06, + "loss": 0.81908882, + "num_input_tokens_seen": 62553810, + "step": 2912, + "time_per_iteration": 2.712541341781616 + }, + { + "auxiliary_loss_clip": 0.012136, + "auxiliary_loss_mlp": 0.01031633, + "balance_loss_clip": 1.06123722, + "balance_loss_mlp": 1.02144063, + "epoch": 0.35026754043167196, + "flos": 25446588658560.0, + "grad_norm": 1.8482001567030635, + "language_loss": 0.70485997, + "learning_rate": 3.017264714361213e-06, + "loss": 0.72731233, + "num_input_tokens_seen": 62573460, + "step": 2913, + "time_per_iteration": 2.67457914352417 + }, + { + "auxiliary_loss_clip": 0.01218256, + "auxiliary_loss_mlp": 0.01380421, + "balance_loss_clip": 1.05693483, + "balance_loss_mlp": 1.00057411, + "epoch": 0.3503877833223111, + "flos": 19573757959680.0, + "grad_norm": 2.48968374292679, + "language_loss": 0.82422984, + "learning_rate": 3.016593955240389e-06, + "loss": 0.85021663, + "num_input_tokens_seen": 62592150, + "step": 2914, + "time_per_iteration": 2.732919692993164 + }, + { + "auxiliary_loss_clip": 0.01110508, + "auxiliary_loss_mlp": 0.01002895, + "balance_loss_clip": 1.0308919, + "balance_loss_mlp": 1.00123775, + "epoch": 0.3505080262129502, + "flos": 65072075880960.0, + "grad_norm": 0.8294071645597316, + "language_loss": 0.63720262, + "learning_rate": 3.015923041903071e-06, + "loss": 0.65833664, + "num_input_tokens_seen": 62658275, + "step": 2915, + "time_per_iteration": 3.304044723510742 + }, + { + "auxiliary_loss_clip": 0.01212375, + "auxiliary_loss_mlp": 0.01032328, + "balance_loss_clip": 1.06287551, + "balance_loss_mlp": 1.02311945, + "epoch": 0.35062826910358924, + "flos": 29314562768640.0, + "grad_norm": 2.502906659391027, + "language_loss": 0.833556, + "learning_rate": 3.0152519744510347e-06, + "loss": 0.85600305, + "num_input_tokens_seen": 62678075, + "step": 2916, + "time_per_iteration": 2.692688465118408 + }, + { + "auxiliary_loss_clip": 0.01228719, + "auxiliary_loss_mlp": 0.01036125, + "balance_loss_clip": 1.05554247, + "balance_loss_mlp": 1.02670121, + "epoch": 0.35074851199422835, + "flos": 23987717775360.0, + "grad_norm": 2.9286232416458664, + "language_loss": 0.82942909, + "learning_rate": 3.014580752986081e-06, + "loss": 0.85207754, + "num_input_tokens_seen": 62696950, + "step": 2917, + "time_per_iteration": 2.7953941822052 + }, + { + "auxiliary_loss_clip": 0.0123959, + "auxiliary_loss_mlp": 0.01029666, + "balance_loss_clip": 1.0584085, + "balance_loss_mlp": 1.02009988, + "epoch": 0.3508687548848674, + "flos": 15224436668160.0, + "grad_norm": 1.9928041406571915, + "language_loss": 0.78776252, + "learning_rate": 3.0139093776100345e-06, + "loss": 0.81045508, + "num_input_tokens_seen": 62713540, + "step": 2918, + "time_per_iteration": 2.6853716373443604 + }, + { + "auxiliary_loss_clip": 0.01197983, + "auxiliary_loss_mlp": 0.01027843, + "balance_loss_clip": 1.05915928, + "balance_loss_mlp": 1.01893806, + "epoch": 0.3509889977755065, + "flos": 21361750185600.0, + "grad_norm": 2.048030433759377, + "language_loss": 0.75473213, + "learning_rate": 3.013237848424741e-06, + "loss": 0.77699047, + "num_input_tokens_seen": 62732925, + "step": 2919, + "time_per_iteration": 2.6186070442199707 + }, + { + "auxiliary_loss_clip": 0.01222717, + "auxiliary_loss_mlp": 0.01029047, + "balance_loss_clip": 1.06062388, + "balance_loss_mlp": 1.01949239, + "epoch": 0.35110924066614563, + "flos": 19135360465920.0, + "grad_norm": 3.9251746252402206, + "language_loss": 0.75174028, + "learning_rate": 3.012566165532072e-06, + "loss": 0.7742579, + "num_input_tokens_seen": 62751715, + "step": 2920, + "time_per_iteration": 2.6482956409454346 + }, + { + "auxiliary_loss_clip": 0.01246576, + "auxiliary_loss_mlp": 0.01038722, + "balance_loss_clip": 1.05177999, + "balance_loss_mlp": 1.02916729, + "epoch": 0.3512294835567847, + "flos": 21980885938560.0, + "grad_norm": 2.1018153451831987, + "language_loss": 0.76978475, + "learning_rate": 3.0118943290339207e-06, + "loss": 0.79263771, + "num_input_tokens_seen": 62771925, + "step": 2921, + "time_per_iteration": 2.7517621517181396 + }, + { + "auxiliary_loss_clip": 0.01211288, + "auxiliary_loss_mlp": 0.01036403, + "balance_loss_clip": 1.05138958, + "balance_loss_mlp": 1.02702713, + "epoch": 0.3513497264474238, + "flos": 17817294896640.0, + "grad_norm": 1.8545825401184959, + "language_loss": 0.68011737, + "learning_rate": 3.011222339032204e-06, + "loss": 0.70259428, + "num_input_tokens_seen": 62790075, + "step": 2922, + "time_per_iteration": 2.6752982139587402 + }, + { + "auxiliary_loss_clip": 0.01201077, + "auxiliary_loss_mlp": 0.01034989, + "balance_loss_clip": 1.06120694, + "balance_loss_mlp": 1.02632236, + "epoch": 0.3514699693380629, + "flos": 26943417239040.0, + "grad_norm": 2.1053062110927536, + "language_loss": 0.69799566, + "learning_rate": 3.0105501956288626e-06, + "loss": 0.72035635, + "num_input_tokens_seen": 62810545, + "step": 2923, + "time_per_iteration": 2.6145572662353516 + }, + { + "auxiliary_loss_clip": 0.01217034, + "auxiliary_loss_mlp": 0.0103092, + "balance_loss_clip": 1.06062579, + "balance_loss_mlp": 1.02149057, + "epoch": 0.35159021222870196, + "flos": 15267565923840.0, + "grad_norm": 2.3568633537906174, + "language_loss": 0.72702473, + "learning_rate": 3.0098778989258602e-06, + "loss": 0.74950427, + "num_input_tokens_seen": 62829155, + "step": 2924, + "time_per_iteration": 2.643968343734741 + }, + { + "auxiliary_loss_clip": 0.01223516, + "auxiliary_loss_mlp": 0.01036186, + "balance_loss_clip": 1.05756259, + "balance_loss_mlp": 1.0266968, + "epoch": 0.35171045511934107, + "flos": 13984154000640.0, + "grad_norm": 1.9412535471942491, + "language_loss": 0.885517, + "learning_rate": 3.009205449025183e-06, + "loss": 0.90811402, + "num_input_tokens_seen": 62845350, + "step": 2925, + "time_per_iteration": 5.426146984100342 + }, + { + "auxiliary_loss_clip": 0.01220388, + "auxiliary_loss_mlp": 0.0103202, + "balance_loss_clip": 1.05221379, + "balance_loss_mlp": 1.02334166, + "epoch": 0.3518306980099802, + "flos": 14283434119680.0, + "grad_norm": 1.8132692118239562, + "language_loss": 0.63173211, + "learning_rate": 3.008532846028842e-06, + "loss": 0.65425622, + "num_input_tokens_seen": 62862110, + "step": 2926, + "time_per_iteration": 2.649099826812744 + }, + { + "auxiliary_loss_clip": 0.01203873, + "auxiliary_loss_mlp": 0.01032307, + "balance_loss_clip": 1.06163526, + "balance_loss_mlp": 1.02220964, + "epoch": 0.35195094090061924, + "flos": 27052872958080.0, + "grad_norm": 4.595762676279448, + "language_loss": 0.71690845, + "learning_rate": 3.0078600900388694e-06, + "loss": 0.73927021, + "num_input_tokens_seen": 62882415, + "step": 2927, + "time_per_iteration": 2.655658483505249 + }, + { + "auxiliary_loss_clip": 0.01211884, + "auxiliary_loss_mlp": 0.0103625, + "balance_loss_clip": 1.05195713, + "balance_loss_mlp": 1.02587879, + "epoch": 0.35207118379125835, + "flos": 25629266252160.0, + "grad_norm": 2.7420291739867637, + "language_loss": 0.74127364, + "learning_rate": 3.007187181157323e-06, + "loss": 0.76375496, + "num_input_tokens_seen": 62902425, + "step": 2928, + "time_per_iteration": 2.711954355239868 + }, + { + "auxiliary_loss_clip": 0.01229894, + "auxiliary_loss_mlp": 0.01030715, + "balance_loss_clip": 1.04960239, + "balance_loss_mlp": 1.02142262, + "epoch": 0.35219142668189746, + "flos": 18004713085440.0, + "grad_norm": 2.795045475669485, + "language_loss": 0.68213129, + "learning_rate": 3.006514119486282e-06, + "loss": 0.70473742, + "num_input_tokens_seen": 62919255, + "step": 2929, + "time_per_iteration": 2.925434112548828 + }, + { + "auxiliary_loss_clip": 0.01221785, + "auxiliary_loss_mlp": 0.01035922, + "balance_loss_clip": 1.05499494, + "balance_loss_mlp": 1.0267607, + "epoch": 0.3523116695725365, + "flos": 14028109269120.0, + "grad_norm": 1.6737051841284258, + "language_loss": 0.69655097, + "learning_rate": 3.005840905127849e-06, + "loss": 0.71912795, + "num_input_tokens_seen": 62936160, + "step": 2930, + "time_per_iteration": 2.6687445640563965 + }, + { + "auxiliary_loss_clip": 0.0120378, + "auxiliary_loss_mlp": 0.0103398, + "balance_loss_clip": 1.06375694, + "balance_loss_mlp": 1.02485478, + "epoch": 0.3524319124631756, + "flos": 21433966479360.0, + "grad_norm": 9.798288450157738, + "language_loss": 0.86890841, + "learning_rate": 3.0051675381841516e-06, + "loss": 0.89128602, + "num_input_tokens_seen": 62953470, + "step": 2931, + "time_per_iteration": 2.692136287689209 + }, + { + "auxiliary_loss_clip": 0.01248677, + "auxiliary_loss_mlp": 0.01380674, + "balance_loss_clip": 1.04887462, + "balance_loss_mlp": 1.00041056, + "epoch": 0.3525521553538147, + "flos": 26322773114880.0, + "grad_norm": 1.5392538132740654, + "language_loss": 0.76700473, + "learning_rate": 3.0044940187573363e-06, + "loss": 0.79329824, + "num_input_tokens_seen": 62974480, + "step": 2932, + "time_per_iteration": 3.853634834289551 + }, + { + "auxiliary_loss_clip": 0.01213347, + "auxiliary_loss_mlp": 0.01032681, + "balance_loss_clip": 1.06024516, + "balance_loss_mlp": 1.02349579, + "epoch": 0.3526723982444538, + "flos": 21543314457600.0, + "grad_norm": 2.1416089570711163, + "language_loss": 0.65188348, + "learning_rate": 3.003820346949578e-06, + "loss": 0.67434371, + "num_input_tokens_seen": 62992560, + "step": 2933, + "time_per_iteration": 3.030496120452881 + }, + { + "auxiliary_loss_clip": 0.01205209, + "auxiliary_loss_mlp": 0.01037133, + "balance_loss_clip": 1.06170106, + "balance_loss_mlp": 1.02722692, + "epoch": 0.3527926411350929, + "flos": 23733649900800.0, + "grad_norm": 2.032236868429703, + "language_loss": 0.79645562, + "learning_rate": 3.003146522863071e-06, + "loss": 0.81887901, + "num_input_tokens_seen": 63013445, + "step": 2934, + "time_per_iteration": 2.756747245788574 + }, + { + "auxiliary_loss_clip": 0.01221598, + "auxiliary_loss_mlp": 0.01034256, + "balance_loss_clip": 1.06128478, + "balance_loss_mlp": 1.02545214, + "epoch": 0.35291288402573195, + "flos": 30445461544320.0, + "grad_norm": 2.2046323674221058, + "language_loss": 0.85855556, + "learning_rate": 3.0024725466000345e-06, + "loss": 0.88111407, + "num_input_tokens_seen": 63033400, + "step": 2935, + "time_per_iteration": 2.7236390113830566 + }, + { + "auxiliary_loss_clip": 0.01213113, + "auxiliary_loss_mlp": 0.01028168, + "balance_loss_clip": 1.06278431, + "balance_loss_mlp": 1.01916146, + "epoch": 0.35303312691637107, + "flos": 23112179763840.0, + "grad_norm": 6.100841904918143, + "language_loss": 0.79171848, + "learning_rate": 3.0017984182627087e-06, + "loss": 0.81413126, + "num_input_tokens_seen": 63052725, + "step": 2936, + "time_per_iteration": 2.6569244861602783 + }, + { + "auxiliary_loss_clip": 0.01224044, + "auxiliary_loss_mlp": 0.01380211, + "balance_loss_clip": 1.05397415, + "balance_loss_mlp": 1.00046921, + "epoch": 0.3531533698070102, + "flos": 21835699165440.0, + "grad_norm": 2.0542364686370944, + "language_loss": 0.82407391, + "learning_rate": 3.00112413795336e-06, + "loss": 0.85011649, + "num_input_tokens_seen": 63072560, + "step": 2937, + "time_per_iteration": 2.665658950805664 + }, + { + "auxiliary_loss_clip": 0.01211901, + "auxiliary_loss_mlp": 0.01033115, + "balance_loss_clip": 1.05331504, + "balance_loss_mlp": 1.02371526, + "epoch": 0.35327361269764923, + "flos": 15778969810560.0, + "grad_norm": 1.854068760485824, + "language_loss": 0.79725814, + "learning_rate": 3.000449705774275e-06, + "loss": 0.81970829, + "num_input_tokens_seen": 63090800, + "step": 2938, + "time_per_iteration": 2.654789686203003 + }, + { + "auxiliary_loss_clip": 0.01215404, + "auxiliary_loss_mlp": 0.01036033, + "balance_loss_clip": 1.06372488, + "balance_loss_mlp": 1.02609062, + "epoch": 0.35339385558828834, + "flos": 22090413484800.0, + "grad_norm": 2.1672907281210287, + "language_loss": 0.71820176, + "learning_rate": 2.9997751218277654e-06, + "loss": 0.7407161, + "num_input_tokens_seen": 63108955, + "step": 2939, + "time_per_iteration": 2.829714775085449 + }, + { + "auxiliary_loss_clip": 0.01202924, + "auxiliary_loss_mlp": 0.01030964, + "balance_loss_clip": 1.06059444, + "balance_loss_mlp": 1.02179706, + "epoch": 0.35351409847892745, + "flos": 24165008328960.0, + "grad_norm": 1.9306422955967093, + "language_loss": 0.77671194, + "learning_rate": 2.999100386216166e-06, + "loss": 0.79905081, + "num_input_tokens_seen": 63127895, + "step": 2940, + "time_per_iteration": 2.5765316486358643 + }, + { + "auxiliary_loss_clip": 0.01219949, + "auxiliary_loss_mlp": 0.01031081, + "balance_loss_clip": 1.05841374, + "balance_loss_mlp": 1.02218795, + "epoch": 0.3536343413695665, + "flos": 27052298340480.0, + "grad_norm": 2.2965550455604786, + "language_loss": 0.74463642, + "learning_rate": 2.998425499041831e-06, + "loss": 0.76714671, + "num_input_tokens_seen": 63148410, + "step": 2941, + "time_per_iteration": 2.6984009742736816 + }, + { + "auxiliary_loss_clip": 0.01114916, + "auxiliary_loss_mlp": 0.01010216, + "balance_loss_clip": 1.03249764, + "balance_loss_mlp": 1.008726, + "epoch": 0.3537545842602056, + "flos": 65991066370560.0, + "grad_norm": 1.2439017478355003, + "language_loss": 0.6454165, + "learning_rate": 2.997750460407142e-06, + "loss": 0.66666782, + "num_input_tokens_seen": 63209765, + "step": 2942, + "time_per_iteration": 3.2470715045928955 + }, + { + "auxiliary_loss_clip": 0.0123324, + "auxiliary_loss_mlp": 0.01038941, + "balance_loss_clip": 1.05662203, + "balance_loss_mlp": 1.02824211, + "epoch": 0.35387482715084473, + "flos": 18436897526400.0, + "grad_norm": 1.9665239450822343, + "language_loss": 0.70021152, + "learning_rate": 2.997075270414501e-06, + "loss": 0.72293335, + "num_input_tokens_seen": 63226980, + "step": 2943, + "time_per_iteration": 2.743236541748047 + }, + { + "auxiliary_loss_clip": 0.0112795, + "auxiliary_loss_mlp": 0.01002632, + "balance_loss_clip": 1.03482771, + "balance_loss_mlp": 1.0010823, + "epoch": 0.3539950700414838, + "flos": 65588579498880.0, + "grad_norm": 0.701419992144887, + "language_loss": 0.57727659, + "learning_rate": 2.9963999291663347e-06, + "loss": 0.59858245, + "num_input_tokens_seen": 63292760, + "step": 2944, + "time_per_iteration": 3.2716147899627686 + }, + { + "auxiliary_loss_clip": 0.01236095, + "auxiliary_loss_mlp": 0.01032911, + "balance_loss_clip": 1.05683446, + "balance_loss_mlp": 1.0241791, + "epoch": 0.3541153129321229, + "flos": 20521655919360.0, + "grad_norm": 3.1649713627146383, + "language_loss": 0.73701388, + "learning_rate": 2.9957244367650915e-06, + "loss": 0.75970393, + "num_input_tokens_seen": 63309005, + "step": 2945, + "time_per_iteration": 2.768599033355713 + }, + { + "auxiliary_loss_clip": 0.01224511, + "auxiliary_loss_mlp": 0.01035352, + "balance_loss_clip": 1.05224741, + "balance_loss_mlp": 1.02687025, + "epoch": 0.354235555822762, + "flos": 19573578391680.0, + "grad_norm": 2.0825141927087882, + "language_loss": 0.83670753, + "learning_rate": 2.9950487933132425e-06, + "loss": 0.85930616, + "num_input_tokens_seen": 63326420, + "step": 2946, + "time_per_iteration": 2.7576725482940674 + }, + { + "auxiliary_loss_clip": 0.01213058, + "auxiliary_loss_mlp": 0.01034369, + "balance_loss_clip": 1.05850792, + "balance_loss_mlp": 1.02548766, + "epoch": 0.35435579871340106, + "flos": 20777268078720.0, + "grad_norm": 2.0726991188983472, + "language_loss": 0.71705484, + "learning_rate": 2.994372998913283e-06, + "loss": 0.73952901, + "num_input_tokens_seen": 63344925, + "step": 2947, + "time_per_iteration": 2.67643141746521 + }, + { + "auxiliary_loss_clip": 0.01222404, + "auxiliary_loss_mlp": 0.01035052, + "balance_loss_clip": 1.06055069, + "balance_loss_mlp": 1.02593851, + "epoch": 0.35447604160404017, + "flos": 23951807153280.0, + "grad_norm": 2.0930798485189985, + "language_loss": 0.62640953, + "learning_rate": 2.99369705366773e-06, + "loss": 0.64898407, + "num_input_tokens_seen": 63365170, + "step": 2948, + "time_per_iteration": 2.7092151641845703 + }, + { + "auxiliary_loss_clip": 0.0121223, + "auxiliary_loss_mlp": 0.010343, + "balance_loss_clip": 1.05646086, + "balance_loss_mlp": 1.02546024, + "epoch": 0.3545962844946792, + "flos": 23435662671360.0, + "grad_norm": 2.5416165047213797, + "language_loss": 0.82034731, + "learning_rate": 2.9930209576791244e-06, + "loss": 0.8428126, + "num_input_tokens_seen": 63383645, + "step": 2949, + "time_per_iteration": 2.6905972957611084 + }, + { + "auxiliary_loss_clip": 0.01207302, + "auxiliary_loss_mlp": 0.01029617, + "balance_loss_clip": 1.0587101, + "balance_loss_mlp": 1.02061677, + "epoch": 0.35471652738531834, + "flos": 22085134185600.0, + "grad_norm": 2.1631164848310944, + "language_loss": 0.63850093, + "learning_rate": 2.9923447110500285e-06, + "loss": 0.66087008, + "num_input_tokens_seen": 63402390, + "step": 2950, + "time_per_iteration": 3.5444438457489014 + }, + { + "auxiliary_loss_clip": 0.01198835, + "auxiliary_loss_mlp": 0.01034808, + "balance_loss_clip": 1.05894208, + "balance_loss_mlp": 1.02517629, + "epoch": 0.35483677027595745, + "flos": 27341881787520.0, + "grad_norm": 1.4663681420006043, + "language_loss": 0.75282562, + "learning_rate": 2.9916683138830295e-06, + "loss": 0.77516198, + "num_input_tokens_seen": 63423055, + "step": 2951, + "time_per_iteration": 4.493996620178223 + }, + { + "auxiliary_loss_clip": 0.01218445, + "auxiliary_loss_mlp": 0.01031954, + "balance_loss_clip": 1.05923402, + "balance_loss_mlp": 1.02266133, + "epoch": 0.3549570131665965, + "flos": 13516166678400.0, + "grad_norm": 2.1585429293830463, + "language_loss": 0.80656993, + "learning_rate": 2.9909917662807353e-06, + "loss": 0.82907397, + "num_input_tokens_seen": 63440855, + "step": 2952, + "time_per_iteration": 2.669902801513672 + }, + { + "auxiliary_loss_clip": 0.01210332, + "auxiliary_loss_mlp": 0.01031515, + "balance_loss_clip": 1.05984759, + "balance_loss_mlp": 1.02264023, + "epoch": 0.3550772560572356, + "flos": 20887549810560.0, + "grad_norm": 3.440864061144305, + "language_loss": 0.69222307, + "learning_rate": 2.9903150683457783e-06, + "loss": 0.71464157, + "num_input_tokens_seen": 63459400, + "step": 2953, + "time_per_iteration": 2.6321632862091064 + }, + { + "auxiliary_loss_clip": 0.0122214, + "auxiliary_loss_mlp": 0.01033968, + "balance_loss_clip": 1.05813074, + "balance_loss_mlp": 1.0240972, + "epoch": 0.3551974989478747, + "flos": 20194042947840.0, + "grad_norm": 2.023147721035065, + "language_loss": 0.64520603, + "learning_rate": 2.9896382201808126e-06, + "loss": 0.66776711, + "num_input_tokens_seen": 63476800, + "step": 2954, + "time_per_iteration": 2.6445322036743164 + }, + { + "auxiliary_loss_clip": 0.01203876, + "auxiliary_loss_mlp": 0.01038431, + "balance_loss_clip": 1.0606966, + "balance_loss_mlp": 1.02918005, + "epoch": 0.3553177418385138, + "flos": 19828831415040.0, + "grad_norm": 12.197876388087892, + "language_loss": 0.81196004, + "learning_rate": 2.988961221888516e-06, + "loss": 0.83438313, + "num_input_tokens_seen": 63493475, + "step": 2955, + "time_per_iteration": 2.6129469871520996 + }, + { + "auxiliary_loss_clip": 0.01216726, + "auxiliary_loss_mlp": 0.01033542, + "balance_loss_clip": 1.05586672, + "balance_loss_mlp": 1.02478623, + "epoch": 0.3554379847291529, + "flos": 14829132516480.0, + "grad_norm": 8.372200267008429, + "language_loss": 0.79167324, + "learning_rate": 2.988284073571589e-06, + "loss": 0.8141759, + "num_input_tokens_seen": 63509560, + "step": 2956, + "time_per_iteration": 2.619030475616455 + }, + { + "auxiliary_loss_clip": 0.01213183, + "auxiliary_loss_mlp": 0.01380237, + "balance_loss_clip": 1.05956054, + "balance_loss_mlp": 1.00048256, + "epoch": 0.355558227619792, + "flos": 20485350247680.0, + "grad_norm": 3.373226141367241, + "language_loss": 0.72421247, + "learning_rate": 2.9876067753327528e-06, + "loss": 0.75014663, + "num_input_tokens_seen": 63527290, + "step": 2957, + "time_per_iteration": 2.671865940093994 + }, + { + "auxiliary_loss_clip": 0.01214848, + "auxiliary_loss_mlp": 0.01035786, + "balance_loss_clip": 1.05887473, + "balance_loss_mlp": 1.02662456, + "epoch": 0.35567847051043106, + "flos": 37663613256960.0, + "grad_norm": 3.7391795917792408, + "language_loss": 0.80524641, + "learning_rate": 2.986929327274754e-06, + "loss": 0.82775271, + "num_input_tokens_seen": 63547870, + "step": 2958, + "time_per_iteration": 3.663823127746582 + }, + { + "auxiliary_loss_clip": 0.01209239, + "auxiliary_loss_mlp": 0.010288, + "balance_loss_clip": 1.06147361, + "balance_loss_mlp": 1.01997852, + "epoch": 0.35579871340107017, + "flos": 26943058103040.0, + "grad_norm": 1.5809244276052716, + "language_loss": 0.78762364, + "learning_rate": 2.9862517295003617e-06, + "loss": 0.81000406, + "num_input_tokens_seen": 63568285, + "step": 2959, + "time_per_iteration": 2.667668342590332 + }, + { + "auxiliary_loss_clip": 0.01221294, + "auxiliary_loss_mlp": 0.01033686, + "balance_loss_clip": 1.05268526, + "balance_loss_mlp": 1.02461457, + "epoch": 0.3559189562917093, + "flos": 28293335193600.0, + "grad_norm": 1.5810242539271575, + "language_loss": 0.72706485, + "learning_rate": 2.9855739821123654e-06, + "loss": 0.7496146, + "num_input_tokens_seen": 63589865, + "step": 2960, + "time_per_iteration": 2.7988603115081787 + }, + { + "auxiliary_loss_clip": 0.01207915, + "auxiliary_loss_mlp": 0.01032931, + "balance_loss_clip": 1.05975652, + "balance_loss_mlp": 1.02421665, + "epoch": 0.35603919918234833, + "flos": 25664063552640.0, + "grad_norm": 1.7616067676508698, + "language_loss": 0.81811666, + "learning_rate": 2.98489608521358e-06, + "loss": 0.84052515, + "num_input_tokens_seen": 63609805, + "step": 2961, + "time_per_iteration": 2.6811296939849854 + }, + { + "auxiliary_loss_clip": 0.01214027, + "auxiliary_loss_mlp": 0.01379733, + "balance_loss_clip": 1.05988479, + "balance_loss_mlp": 1.00042474, + "epoch": 0.35615944207298744, + "flos": 23000856537600.0, + "grad_norm": 2.5580935013213106, + "language_loss": 0.79572529, + "learning_rate": 2.9842180389068425e-06, + "loss": 0.8216629, + "num_input_tokens_seen": 63627115, + "step": 2962, + "time_per_iteration": 2.655668258666992 + }, + { + "auxiliary_loss_clip": 0.01129291, + "auxiliary_loss_mlp": 0.01001882, + "balance_loss_clip": 1.03203022, + "balance_loss_mlp": 1.00027263, + "epoch": 0.35627968496362655, + "flos": 68251283723520.0, + "grad_norm": 0.7632317065946312, + "language_loss": 0.59230685, + "learning_rate": 2.98353984329501e-06, + "loss": 0.61361861, + "num_input_tokens_seen": 63691460, + "step": 2963, + "time_per_iteration": 3.2662715911865234 + }, + { + "auxiliary_loss_clip": 0.01222096, + "auxiliary_loss_mlp": 0.01030964, + "balance_loss_clip": 1.0597136, + "balance_loss_mlp": 1.02164781, + "epoch": 0.3563999278542656, + "flos": 22641714403200.0, + "grad_norm": 1.5867849341729798, + "language_loss": 0.70310497, + "learning_rate": 2.982861498480965e-06, + "loss": 0.72563553, + "num_input_tokens_seen": 63713840, + "step": 2964, + "time_per_iteration": 2.7210795879364014 + }, + { + "auxiliary_loss_clip": 0.01221159, + "auxiliary_loss_mlp": 0.01035066, + "balance_loss_clip": 1.05470729, + "balance_loss_mlp": 1.02626228, + "epoch": 0.3565201707449047, + "flos": 25952533678080.0, + "grad_norm": 2.710129603076383, + "language_loss": 0.82943261, + "learning_rate": 2.9821830045676122e-06, + "loss": 0.85199487, + "num_input_tokens_seen": 63733540, + "step": 2965, + "time_per_iteration": 2.749983072280884 + }, + { + "auxiliary_loss_clip": 0.01203916, + "auxiliary_loss_mlp": 0.01034171, + "balance_loss_clip": 1.06254447, + "balance_loss_mlp": 1.02468765, + "epoch": 0.3566404136355438, + "flos": 28475725478400.0, + "grad_norm": 6.423939704682871, + "language_loss": 0.72933698, + "learning_rate": 2.9815043616578793e-06, + "loss": 0.75171781, + "num_input_tokens_seen": 63754335, + "step": 2966, + "time_per_iteration": 2.6473116874694824 + }, + { + "auxiliary_loss_clip": 0.01223028, + "auxiliary_loss_mlp": 0.01036217, + "balance_loss_clip": 1.05626845, + "balance_loss_mlp": 1.02759206, + "epoch": 0.3567606565261829, + "flos": 38363117690880.0, + "grad_norm": 1.8350523901442812, + "language_loss": 0.76978004, + "learning_rate": 2.9808255698547145e-06, + "loss": 0.79237252, + "num_input_tokens_seen": 63777135, + "step": 2967, + "time_per_iteration": 2.89092755317688 + }, + { + "auxiliary_loss_clip": 0.01212233, + "auxiliary_loss_mlp": 0.01035381, + "balance_loss_clip": 1.06243372, + "balance_loss_mlp": 1.02598691, + "epoch": 0.356880899416822, + "flos": 21981029592960.0, + "grad_norm": 2.455692594298399, + "language_loss": 0.7895261, + "learning_rate": 2.9801466292610913e-06, + "loss": 0.81200218, + "num_input_tokens_seen": 63797020, + "step": 2968, + "time_per_iteration": 2.578160524368286 + }, + { + "auxiliary_loss_clip": 0.0121416, + "auxiliary_loss_mlp": 0.01032881, + "balance_loss_clip": 1.06256306, + "balance_loss_mlp": 1.02394044, + "epoch": 0.35700114230746105, + "flos": 18989132198400.0, + "grad_norm": 2.7140281965082416, + "language_loss": 0.80972111, + "learning_rate": 2.979467539980003e-06, + "loss": 0.83219147, + "num_input_tokens_seen": 63813810, + "step": 2969, + "time_per_iteration": 2.633349895477295 + }, + { + "auxiliary_loss_clip": 0.01215039, + "auxiliary_loss_mlp": 0.01030775, + "balance_loss_clip": 1.0617274, + "balance_loss_mlp": 1.02155387, + "epoch": 0.35712138519810016, + "flos": 19756112330880.0, + "grad_norm": 1.8652258945589724, + "language_loss": 0.76497674, + "learning_rate": 2.978788302114468e-06, + "loss": 0.78743488, + "num_input_tokens_seen": 63830925, + "step": 2970, + "time_per_iteration": 2.616424322128296 + }, + { + "auxiliary_loss_clip": 0.01209658, + "auxiliary_loss_mlp": 0.01033244, + "balance_loss_clip": 1.06058979, + "balance_loss_mlp": 1.02395737, + "epoch": 0.35724162808873927, + "flos": 35183012008320.0, + "grad_norm": 2.861198459024382, + "language_loss": 0.83247125, + "learning_rate": 2.9781089157675255e-06, + "loss": 0.8549003, + "num_input_tokens_seen": 63849385, + "step": 2971, + "time_per_iteration": 2.773568630218506 + }, + { + "auxiliary_loss_clip": 0.01205832, + "auxiliary_loss_mlp": 0.01028994, + "balance_loss_clip": 1.05902147, + "balance_loss_mlp": 1.02056003, + "epoch": 0.3573618709793783, + "flos": 25556726736000.0, + "grad_norm": 1.955101473603009, + "language_loss": 0.88187349, + "learning_rate": 2.977429381042238e-06, + "loss": 0.90422171, + "num_input_tokens_seen": 63870060, + "step": 2972, + "time_per_iteration": 2.713210105895996 + }, + { + "auxiliary_loss_clip": 0.01219481, + "auxiliary_loss_mlp": 0.01036904, + "balance_loss_clip": 1.05808055, + "balance_loss_mlp": 1.02841592, + "epoch": 0.35748211387001744, + "flos": 29132352051840.0, + "grad_norm": 2.794298601790641, + "language_loss": 0.88324726, + "learning_rate": 2.9767496980416913e-06, + "loss": 0.90581113, + "num_input_tokens_seen": 63889355, + "step": 2973, + "time_per_iteration": 2.7279725074768066 + }, + { + "auxiliary_loss_clip": 0.01210331, + "auxiliary_loss_mlp": 0.01031773, + "balance_loss_clip": 1.05546486, + "balance_loss_mlp": 1.02143121, + "epoch": 0.35760235676065655, + "flos": 13954169122560.0, + "grad_norm": 2.5885943583572906, + "language_loss": 0.80824387, + "learning_rate": 2.9760698668689914e-06, + "loss": 0.83066487, + "num_input_tokens_seen": 63905580, + "step": 2974, + "time_per_iteration": 2.6165881156921387 + }, + { + "auxiliary_loss_clip": 0.01211051, + "auxiliary_loss_mlp": 0.01027751, + "balance_loss_clip": 1.05940711, + "balance_loss_mlp": 1.01907849, + "epoch": 0.3577225996512956, + "flos": 44018688977280.0, + "grad_norm": 1.9262306799339317, + "language_loss": 0.71300375, + "learning_rate": 2.975389887627269e-06, + "loss": 0.73539186, + "num_input_tokens_seen": 63928180, + "step": 2975, + "time_per_iteration": 2.832486867904663 + }, + { + "auxiliary_loss_clip": 0.01234564, + "auxiliary_loss_mlp": 0.01038307, + "balance_loss_clip": 1.05900955, + "balance_loss_mlp": 1.02978921, + "epoch": 0.3578428425419347, + "flos": 17055199013760.0, + "grad_norm": 2.1325268417745895, + "language_loss": 0.90083694, + "learning_rate": 2.9747097604196764e-06, + "loss": 0.92356569, + "num_input_tokens_seen": 63944825, + "step": 2976, + "time_per_iteration": 2.693110704421997 + }, + { + "auxiliary_loss_clip": 0.01145661, + "auxiliary_loss_mlp": 0.01008773, + "balance_loss_clip": 1.02804744, + "balance_loss_mlp": 1.00736654, + "epoch": 0.3579630854325738, + "flos": 71676550707840.0, + "grad_norm": 0.6723195208136589, + "language_loss": 0.56675893, + "learning_rate": 2.9740294853493875e-06, + "loss": 0.58830327, + "num_input_tokens_seen": 64016385, + "step": 2977, + "time_per_iteration": 4.451757431030273 + }, + { + "auxiliary_loss_clip": 0.01240674, + "auxiliary_loss_mlp": 0.01026406, + "balance_loss_clip": 1.05443382, + "balance_loss_mlp": 1.0179956, + "epoch": 0.3580833283232129, + "flos": 25046651652480.0, + "grad_norm": 2.0708394314606204, + "language_loss": 0.67128843, + "learning_rate": 2.9733490625196008e-06, + "loss": 0.69395924, + "num_input_tokens_seen": 64036245, + "step": 2978, + "time_per_iteration": 2.7359511852264404 + }, + { + "auxiliary_loss_clip": 0.01215109, + "auxiliary_loss_mlp": 0.01029867, + "balance_loss_clip": 1.05367362, + "balance_loss_mlp": 1.02061057, + "epoch": 0.358203571213852, + "flos": 13953127628160.0, + "grad_norm": 3.4971889727851457, + "language_loss": 0.75649583, + "learning_rate": 2.9726684920335353e-06, + "loss": 0.77894557, + "num_input_tokens_seen": 64054110, + "step": 2979, + "time_per_iteration": 2.66110897064209 + }, + { + "auxiliary_loss_clip": 0.01203808, + "auxiliary_loss_mlp": 0.01380483, + "balance_loss_clip": 1.06082082, + "balance_loss_mlp": 1.00047946, + "epoch": 0.35832381410449105, + "flos": 20302457172480.0, + "grad_norm": 2.2258711378308185, + "language_loss": 0.8126899, + "learning_rate": 2.971987773994432e-06, + "loss": 0.83853281, + "num_input_tokens_seen": 64070295, + "step": 2980, + "time_per_iteration": 2.586182117462158 + }, + { + "auxiliary_loss_clip": 0.01199606, + "auxiliary_loss_mlp": 0.01028003, + "balance_loss_clip": 1.05612803, + "balance_loss_mlp": 1.01883006, + "epoch": 0.35844405699513016, + "flos": 16983234115200.0, + "grad_norm": 3.3574923238138323, + "language_loss": 0.83085644, + "learning_rate": 2.9713069085055566e-06, + "loss": 0.85313255, + "num_input_tokens_seen": 64088605, + "step": 2981, + "time_per_iteration": 2.6178789138793945 + }, + { + "auxiliary_loss_clip": 0.01230302, + "auxiliary_loss_mlp": 0.01028478, + "balance_loss_clip": 1.058851, + "balance_loss_mlp": 1.01960886, + "epoch": 0.35856429988576927, + "flos": 23216858974080.0, + "grad_norm": 2.2169518794720946, + "language_loss": 0.78594112, + "learning_rate": 2.9706258956701958e-06, + "loss": 0.8085289, + "num_input_tokens_seen": 64108595, + "step": 2982, + "time_per_iteration": 2.7203173637390137 + }, + { + "auxiliary_loss_clip": 0.01214922, + "auxiliary_loss_mlp": 0.01032568, + "balance_loss_clip": 1.06094861, + "balance_loss_mlp": 1.02351987, + "epoch": 0.3586845427764083, + "flos": 23034576430080.0, + "grad_norm": 2.9255775448823154, + "language_loss": 0.77400357, + "learning_rate": 2.9699447355916575e-06, + "loss": 0.79647845, + "num_input_tokens_seen": 64127405, + "step": 2983, + "time_per_iteration": 2.6564908027648926 + }, + { + "auxiliary_loss_clip": 0.01199092, + "auxiliary_loss_mlp": 0.01379799, + "balance_loss_clip": 1.05836177, + "balance_loss_mlp": 1.00030541, + "epoch": 0.35880478566704743, + "flos": 20010682995840.0, + "grad_norm": 1.940044411760548, + "language_loss": 0.73861152, + "learning_rate": 2.969263428373275e-06, + "loss": 0.76440048, + "num_input_tokens_seen": 64145755, + "step": 2984, + "time_per_iteration": 3.470937967300415 + }, + { + "auxiliary_loss_clip": 0.01216865, + "auxiliary_loss_mlp": 0.01026208, + "balance_loss_clip": 1.05591035, + "balance_loss_mlp": 1.01766622, + "epoch": 0.35892502855768654, + "flos": 13699095667200.0, + "grad_norm": 2.081421552613037, + "language_loss": 0.79349744, + "learning_rate": 2.9685819741184007e-06, + "loss": 0.81592822, + "num_input_tokens_seen": 64164195, + "step": 2985, + "time_per_iteration": 2.687962293624878 + }, + { + "auxiliary_loss_clip": 0.01220199, + "auxiliary_loss_mlp": 0.01031405, + "balance_loss_clip": 1.05342031, + "balance_loss_mlp": 1.02247047, + "epoch": 0.3590452714483256, + "flos": 18114096977280.0, + "grad_norm": 3.3857988505831798, + "language_loss": 0.67846167, + "learning_rate": 2.967900372930411e-06, + "loss": 0.70097768, + "num_input_tokens_seen": 64182705, + "step": 2986, + "time_per_iteration": 2.6604442596435547 + }, + { + "auxiliary_loss_clip": 0.0120493, + "auxiliary_loss_mlp": 0.01034821, + "balance_loss_clip": 1.05271935, + "balance_loss_mlp": 1.02529645, + "epoch": 0.3591655143389647, + "flos": 17749352321280.0, + "grad_norm": 2.4611774004416547, + "language_loss": 0.79463673, + "learning_rate": 2.9672186249127046e-06, + "loss": 0.81703424, + "num_input_tokens_seen": 64202170, + "step": 2987, + "time_per_iteration": 2.6988136768341064 + }, + { + "auxiliary_loss_clip": 0.01215499, + "auxiliary_loss_mlp": 0.01030922, + "balance_loss_clip": 1.05609918, + "balance_loss_mlp": 1.0222553, + "epoch": 0.3592857572296038, + "flos": 25224409082880.0, + "grad_norm": 1.8999602275864012, + "language_loss": 0.788836, + "learning_rate": 2.9665367301687014e-06, + "loss": 0.81130016, + "num_input_tokens_seen": 64220415, + "step": 2988, + "time_per_iteration": 2.684123992919922 + }, + { + "auxiliary_loss_clip": 0.01205234, + "auxiliary_loss_mlp": 0.01023258, + "balance_loss_clip": 1.05335975, + "balance_loss_mlp": 1.01452637, + "epoch": 0.3594060001202429, + "flos": 29384408764800.0, + "grad_norm": 1.8273530139284784, + "language_loss": 0.76716238, + "learning_rate": 2.965854688801845e-06, + "loss": 0.78944731, + "num_input_tokens_seen": 64242475, + "step": 2989, + "time_per_iteration": 2.8468761444091797 + }, + { + "auxiliary_loss_clip": 0.01202661, + "auxiliary_loss_mlp": 0.01026383, + "balance_loss_clip": 1.05268013, + "balance_loss_mlp": 1.01788902, + "epoch": 0.359526243010882, + "flos": 17052900543360.0, + "grad_norm": 1.9747411224282612, + "language_loss": 0.76272619, + "learning_rate": 2.9651725009156005e-06, + "loss": 0.78501666, + "num_input_tokens_seen": 64260220, + "step": 2990, + "time_per_iteration": 2.662343978881836 + }, + { + "auxiliary_loss_clip": 0.01206077, + "auxiliary_loss_mlp": 0.01035692, + "balance_loss_clip": 1.05220866, + "balance_loss_mlp": 1.0270437, + "epoch": 0.3596464859015211, + "flos": 22965089569920.0, + "grad_norm": 1.9999128741623502, + "language_loss": 0.74079901, + "learning_rate": 2.964490166613454e-06, + "loss": 0.76321673, + "num_input_tokens_seen": 64280145, + "step": 2991, + "time_per_iteration": 2.6848042011260986 + }, + { + "auxiliary_loss_clip": 0.01101133, + "auxiliary_loss_mlp": 0.0100824, + "balance_loss_clip": 1.03076458, + "balance_loss_mlp": 1.00685716, + "epoch": 0.35976672879216015, + "flos": 54739462590720.0, + "grad_norm": 0.7644953815735397, + "language_loss": 0.5769369, + "learning_rate": 2.963807685998917e-06, + "loss": 0.59803075, + "num_input_tokens_seen": 64336010, + "step": 2992, + "time_per_iteration": 3.042863607406616 + }, + { + "auxiliary_loss_clip": 0.01238196, + "auxiliary_loss_mlp": 0.01031469, + "balance_loss_clip": 1.05261934, + "balance_loss_mlp": 1.02353597, + "epoch": 0.35988697168279926, + "flos": 43139020901760.0, + "grad_norm": 2.0708639981057395, + "language_loss": 0.77999902, + "learning_rate": 2.9631250591755196e-06, + "loss": 0.80269569, + "num_input_tokens_seen": 64358725, + "step": 2993, + "time_per_iteration": 2.9169728755950928 + }, + { + "auxiliary_loss_clip": 0.01210207, + "auxiliary_loss_mlp": 0.0102647, + "balance_loss_clip": 1.05634248, + "balance_loss_mlp": 1.01777935, + "epoch": 0.36000721457343837, + "flos": 35845600239360.0, + "grad_norm": 1.7430136116152175, + "language_loss": 0.57723713, + "learning_rate": 2.962442286246817e-06, + "loss": 0.59960389, + "num_input_tokens_seen": 64381555, + "step": 2994, + "time_per_iteration": 2.8926472663879395 + }, + { + "auxiliary_loss_clip": 0.01219717, + "auxiliary_loss_mlp": 0.01030072, + "balance_loss_clip": 1.05622017, + "balance_loss_mlp": 1.02119672, + "epoch": 0.3601274574640774, + "flos": 18291100222080.0, + "grad_norm": 1.6987619480937304, + "language_loss": 0.6934312, + "learning_rate": 2.9617593673163853e-06, + "loss": 0.71592909, + "num_input_tokens_seen": 64400375, + "step": 2995, + "time_per_iteration": 2.6507062911987305 + }, + { + "auxiliary_loss_clip": 0.01219266, + "auxiliary_loss_mlp": 0.01029154, + "balance_loss_clip": 1.05587602, + "balance_loss_mlp": 1.02066064, + "epoch": 0.36024770035471654, + "flos": 13333955961600.0, + "grad_norm": 2.1462688258519615, + "language_loss": 0.77248603, + "learning_rate": 2.9610763024878216e-06, + "loss": 0.79497021, + "num_input_tokens_seen": 64415880, + "step": 2996, + "time_per_iteration": 2.747339963912964 + }, + { + "auxiliary_loss_clip": 0.01206693, + "auxiliary_loss_mlp": 0.01024769, + "balance_loss_clip": 1.05344009, + "balance_loss_mlp": 1.01625729, + "epoch": 0.3603679432453556, + "flos": 20267013427200.0, + "grad_norm": 1.8732297215099594, + "language_loss": 0.91391587, + "learning_rate": 2.960393091864747e-06, + "loss": 0.93623042, + "num_input_tokens_seen": 64434260, + "step": 2997, + "time_per_iteration": 2.6600611209869385 + }, + { + "auxiliary_loss_clip": 0.01217969, + "auxiliary_loss_mlp": 0.01026795, + "balance_loss_clip": 1.05931473, + "balance_loss_mlp": 1.01820028, + "epoch": 0.3604881861359947, + "flos": 22451135817600.0, + "grad_norm": 1.783268323878724, + "language_loss": 0.74549031, + "learning_rate": 2.959709735550804e-06, + "loss": 0.76793802, + "num_input_tokens_seen": 64453855, + "step": 2998, + "time_per_iteration": 2.6882379055023193 + }, + { + "auxiliary_loss_clip": 0.01238725, + "auxiliary_loss_mlp": 0.01032619, + "balance_loss_clip": 1.05255616, + "balance_loss_mlp": 1.02367795, + "epoch": 0.3606084290266338, + "flos": 22054251467520.0, + "grad_norm": 2.290904150339179, + "language_loss": 0.75810468, + "learning_rate": 2.9590262336496575e-06, + "loss": 0.7808181, + "num_input_tokens_seen": 64473585, + "step": 2999, + "time_per_iteration": 2.742124557495117 + }, + { + "auxiliary_loss_clip": 0.01219264, + "auxiliary_loss_mlp": 0.01028959, + "balance_loss_clip": 1.05456543, + "balance_loss_mlp": 1.01967835, + "epoch": 0.36072867191727287, + "flos": 15632921111040.0, + "grad_norm": 1.9773328296013897, + "language_loss": 0.85340273, + "learning_rate": 2.9583425862649936e-06, + "loss": 0.87588489, + "num_input_tokens_seen": 64491720, + "step": 3000, + "time_per_iteration": 2.679565191268921 + }, + { + "auxiliary_loss_clip": 0.01202612, + "auxiliary_loss_mlp": 0.01033036, + "balance_loss_clip": 1.06083453, + "balance_loss_mlp": 1.02404213, + "epoch": 0.360848914807912, + "flos": 19677000625920.0, + "grad_norm": 2.0955498466241385, + "language_loss": 0.73659855, + "learning_rate": 2.9576587935005215e-06, + "loss": 0.75895506, + "num_input_tokens_seen": 64509800, + "step": 3001, + "time_per_iteration": 2.6291093826293945 + }, + { + "auxiliary_loss_clip": 0.01209628, + "auxiliary_loss_mlp": 0.01033379, + "balance_loss_clip": 1.05624676, + "balance_loss_mlp": 1.02372313, + "epoch": 0.3609691576985511, + "flos": 18877808972160.0, + "grad_norm": 2.236435984797161, + "language_loss": 0.71844947, + "learning_rate": 2.9569748554599713e-06, + "loss": 0.74087954, + "num_input_tokens_seen": 64525410, + "step": 3002, + "time_per_iteration": 2.6608164310455322 + }, + { + "auxiliary_loss_clip": 0.01215031, + "auxiliary_loss_mlp": 0.01032366, + "balance_loss_clip": 1.05670035, + "balance_loss_mlp": 1.02360439, + "epoch": 0.36108940058919015, + "flos": 42224088648960.0, + "grad_norm": 1.9268194828993321, + "language_loss": 0.73166764, + "learning_rate": 2.956290772247097e-06, + "loss": 0.75414157, + "num_input_tokens_seen": 64544085, + "step": 3003, + "time_per_iteration": 4.595839738845825 + }, + { + "auxiliary_loss_clip": 0.01222909, + "auxiliary_loss_mlp": 0.01028438, + "balance_loss_clip": 1.05267334, + "balance_loss_mlp": 1.01931798, + "epoch": 0.36120964347982926, + "flos": 23185150243200.0, + "grad_norm": 4.140176342340761, + "language_loss": 0.72984159, + "learning_rate": 2.9556065439656724e-06, + "loss": 0.7523551, + "num_input_tokens_seen": 64563135, + "step": 3004, + "time_per_iteration": 2.8252832889556885 + }, + { + "auxiliary_loss_clip": 0.01227975, + "auxiliary_loss_mlp": 0.01028915, + "balance_loss_clip": 1.04727602, + "balance_loss_mlp": 1.02033222, + "epoch": 0.36132988637046837, + "flos": 18113055482880.0, + "grad_norm": 2.102088417084044, + "language_loss": 0.81774241, + "learning_rate": 2.9549221707194952e-06, + "loss": 0.84031135, + "num_input_tokens_seen": 64581985, + "step": 3005, + "time_per_iteration": 2.844987154006958 + }, + { + "auxiliary_loss_clip": 0.0121144, + "auxiliary_loss_mlp": 0.01027012, + "balance_loss_clip": 1.05918503, + "balance_loss_mlp": 1.01820862, + "epoch": 0.3614501292611074, + "flos": 27813101333760.0, + "grad_norm": 2.1067148826663957, + "language_loss": 0.72826171, + "learning_rate": 2.954237652612384e-06, + "loss": 0.75064623, + "num_input_tokens_seen": 64601035, + "step": 3006, + "time_per_iteration": 2.6951444149017334 + }, + { + "auxiliary_loss_clip": 0.01215722, + "auxiliary_loss_mlp": 0.01033752, + "balance_loss_clip": 1.05683041, + "balance_loss_mlp": 1.02538967, + "epoch": 0.36157037215174653, + "flos": 22634926732800.0, + "grad_norm": 2.0817686655734913, + "language_loss": 0.84103799, + "learning_rate": 2.9535529897481796e-06, + "loss": 0.86353266, + "num_input_tokens_seen": 64618580, + "step": 3007, + "time_per_iteration": 2.659291982650757 + }, + { + "auxiliary_loss_clip": 0.01197879, + "auxiliary_loss_mlp": 0.0102989, + "balance_loss_clip": 1.05670977, + "balance_loss_mlp": 1.0213666, + "epoch": 0.36169061504238564, + "flos": 12600839376000.0, + "grad_norm": 2.3046293543707925, + "language_loss": 0.76483226, + "learning_rate": 2.9528681822307446e-06, + "loss": 0.78710997, + "num_input_tokens_seen": 64635430, + "step": 3008, + "time_per_iteration": 2.6292145252227783 + }, + { + "auxiliary_loss_clip": 0.0120856, + "auxiliary_loss_mlp": 0.01379368, + "balance_loss_clip": 1.06214333, + "balance_loss_mlp": 1.00027823, + "epoch": 0.3618108579330247, + "flos": 26684644682880.0, + "grad_norm": 2.013012196639431, + "language_loss": 0.82507747, + "learning_rate": 2.952183230163964e-06, + "loss": 0.85095674, + "num_input_tokens_seen": 64655005, + "step": 3009, + "time_per_iteration": 2.6695492267608643 + }, + { + "auxiliary_loss_clip": 0.01215145, + "auxiliary_loss_mlp": 0.01027382, + "balance_loss_clip": 1.05217624, + "balance_loss_mlp": 1.01876318, + "epoch": 0.3619311008236638, + "flos": 22817029708800.0, + "grad_norm": 2.8114714852911438, + "language_loss": 0.73436046, + "learning_rate": 2.9514981336517448e-06, + "loss": 0.75678575, + "num_input_tokens_seen": 64674775, + "step": 3010, + "time_per_iteration": 3.669426918029785 + }, + { + "auxiliary_loss_clip": 0.01205955, + "auxiliary_loss_mlp": 0.01028586, + "balance_loss_clip": 1.05936432, + "balance_loss_mlp": 1.02038431, + "epoch": 0.36205134371430286, + "flos": 25919603884800.0, + "grad_norm": 2.015410856462736, + "language_loss": 0.81333786, + "learning_rate": 2.950812892798015e-06, + "loss": 0.83568323, + "num_input_tokens_seen": 64695670, + "step": 3011, + "time_per_iteration": 2.7043893337249756 + }, + { + "auxiliary_loss_clip": 0.01222356, + "auxiliary_loss_mlp": 0.01379456, + "balance_loss_clip": 1.05158043, + "balance_loss_mlp": 1.00035918, + "epoch": 0.362171586604942, + "flos": 26139592730880.0, + "grad_norm": 1.748209598080257, + "language_loss": 0.87052345, + "learning_rate": 2.9501275077067256e-06, + "loss": 0.89654148, + "num_input_tokens_seen": 64716290, + "step": 3012, + "time_per_iteration": 2.7838168144226074 + }, + { + "auxiliary_loss_clip": 0.01220655, + "auxiliary_loss_mlp": 0.01027372, + "balance_loss_clip": 1.04677403, + "balance_loss_mlp": 1.0189383, + "epoch": 0.3622918294955811, + "flos": 28074208273920.0, + "grad_norm": 1.4251959415599755, + "language_loss": 0.88582623, + "learning_rate": 2.949441978481848e-06, + "loss": 0.90830648, + "num_input_tokens_seen": 64737190, + "step": 3013, + "time_per_iteration": 2.8330087661743164 + }, + { + "auxiliary_loss_clip": 0.01227606, + "auxiliary_loss_mlp": 0.01032586, + "balance_loss_clip": 1.05400348, + "balance_loss_mlp": 1.02288842, + "epoch": 0.36241207238622014, + "flos": 19828005402240.0, + "grad_norm": 2.6785879983426826, + "language_loss": 0.79690993, + "learning_rate": 2.9487563052273778e-06, + "loss": 0.81951183, + "num_input_tokens_seen": 64753950, + "step": 3014, + "time_per_iteration": 2.6813247203826904 + }, + { + "auxiliary_loss_clip": 0.01203941, + "auxiliary_loss_mlp": 0.01029171, + "balance_loss_clip": 1.05722106, + "balance_loss_mlp": 1.02070737, + "epoch": 0.36253231527685925, + "flos": 21397158017280.0, + "grad_norm": 1.7730297007382712, + "language_loss": 0.85487783, + "learning_rate": 2.94807048804733e-06, + "loss": 0.87720895, + "num_input_tokens_seen": 64773570, + "step": 3015, + "time_per_iteration": 2.6889312267303467 + }, + { + "auxiliary_loss_clip": 0.01225556, + "auxiliary_loss_mlp": 0.01035401, + "balance_loss_clip": 1.05229139, + "balance_loss_mlp": 1.02621627, + "epoch": 0.36265255816749836, + "flos": 18362885552640.0, + "grad_norm": 1.7239074403234642, + "language_loss": 0.89885956, + "learning_rate": 2.9473845270457434e-06, + "loss": 0.92146915, + "num_input_tokens_seen": 64790385, + "step": 3016, + "time_per_iteration": 2.704698085784912 + }, + { + "auxiliary_loss_clip": 0.01208831, + "auxiliary_loss_mlp": 0.01026309, + "balance_loss_clip": 1.05487871, + "balance_loss_mlp": 1.01775002, + "epoch": 0.3627728010581374, + "flos": 18660046769280.0, + "grad_norm": 2.191545609160639, + "language_loss": 0.70455205, + "learning_rate": 2.946698422326677e-06, + "loss": 0.72690344, + "num_input_tokens_seen": 64807845, + "step": 3017, + "time_per_iteration": 2.76499080657959 + }, + { + "auxiliary_loss_clip": 0.01228273, + "auxiliary_loss_mlp": 0.01036006, + "balance_loss_clip": 1.04950047, + "balance_loss_mlp": 1.02724981, + "epoch": 0.36289304394877653, + "flos": 27524272072320.0, + "grad_norm": 2.108329845943566, + "language_loss": 0.79556918, + "learning_rate": 2.946012173994213e-06, + "loss": 0.81821197, + "num_input_tokens_seen": 64827630, + "step": 3018, + "time_per_iteration": 2.7754719257354736 + }, + { + "auxiliary_loss_clip": 0.01201898, + "auxiliary_loss_mlp": 0.01033077, + "balance_loss_clip": 1.05839741, + "balance_loss_mlp": 1.02441049, + "epoch": 0.36301328683941564, + "flos": 34533244932480.0, + "grad_norm": 1.3810199539520007, + "language_loss": 0.67763078, + "learning_rate": 2.945325782152454e-06, + "loss": 0.69998062, + "num_input_tokens_seen": 64850665, + "step": 3019, + "time_per_iteration": 2.7825400829315186 + }, + { + "auxiliary_loss_clip": 0.01214937, + "auxiliary_loss_mlp": 0.01031478, + "balance_loss_clip": 1.05352008, + "balance_loss_mlp": 1.02341962, + "epoch": 0.3631335297300547, + "flos": 19025976574080.0, + "grad_norm": 2.2031633387516183, + "language_loss": 0.7881974, + "learning_rate": 2.9446392469055257e-06, + "loss": 0.81066155, + "num_input_tokens_seen": 64868700, + "step": 3020, + "time_per_iteration": 2.6972603797912598 + }, + { + "auxiliary_loss_clip": 0.0121943, + "auxiliary_loss_mlp": 0.01035116, + "balance_loss_clip": 1.05705833, + "balance_loss_mlp": 1.02695322, + "epoch": 0.3632537726206938, + "flos": 19536769929600.0, + "grad_norm": 1.7938943429302803, + "language_loss": 0.79990512, + "learning_rate": 2.9439525683575745e-06, + "loss": 0.82245058, + "num_input_tokens_seen": 64887620, + "step": 3021, + "time_per_iteration": 2.6853737831115723 + }, + { + "auxiliary_loss_clip": 0.01203849, + "auxiliary_loss_mlp": 0.01035634, + "balance_loss_clip": 1.06176877, + "balance_loss_mlp": 1.02524471, + "epoch": 0.3633740155113329, + "flos": 21068611292160.0, + "grad_norm": 2.1221782273872223, + "language_loss": 0.74782896, + "learning_rate": 2.9432657466127694e-06, + "loss": 0.7702238, + "num_input_tokens_seen": 64907190, + "step": 3022, + "time_per_iteration": 2.6700620651245117 + }, + { + "auxiliary_loss_clip": 0.01232298, + "auxiliary_loss_mlp": 0.01032243, + "balance_loss_clip": 1.05716968, + "balance_loss_mlp": 1.02338541, + "epoch": 0.36349425840197197, + "flos": 20298722158080.0, + "grad_norm": 2.009437575710006, + "language_loss": 0.76401913, + "learning_rate": 2.9425787817753007e-06, + "loss": 0.78666449, + "num_input_tokens_seen": 64925850, + "step": 3023, + "time_per_iteration": 2.733124017715454 + }, + { + "auxiliary_loss_clip": 0.01222919, + "auxiliary_loss_mlp": 0.0103217, + "balance_loss_clip": 1.05325711, + "balance_loss_mlp": 1.02312827, + "epoch": 0.3636145012926111, + "flos": 29716762331520.0, + "grad_norm": 1.8444883174694153, + "language_loss": 0.71295005, + "learning_rate": 2.94189167394938e-06, + "loss": 0.73550093, + "num_input_tokens_seen": 64948285, + "step": 3024, + "time_per_iteration": 2.799553155899048 + }, + { + "auxiliary_loss_clip": 0.01200262, + "auxiliary_loss_mlp": 0.01033046, + "balance_loss_clip": 1.06151474, + "balance_loss_mlp": 1.02456462, + "epoch": 0.3637347441832502, + "flos": 21431847576960.0, + "grad_norm": 1.8079217347341752, + "language_loss": 0.80998862, + "learning_rate": 2.941204423239241e-06, + "loss": 0.83232164, + "num_input_tokens_seen": 64967160, + "step": 3025, + "time_per_iteration": 2.585726022720337 + }, + { + "auxiliary_loss_clip": 0.01205903, + "auxiliary_loss_mlp": 0.01028682, + "balance_loss_clip": 1.05862451, + "balance_loss_mlp": 1.01999152, + "epoch": 0.36385498707388925, + "flos": 29533941083520.0, + "grad_norm": 1.811337199183616, + "language_loss": 0.75953597, + "learning_rate": 2.9405170297491395e-06, + "loss": 0.78188181, + "num_input_tokens_seen": 64987155, + "step": 3026, + "time_per_iteration": 2.7282350063323975 + }, + { + "auxiliary_loss_clip": 0.01230531, + "auxiliary_loss_mlp": 0.01380061, + "balance_loss_clip": 1.051687, + "balance_loss_mlp": 1.0004909, + "epoch": 0.36397522996452836, + "flos": 22236569925120.0, + "grad_norm": 1.9182226319173676, + "language_loss": 0.80413252, + "learning_rate": 2.939829493583353e-06, + "loss": 0.83023846, + "num_input_tokens_seen": 65003800, + "step": 3027, + "time_per_iteration": 2.758251428604126 + }, + { + "auxiliary_loss_clip": 0.01213276, + "auxiliary_loss_mlp": 0.01032018, + "balance_loss_clip": 1.04938161, + "balance_loss_mlp": 1.02339959, + "epoch": 0.3640954728551674, + "flos": 21506505995520.0, + "grad_norm": 2.470337136868273, + "language_loss": 0.8294782, + "learning_rate": 2.939141814846179e-06, + "loss": 0.85193115, + "num_input_tokens_seen": 65021215, + "step": 3028, + "time_per_iteration": 3.5967516899108887 + }, + { + "auxiliary_loss_clip": 0.01214235, + "auxiliary_loss_mlp": 0.01029247, + "balance_loss_clip": 1.05357957, + "balance_loss_mlp": 1.02066958, + "epoch": 0.3642157157458065, + "flos": 17712867081600.0, + "grad_norm": 2.02079311470927, + "language_loss": 0.82353336, + "learning_rate": 2.938453993641938e-06, + "loss": 0.84596819, + "num_input_tokens_seen": 65039590, + "step": 3029, + "time_per_iteration": 4.523404359817505 + }, + { + "auxiliary_loss_clip": 0.0121595, + "auxiliary_loss_mlp": 0.01032058, + "balance_loss_clip": 1.06142354, + "balance_loss_mlp": 1.02309942, + "epoch": 0.36433595863644563, + "flos": 17639537466240.0, + "grad_norm": 2.3413665870841314, + "language_loss": 0.70138633, + "learning_rate": 2.937766030074973e-06, + "loss": 0.72386646, + "num_input_tokens_seen": 65056845, + "step": 3030, + "time_per_iteration": 2.644622564315796 + }, + { + "auxiliary_loss_clip": 0.01227783, + "auxiliary_loss_mlp": 0.0102824, + "balance_loss_clip": 1.05487657, + "balance_loss_mlp": 1.0194844, + "epoch": 0.3644562015270847, + "flos": 26833279161600.0, + "grad_norm": 2.0840305515942164, + "language_loss": 0.82538283, + "learning_rate": 2.937077924249646e-06, + "loss": 0.84794307, + "num_input_tokens_seen": 65079435, + "step": 3031, + "time_per_iteration": 2.752495527267456 + }, + { + "auxiliary_loss_clip": 0.01220265, + "auxiliary_loss_mlp": 0.01027537, + "balance_loss_clip": 1.05470777, + "balance_loss_mlp": 1.01853061, + "epoch": 0.3645764444177238, + "flos": 14282715847680.0, + "grad_norm": 2.128582128405994, + "language_loss": 0.7554217, + "learning_rate": 2.9363896762703443e-06, + "loss": 0.77789974, + "num_input_tokens_seen": 65096500, + "step": 3032, + "time_per_iteration": 2.723902463912964 + }, + { + "auxiliary_loss_clip": 0.01198012, + "auxiliary_loss_mlp": 0.01028841, + "balance_loss_clip": 1.05823076, + "balance_loss_mlp": 1.01995373, + "epoch": 0.3646966873083629, + "flos": 20667489137280.0, + "grad_norm": 1.896375636907242, + "language_loss": 0.84301209, + "learning_rate": 2.9357012862414725e-06, + "loss": 0.86528063, + "num_input_tokens_seen": 65115860, + "step": 3033, + "time_per_iteration": 2.5999443531036377 + }, + { + "auxiliary_loss_clip": 0.01211497, + "auxiliary_loss_mlp": 0.01027539, + "balance_loss_clip": 1.05961514, + "balance_loss_mlp": 1.01896155, + "epoch": 0.36481693019900197, + "flos": 27782613665280.0, + "grad_norm": 2.028975818954495, + "language_loss": 0.71420306, + "learning_rate": 2.9350127542674593e-06, + "loss": 0.73659343, + "num_input_tokens_seen": 65138070, + "step": 3034, + "time_per_iteration": 2.688094139099121 + }, + { + "auxiliary_loss_clip": 0.01221269, + "auxiliary_loss_mlp": 0.01035262, + "balance_loss_clip": 1.0571208, + "balance_loss_mlp": 1.02649438, + "epoch": 0.3649371730896411, + "flos": 19712588025600.0, + "grad_norm": 2.135654674125237, + "language_loss": 0.76362532, + "learning_rate": 2.934324080452755e-06, + "loss": 0.78619063, + "num_input_tokens_seen": 65155860, + "step": 3035, + "time_per_iteration": 2.6646101474761963 + }, + { + "auxiliary_loss_clip": 0.01207292, + "auxiliary_loss_mlp": 0.0138002, + "balance_loss_clip": 1.04925525, + "balance_loss_mlp": 1.00033331, + "epoch": 0.3650574159802802, + "flos": 24750496016640.0, + "grad_norm": 1.4488565980919592, + "language_loss": 0.7822758, + "learning_rate": 2.9336352649018307e-06, + "loss": 0.80814898, + "num_input_tokens_seen": 65175930, + "step": 3036, + "time_per_iteration": 3.611687660217285 + }, + { + "auxiliary_loss_clip": 0.01217364, + "auxiliary_loss_mlp": 0.01036406, + "balance_loss_clip": 1.05753553, + "balance_loss_mlp": 1.02732205, + "epoch": 0.36517765887091924, + "flos": 32853487363200.0, + "grad_norm": 1.979380553317927, + "language_loss": 0.69982541, + "learning_rate": 2.9329463077191783e-06, + "loss": 0.72236311, + "num_input_tokens_seen": 65199305, + "step": 3037, + "time_per_iteration": 2.752635955810547 + }, + { + "auxiliary_loss_clip": 0.01230561, + "auxiliary_loss_mlp": 0.01028668, + "balance_loss_clip": 1.05213439, + "balance_loss_mlp": 1.01973295, + "epoch": 0.36529790176155835, + "flos": 20120318282880.0, + "grad_norm": 2.109298244966041, + "language_loss": 0.64101934, + "learning_rate": 2.9322572090093135e-06, + "loss": 0.66361165, + "num_input_tokens_seen": 65218010, + "step": 3038, + "time_per_iteration": 2.7669758796691895 + }, + { + "auxiliary_loss_clip": 0.01230256, + "auxiliary_loss_mlp": 0.01027837, + "balance_loss_clip": 1.05144441, + "balance_loss_mlp": 1.01890254, + "epoch": 0.36541814465219746, + "flos": 17639573379840.0, + "grad_norm": 2.7559963716826275, + "language_loss": 0.7671532, + "learning_rate": 2.9315679688767713e-06, + "loss": 0.78973413, + "num_input_tokens_seen": 65236020, + "step": 3039, + "time_per_iteration": 2.708390474319458 + }, + { + "auxiliary_loss_clip": 0.01206729, + "auxiliary_loss_mlp": 0.01036957, + "balance_loss_clip": 1.05216646, + "balance_loss_mlp": 1.02872562, + "epoch": 0.3655383875428365, + "flos": 22674356887680.0, + "grad_norm": 1.526091408856422, + "language_loss": 0.66245425, + "learning_rate": 2.9308785874261085e-06, + "loss": 0.6848911, + "num_input_tokens_seen": 65256210, + "step": 3040, + "time_per_iteration": 2.66568660736084 + }, + { + "auxiliary_loss_clip": 0.0119934, + "auxiliary_loss_mlp": 0.01034165, + "balance_loss_clip": 1.05942869, + "balance_loss_mlp": 1.02528381, + "epoch": 0.36565863043347563, + "flos": 21981173247360.0, + "grad_norm": 1.6425049813077752, + "language_loss": 0.82052064, + "learning_rate": 2.9301890647619045e-06, + "loss": 0.84285569, + "num_input_tokens_seen": 65275505, + "step": 3041, + "time_per_iteration": 2.640352487564087 + }, + { + "auxiliary_loss_clip": 0.01221886, + "auxiliary_loss_mlp": 0.01034837, + "balance_loss_clip": 1.05753493, + "balance_loss_mlp": 1.02584839, + "epoch": 0.36577887332411474, + "flos": 24827632473600.0, + "grad_norm": 1.956322830956772, + "language_loss": 0.79995131, + "learning_rate": 2.929499400988759e-06, + "loss": 0.82251847, + "num_input_tokens_seen": 65296665, + "step": 3042, + "time_per_iteration": 2.6638035774230957 + }, + { + "auxiliary_loss_clip": 0.01205516, + "auxiliary_loss_mlp": 0.01035258, + "balance_loss_clip": 1.05658734, + "balance_loss_mlp": 1.02595401, + "epoch": 0.3658991162147538, + "flos": 28293191539200.0, + "grad_norm": 1.8339087141252766, + "language_loss": 0.65202904, + "learning_rate": 2.9288095962112927e-06, + "loss": 0.67443675, + "num_input_tokens_seen": 65317370, + "step": 3043, + "time_per_iteration": 2.712254762649536 + }, + { + "auxiliary_loss_clip": 0.01196808, + "auxiliary_loss_mlp": 0.01032424, + "balance_loss_clip": 1.05848193, + "balance_loss_mlp": 1.02375102, + "epoch": 0.3660193591053929, + "flos": 17785550252160.0, + "grad_norm": 2.3608261359567777, + "language_loss": 0.85275406, + "learning_rate": 2.9281196505341503e-06, + "loss": 0.87504637, + "num_input_tokens_seen": 65334540, + "step": 3044, + "time_per_iteration": 2.535458564758301 + }, + { + "auxiliary_loss_clip": 0.01222969, + "auxiliary_loss_mlp": 0.01379846, + "balance_loss_clip": 1.04944944, + "balance_loss_mlp": 1.00041807, + "epoch": 0.36613960199603196, + "flos": 10342776839040.0, + "grad_norm": 2.1945950519501216, + "language_loss": 0.7834475, + "learning_rate": 2.9274295640619946e-06, + "loss": 0.80947566, + "num_input_tokens_seen": 65351670, + "step": 3045, + "time_per_iteration": 2.789767265319824 + }, + { + "auxiliary_loss_clip": 0.0121791, + "auxiliary_loss_mlp": 0.0103317, + "balance_loss_clip": 1.04935133, + "balance_loss_mlp": 1.02520084, + "epoch": 0.36625984488667107, + "flos": 19755609540480.0, + "grad_norm": 1.7207493064412525, + "language_loss": 0.78507787, + "learning_rate": 2.9267393368995103e-06, + "loss": 0.8075887, + "num_input_tokens_seen": 65370900, + "step": 3046, + "time_per_iteration": 2.678529977798462 + }, + { + "auxiliary_loss_clip": 0.01198326, + "auxiliary_loss_mlp": 0.0102944, + "balance_loss_clip": 1.05866003, + "balance_loss_mlp": 1.0206368, + "epoch": 0.3663800877773102, + "flos": 17674262939520.0, + "grad_norm": 2.246647639096411, + "language_loss": 0.74239755, + "learning_rate": 2.926048969151407e-06, + "loss": 0.76467526, + "num_input_tokens_seen": 65388185, + "step": 3047, + "time_per_iteration": 2.6350815296173096 + }, + { + "auxiliary_loss_clip": 0.0122089, + "auxiliary_loss_mlp": 0.0102849, + "balance_loss_clip": 1.05153608, + "balance_loss_mlp": 1.01971662, + "epoch": 0.36650033066794924, + "flos": 20303606407680.0, + "grad_norm": 2.529915472158864, + "language_loss": 0.68836534, + "learning_rate": 2.92535846092241e-06, + "loss": 0.71085912, + "num_input_tokens_seen": 65407200, + "step": 3048, + "time_per_iteration": 2.7587451934814453 + }, + { + "auxiliary_loss_clip": 0.01218011, + "auxiliary_loss_mlp": 0.01029965, + "balance_loss_clip": 1.05654001, + "balance_loss_mlp": 1.01985562, + "epoch": 0.36662057355858835, + "flos": 24716237420160.0, + "grad_norm": 1.982028527262054, + "language_loss": 0.8246454, + "learning_rate": 2.9246678123172704e-06, + "loss": 0.84712517, + "num_input_tokens_seen": 65427290, + "step": 3049, + "time_per_iteration": 2.684156894683838 + }, + { + "auxiliary_loss_clip": 0.01199538, + "auxiliary_loss_mlp": 0.01027767, + "balance_loss_clip": 1.05878973, + "balance_loss_mlp": 1.01855767, + "epoch": 0.36674081644922746, + "flos": 12385267902720.0, + "grad_norm": 2.727715381904093, + "language_loss": 0.74354064, + "learning_rate": 2.9239770234407596e-06, + "loss": 0.76581365, + "num_input_tokens_seen": 65445595, + "step": 3050, + "time_per_iteration": 2.5623903274536133 + }, + { + "auxiliary_loss_clip": 0.01207604, + "auxiliary_loss_mlp": 0.01031382, + "balance_loss_clip": 1.05690873, + "balance_loss_mlp": 1.02295995, + "epoch": 0.3668610593398665, + "flos": 21105922544640.0, + "grad_norm": 1.718095849557224, + "language_loss": 0.6837678, + "learning_rate": 2.9232860943976686e-06, + "loss": 0.70615768, + "num_input_tokens_seen": 65466330, + "step": 3051, + "time_per_iteration": 2.6597328186035156 + }, + { + "auxiliary_loss_clip": 0.01210269, + "auxiliary_loss_mlp": 0.01031938, + "balance_loss_clip": 1.05350733, + "balance_loss_mlp": 1.02381372, + "epoch": 0.3669813022305056, + "flos": 26758082039040.0, + "grad_norm": 1.6410055694223673, + "language_loss": 0.84182119, + "learning_rate": 2.9225950252928115e-06, + "loss": 0.86424327, + "num_input_tokens_seen": 65487180, + "step": 3052, + "time_per_iteration": 2.6542792320251465 + }, + { + "auxiliary_loss_clip": 0.01206025, + "auxiliary_loss_mlp": 0.01035958, + "balance_loss_clip": 1.05939507, + "balance_loss_mlp": 1.02681518, + "epoch": 0.36710154512114473, + "flos": 19099521671040.0, + "grad_norm": 2.428725036265573, + "language_loss": 0.81471968, + "learning_rate": 2.9219038162310217e-06, + "loss": 0.83713949, + "num_input_tokens_seen": 65505380, + "step": 3053, + "time_per_iteration": 2.6114113330841064 + }, + { + "auxiliary_loss_clip": 0.01246011, + "auxiliary_loss_mlp": 0.01379977, + "balance_loss_clip": 1.04699826, + "balance_loss_mlp": 1.00039458, + "epoch": 0.3672217880117838, + "flos": 20812029465600.0, + "grad_norm": 2.11737306688722, + "language_loss": 0.82503444, + "learning_rate": 2.921212467317157e-06, + "loss": 0.85129428, + "num_input_tokens_seen": 65524825, + "step": 3054, + "time_per_iteration": 3.8865232467651367 + }, + { + "auxiliary_loss_clip": 0.01193081, + "auxiliary_loss_mlp": 0.0102824, + "balance_loss_clip": 1.04925013, + "balance_loss_mlp": 1.01910841, + "epoch": 0.3673420309024229, + "flos": 13590394133760.0, + "grad_norm": 1.986822088334943, + "language_loss": 0.79957265, + "learning_rate": 2.920520978656093e-06, + "loss": 0.82178587, + "num_input_tokens_seen": 65541790, + "step": 3055, + "time_per_iteration": 5.1599626541137695 + }, + { + "auxiliary_loss_clip": 0.0119433, + "auxiliary_loss_mlp": 0.01379566, + "balance_loss_clip": 1.05627537, + "balance_loss_mlp": 1.00027752, + "epoch": 0.367462273793062, + "flos": 28986877969920.0, + "grad_norm": 1.8890439968492614, + "language_loss": 0.76887852, + "learning_rate": 2.919829350352729e-06, + "loss": 0.79461747, + "num_input_tokens_seen": 65563395, + "step": 3056, + "time_per_iteration": 2.662691593170166 + }, + { + "auxiliary_loss_clip": 0.01092664, + "auxiliary_loss_mlp": 0.01004284, + "balance_loss_clip": 1.02441669, + "balance_loss_mlp": 1.00281131, + "epoch": 0.36758251668370107, + "flos": 62643148346880.0, + "grad_norm": 0.7849546604015271, + "language_loss": 0.59998107, + "learning_rate": 2.919137582511983e-06, + "loss": 0.62095058, + "num_input_tokens_seen": 65619835, + "step": 3057, + "time_per_iteration": 3.1007769107818604 + }, + { + "auxiliary_loss_clip": 0.0122809, + "auxiliary_loss_mlp": 0.01035228, + "balance_loss_clip": 1.05559671, + "balance_loss_mlp": 1.02651954, + "epoch": 0.3677027595743402, + "flos": 12713886455040.0, + "grad_norm": 2.6350429896317493, + "language_loss": 0.64126408, + "learning_rate": 2.918445675238797e-06, + "loss": 0.66389728, + "num_input_tokens_seen": 65636760, + "step": 3058, + "time_per_iteration": 2.7211203575134277 + }, + { + "auxiliary_loss_clip": 0.01195477, + "auxiliary_loss_mlp": 0.01031586, + "balance_loss_clip": 1.05658865, + "balance_loss_mlp": 1.02289534, + "epoch": 0.36782300246497923, + "flos": 25046579825280.0, + "grad_norm": 1.7888714319496601, + "language_loss": 0.69591755, + "learning_rate": 2.917753628638132e-06, + "loss": 0.71818817, + "num_input_tokens_seen": 65657065, + "step": 3059, + "time_per_iteration": 2.649698495864868 + }, + { + "auxiliary_loss_clip": 0.01216958, + "auxiliary_loss_mlp": 0.01031336, + "balance_loss_clip": 1.05792797, + "balance_loss_mlp": 1.0224669, + "epoch": 0.36794324535561834, + "flos": 17419512706560.0, + "grad_norm": 2.341558854345321, + "language_loss": 0.70532715, + "learning_rate": 2.9170614428149716e-06, + "loss": 0.72781014, + "num_input_tokens_seen": 65675400, + "step": 3060, + "time_per_iteration": 2.717766284942627 + }, + { + "auxiliary_loss_clip": 0.01216643, + "auxiliary_loss_mlp": 0.01028179, + "balance_loss_clip": 1.05459917, + "balance_loss_mlp": 1.01931548, + "epoch": 0.36806348824625745, + "flos": 24089128848000.0, + "grad_norm": 2.4804360018592515, + "language_loss": 0.86791545, + "learning_rate": 2.9163691178743195e-06, + "loss": 0.89036369, + "num_input_tokens_seen": 65694050, + "step": 3061, + "time_per_iteration": 2.750016212463379 + }, + { + "auxiliary_loss_clip": 0.01202272, + "auxiliary_loss_mlp": 0.01030545, + "balance_loss_clip": 1.05632615, + "balance_loss_mlp": 1.02199221, + "epoch": 0.3681837311368965, + "flos": 20521871400960.0, + "grad_norm": 2.137001598163041, + "language_loss": 0.77532494, + "learning_rate": 2.9156766539212006e-06, + "loss": 0.79765308, + "num_input_tokens_seen": 65711695, + "step": 3062, + "time_per_iteration": 3.5848209857940674 + }, + { + "auxiliary_loss_clip": 0.0120678, + "auxiliary_loss_mlp": 0.01026537, + "balance_loss_clip": 1.05540884, + "balance_loss_mlp": 1.01769757, + "epoch": 0.3683039740275356, + "flos": 21466644877440.0, + "grad_norm": 2.0887474821606724, + "language_loss": 0.71961939, + "learning_rate": 2.9149840510606614e-06, + "loss": 0.7419526, + "num_input_tokens_seen": 65730350, + "step": 3063, + "time_per_iteration": 2.6244890689849854 + }, + { + "auxiliary_loss_clip": 0.01102477, + "auxiliary_loss_mlp": 0.01375518, + "balance_loss_clip": 1.02256453, + "balance_loss_mlp": 1.00016546, + "epoch": 0.36842421691817473, + "flos": 70380999987840.0, + "grad_norm": 1.0259811896005333, + "language_loss": 0.64181423, + "learning_rate": 2.914291309397769e-06, + "loss": 0.66659427, + "num_input_tokens_seen": 65787820, + "step": 3064, + "time_per_iteration": 3.3042213916778564 + }, + { + "auxiliary_loss_clip": 0.01219728, + "auxiliary_loss_mlp": 0.01030864, + "balance_loss_clip": 1.046731, + "balance_loss_mlp": 1.02182817, + "epoch": 0.3685444598088138, + "flos": 23331378510720.0, + "grad_norm": 2.7189074250742227, + "language_loss": 0.78483123, + "learning_rate": 2.9135984290376117e-06, + "loss": 0.80733716, + "num_input_tokens_seen": 65806685, + "step": 3065, + "time_per_iteration": 2.799553155899048 + }, + { + "auxiliary_loss_clip": 0.01233347, + "auxiliary_loss_mlp": 0.01041666, + "balance_loss_clip": 1.0486145, + "balance_loss_mlp": 1.03244567, + "epoch": 0.3686647026994529, + "flos": 23070271570560.0, + "grad_norm": 1.817434340810952, + "language_loss": 0.82533807, + "learning_rate": 2.9129054100853e-06, + "loss": 0.8480882, + "num_input_tokens_seen": 65825525, + "step": 3066, + "time_per_iteration": 2.7661945819854736 + }, + { + "auxiliary_loss_clip": 0.01214297, + "auxiliary_loss_mlp": 0.01027944, + "balance_loss_clip": 1.05232942, + "balance_loss_mlp": 1.01917577, + "epoch": 0.368784945590092, + "flos": 25119909440640.0, + "grad_norm": 1.7851609201697944, + "language_loss": 0.76295334, + "learning_rate": 2.912212252645963e-06, + "loss": 0.78537571, + "num_input_tokens_seen": 65848110, + "step": 3067, + "time_per_iteration": 2.7223129272460938 + }, + { + "auxiliary_loss_clip": 0.01209685, + "auxiliary_loss_mlp": 0.01029632, + "balance_loss_clip": 1.05592489, + "balance_loss_mlp": 1.01994073, + "epoch": 0.36890518848073106, + "flos": 18442284566400.0, + "grad_norm": 2.315202727041285, + "language_loss": 0.76380444, + "learning_rate": 2.9115189568247523e-06, + "loss": 0.78619766, + "num_input_tokens_seen": 65865670, + "step": 3068, + "time_per_iteration": 2.6105825901031494 + }, + { + "auxiliary_loss_clip": 0.01220636, + "auxiliary_loss_mlp": 0.01030007, + "balance_loss_clip": 1.05474341, + "balance_loss_mlp": 1.0216918, + "epoch": 0.36902543137137017, + "flos": 16362446336640.0, + "grad_norm": 1.9349843663223967, + "language_loss": 0.91793782, + "learning_rate": 2.910825522726841e-06, + "loss": 0.94044423, + "num_input_tokens_seen": 65883195, + "step": 3069, + "time_per_iteration": 2.774934768676758 + }, + { + "auxiliary_loss_clip": 0.0122556, + "auxiliary_loss_mlp": 0.01030462, + "balance_loss_clip": 1.05035102, + "balance_loss_mlp": 1.02183104, + "epoch": 0.3691456742620093, + "flos": 12275596702080.0, + "grad_norm": 1.8851535730395519, + "language_loss": 0.77337921, + "learning_rate": 2.9101319504574215e-06, + "loss": 0.79593945, + "num_input_tokens_seen": 65899635, + "step": 3070, + "time_per_iteration": 2.6808054447174072 + }, + { + "auxiliary_loss_clip": 0.01190119, + "auxiliary_loss_mlp": 0.01031053, + "balance_loss_clip": 1.0516175, + "balance_loss_mlp": 1.02206492, + "epoch": 0.36926591715264834, + "flos": 17786412178560.0, + "grad_norm": 1.9760250366177547, + "language_loss": 0.76261628, + "learning_rate": 2.909438240121709e-06, + "loss": 0.78482795, + "num_input_tokens_seen": 65919910, + "step": 3071, + "time_per_iteration": 2.725998640060425 + }, + { + "auxiliary_loss_clip": 0.01203198, + "auxiliary_loss_mlp": 0.01030652, + "balance_loss_clip": 1.05322075, + "balance_loss_mlp": 1.02199793, + "epoch": 0.36938616004328745, + "flos": 28948309741440.0, + "grad_norm": 1.7616011285951978, + "language_loss": 0.70090985, + "learning_rate": 2.908744391824939e-06, + "loss": 0.72324836, + "num_input_tokens_seen": 65940930, + "step": 3072, + "time_per_iteration": 2.703395128250122 + }, + { + "auxiliary_loss_clip": 0.01234822, + "auxiliary_loss_mlp": 0.0103049, + "balance_loss_clip": 1.04516888, + "balance_loss_mlp": 1.02147806, + "epoch": 0.36950640293392656, + "flos": 29205394358400.0, + "grad_norm": 1.7585813334308744, + "language_loss": 0.79118568, + "learning_rate": 2.908050405672367e-06, + "loss": 0.81383884, + "num_input_tokens_seen": 65960475, + "step": 3073, + "time_per_iteration": 2.843437910079956 + }, + { + "auxiliary_loss_clip": 0.01218297, + "auxiliary_loss_mlp": 0.01033489, + "balance_loss_clip": 1.0528667, + "balance_loss_mlp": 1.02443528, + "epoch": 0.3696266458245656, + "flos": 24827776128000.0, + "grad_norm": 1.7686613947012348, + "language_loss": 0.79547822, + "learning_rate": 2.9073562817692703e-06, + "loss": 0.81799608, + "num_input_tokens_seen": 65979160, + "step": 3074, + "time_per_iteration": 2.6913058757781982 + }, + { + "auxiliary_loss_clip": 0.01141786, + "auxiliary_loss_mlp": 0.00999269, + "balance_loss_clip": 1.02009606, + "balance_loss_mlp": 0.99776697, + "epoch": 0.3697468887152047, + "flos": 59887257264000.0, + "grad_norm": 0.7268672695115089, + "language_loss": 0.56537145, + "learning_rate": 2.9066620202209468e-06, + "loss": 0.58678204, + "num_input_tokens_seen": 66041650, + "step": 3075, + "time_per_iteration": 3.296022891998291 + }, + { + "auxiliary_loss_clip": 0.0120651, + "auxiliary_loss_mlp": 0.01029411, + "balance_loss_clip": 1.05021083, + "balance_loss_mlp": 1.02051187, + "epoch": 0.3698671316058438, + "flos": 26137581569280.0, + "grad_norm": 2.684873789304768, + "language_loss": 0.77725381, + "learning_rate": 2.905967621132716e-06, + "loss": 0.799613, + "num_input_tokens_seen": 66059260, + "step": 3076, + "time_per_iteration": 2.794832944869995 + }, + { + "auxiliary_loss_clip": 0.01216906, + "auxiliary_loss_mlp": 0.01031851, + "balance_loss_clip": 1.0533309, + "balance_loss_mlp": 1.02272534, + "epoch": 0.3699873744964829, + "flos": 24607464059520.0, + "grad_norm": 2.033297020881011, + "language_loss": 0.75169945, + "learning_rate": 2.9052730846099172e-06, + "loss": 0.77418697, + "num_input_tokens_seen": 66080605, + "step": 3077, + "time_per_iteration": 2.73144793510437 + }, + { + "auxiliary_loss_clip": 0.01112888, + "auxiliary_loss_mlp": 0.01005307, + "balance_loss_clip": 1.02069497, + "balance_loss_mlp": 1.00388861, + "epoch": 0.370107617387122, + "flos": 64885340050560.0, + "grad_norm": 0.850581472756907, + "language_loss": 0.60950625, + "learning_rate": 2.9045784107579123e-06, + "loss": 0.63068831, + "num_input_tokens_seen": 66140710, + "step": 3078, + "time_per_iteration": 3.3177599906921387 + }, + { + "auxiliary_loss_clip": 0.01197002, + "auxiliary_loss_mlp": 0.01026108, + "balance_loss_clip": 1.05855942, + "balance_loss_mlp": 1.01760268, + "epoch": 0.37022786027776106, + "flos": 15961683317760.0, + "grad_norm": 1.8344049771261015, + "language_loss": 0.67124879, + "learning_rate": 2.9038835996820807e-06, + "loss": 0.6934799, + "num_input_tokens_seen": 66158320, + "step": 3079, + "time_per_iteration": 3.4713635444641113 + }, + { + "auxiliary_loss_clip": 0.01223848, + "auxiliary_loss_mlp": 0.01032832, + "balance_loss_clip": 1.05027461, + "balance_loss_mlp": 1.0245471, + "epoch": 0.37034810316840017, + "flos": 18546927863040.0, + "grad_norm": 1.8011724766458952, + "language_loss": 0.79545879, + "learning_rate": 2.903188651487826e-06, + "loss": 0.81802559, + "num_input_tokens_seen": 66176875, + "step": 3080, + "time_per_iteration": 3.874631881713867 + }, + { + "auxiliary_loss_clip": 0.01209803, + "auxiliary_loss_mlp": 0.01030939, + "balance_loss_clip": 1.05835354, + "balance_loss_mlp": 1.0221473, + "epoch": 0.3704683460590393, + "flos": 17821927751040.0, + "grad_norm": 2.302696362416915, + "language_loss": 0.8697716, + "learning_rate": 2.902493566280571e-06, + "loss": 0.89217895, + "num_input_tokens_seen": 66194980, + "step": 3081, + "time_per_iteration": 3.5804197788238525 + }, + { + "auxiliary_loss_clip": 0.01210145, + "auxiliary_loss_mlp": 0.01029525, + "balance_loss_clip": 1.05330122, + "balance_loss_mlp": 1.02044749, + "epoch": 0.37058858894967833, + "flos": 14134081368960.0, + "grad_norm": 2.00209917727571, + "language_loss": 0.81146264, + "learning_rate": 2.9017983441657595e-06, + "loss": 0.83385938, + "num_input_tokens_seen": 66212310, + "step": 3082, + "time_per_iteration": 2.7027859687805176 + }, + { + "auxiliary_loss_clip": 0.01225047, + "auxiliary_loss_mlp": 0.01035012, + "balance_loss_clip": 1.04681182, + "balance_loss_mlp": 1.02664995, + "epoch": 0.37070883184031744, + "flos": 13954492344960.0, + "grad_norm": 2.258322122756791, + "language_loss": 0.749111, + "learning_rate": 2.9011029852488564e-06, + "loss": 0.77171165, + "num_input_tokens_seen": 66229545, + "step": 3083, + "time_per_iteration": 2.643192768096924 + }, + { + "auxiliary_loss_clip": 0.01087915, + "auxiliary_loss_mlp": 0.01001933, + "balance_loss_clip": 1.01912642, + "balance_loss_mlp": 1.00041342, + "epoch": 0.37082907473095655, + "flos": 52315419306240.0, + "grad_norm": 1.0032807560709696, + "language_loss": 0.62471688, + "learning_rate": 2.9004074896353465e-06, + "loss": 0.6456154, + "num_input_tokens_seen": 66283545, + "step": 3084, + "time_per_iteration": 3.136556386947632 + }, + { + "auxiliary_loss_clip": 0.01197498, + "auxiliary_loss_mlp": 0.01029722, + "balance_loss_clip": 1.06038666, + "balance_loss_mlp": 1.02122796, + "epoch": 0.3709493176215956, + "flos": 15998096730240.0, + "grad_norm": 1.740542808530673, + "language_loss": 0.81582665, + "learning_rate": 2.8997118574307362e-06, + "loss": 0.83809888, + "num_input_tokens_seen": 66300500, + "step": 3085, + "time_per_iteration": 2.540501117706299 + }, + { + "auxiliary_loss_clip": 0.01230486, + "auxiliary_loss_mlp": 0.01035344, + "balance_loss_clip": 1.05509484, + "balance_loss_mlp": 1.02637982, + "epoch": 0.3710695605122347, + "flos": 20959837931520.0, + "grad_norm": 2.078368841508547, + "language_loss": 0.74378216, + "learning_rate": 2.899016088740553e-06, + "loss": 0.76644045, + "num_input_tokens_seen": 66318610, + "step": 3086, + "time_per_iteration": 2.7449018955230713 + }, + { + "auxiliary_loss_clip": 0.01228906, + "auxiliary_loss_mlp": 0.01037561, + "balance_loss_clip": 1.05203068, + "balance_loss_mlp": 1.02838826, + "epoch": 0.37118980340287383, + "flos": 14355578586240.0, + "grad_norm": 1.8512764563298374, + "language_loss": 0.79576766, + "learning_rate": 2.898320183670344e-06, + "loss": 0.81843239, + "num_input_tokens_seen": 66336025, + "step": 3087, + "time_per_iteration": 2.723073720932007 + }, + { + "auxiliary_loss_clip": 0.0122887, + "auxiliary_loss_mlp": 0.01028169, + "balance_loss_clip": 1.0554204, + "balance_loss_mlp": 1.01882315, + "epoch": 0.3713100462935129, + "flos": 25885381201920.0, + "grad_norm": 2.590835455351196, + "language_loss": 0.88727462, + "learning_rate": 2.8976241423256767e-06, + "loss": 0.90984511, + "num_input_tokens_seen": 66356120, + "step": 3088, + "time_per_iteration": 3.784844160079956 + }, + { + "auxiliary_loss_clip": 0.01205843, + "auxiliary_loss_mlp": 0.01024714, + "balance_loss_clip": 1.05284989, + "balance_loss_mlp": 1.01595759, + "epoch": 0.371430289184152, + "flos": 30518934814080.0, + "grad_norm": 2.139480468942054, + "language_loss": 0.68541068, + "learning_rate": 2.896927964812142e-06, + "loss": 0.70771623, + "num_input_tokens_seen": 66376685, + "step": 3089, + "time_per_iteration": 2.827488899230957 + }, + { + "auxiliary_loss_clip": 0.01213075, + "auxiliary_loss_mlp": 0.01028099, + "balance_loss_clip": 1.05607045, + "balance_loss_mlp": 1.01929522, + "epoch": 0.37155053207479105, + "flos": 15742233175680.0, + "grad_norm": 2.762662682207077, + "language_loss": 0.74913996, + "learning_rate": 2.8962316512353465e-06, + "loss": 0.77155173, + "num_input_tokens_seen": 66394230, + "step": 3090, + "time_per_iteration": 2.6487934589385986 + }, + { + "auxiliary_loss_clip": 0.0123238, + "auxiliary_loss_mlp": 0.01030049, + "balance_loss_clip": 1.04632759, + "balance_loss_mlp": 1.02123904, + "epoch": 0.37167077496543016, + "flos": 23404061681280.0, + "grad_norm": 1.5536733701231478, + "language_loss": 0.7515732, + "learning_rate": 2.8955352017009233e-06, + "loss": 0.77419746, + "num_input_tokens_seen": 66413475, + "step": 3091, + "time_per_iteration": 2.7836415767669678 + }, + { + "auxiliary_loss_clip": 0.01211748, + "auxiliary_loss_mlp": 0.01035514, + "balance_loss_clip": 1.05535197, + "balance_loss_mlp": 1.02664495, + "epoch": 0.3717910178560693, + "flos": 22088653718400.0, + "grad_norm": 1.8957716777265006, + "language_loss": 0.77230489, + "learning_rate": 2.8948386163145212e-06, + "loss": 0.79477751, + "num_input_tokens_seen": 66432685, + "step": 3092, + "time_per_iteration": 2.6823341846466064 + }, + { + "auxiliary_loss_clip": 0.01207766, + "auxiliary_loss_mlp": 0.01031036, + "balance_loss_clip": 1.05581927, + "balance_loss_mlp": 1.02222657, + "epoch": 0.3719112607467083, + "flos": 26939969533440.0, + "grad_norm": 1.8683751034521363, + "language_loss": 0.79535943, + "learning_rate": 2.8941418951818135e-06, + "loss": 0.81774741, + "num_input_tokens_seen": 66452245, + "step": 3093, + "time_per_iteration": 2.6912686824798584 + }, + { + "auxiliary_loss_clip": 0.01218647, + "auxiliary_loss_mlp": 0.01030617, + "balance_loss_clip": 1.04917896, + "balance_loss_mlp": 1.02189112, + "epoch": 0.37203150363734744, + "flos": 12166500119040.0, + "grad_norm": 2.094441287309375, + "language_loss": 0.70856011, + "learning_rate": 2.8934450384084903e-06, + "loss": 0.73105276, + "num_input_tokens_seen": 66469760, + "step": 3094, + "time_per_iteration": 2.64748477935791 + }, + { + "auxiliary_loss_clip": 0.01203488, + "auxiliary_loss_mlp": 0.01029835, + "balance_loss_clip": 1.05131364, + "balance_loss_mlp": 1.02110314, + "epoch": 0.37215174652798655, + "flos": 23697595624320.0, + "grad_norm": 2.2199891439800306, + "language_loss": 0.69651073, + "learning_rate": 2.8927480461002653e-06, + "loss": 0.718844, + "num_input_tokens_seen": 66489730, + "step": 3095, + "time_per_iteration": 2.699679374694824 + }, + { + "auxiliary_loss_clip": 0.01210032, + "auxiliary_loss_mlp": 0.01032542, + "balance_loss_clip": 1.05246329, + "balance_loss_mlp": 1.02311254, + "epoch": 0.3722719894186256, + "flos": 17887751424000.0, + "grad_norm": 2.4016963712697548, + "language_loss": 0.8603667, + "learning_rate": 2.892050918362872e-06, + "loss": 0.88279247, + "num_input_tokens_seen": 66504785, + "step": 3096, + "time_per_iteration": 2.652116298675537 + }, + { + "auxiliary_loss_clip": 0.01153199, + "auxiliary_loss_mlp": 0.01004318, + "balance_loss_clip": 1.01666164, + "balance_loss_mlp": 1.00285149, + "epoch": 0.3723922323092647, + "flos": 62419891363200.0, + "grad_norm": 0.8408110829337274, + "language_loss": 0.55876541, + "learning_rate": 2.8913536553020626e-06, + "loss": 0.58034062, + "num_input_tokens_seen": 66558840, + "step": 3097, + "time_per_iteration": 3.5406572818756104 + }, + { + "auxiliary_loss_clip": 0.01215496, + "auxiliary_loss_mlp": 0.01039181, + "balance_loss_clip": 1.04663491, + "balance_loss_mlp": 1.02997231, + "epoch": 0.3725124751999038, + "flos": 23039747988480.0, + "grad_norm": 1.9456129516593694, + "language_loss": 0.85201675, + "learning_rate": 2.8906562570236137e-06, + "loss": 0.87456352, + "num_input_tokens_seen": 66576750, + "step": 3098, + "time_per_iteration": 2.9581682682037354 + }, + { + "auxiliary_loss_clip": 0.01229584, + "auxiliary_loss_mlp": 0.01026124, + "balance_loss_clip": 1.04371786, + "balance_loss_mlp": 1.01760626, + "epoch": 0.3726327180905429, + "flos": 20920551431040.0, + "grad_norm": 1.5576187018870153, + "language_loss": 0.76785111, + "learning_rate": 2.889958723633318e-06, + "loss": 0.79040819, + "num_input_tokens_seen": 66595690, + "step": 3099, + "time_per_iteration": 2.865122079849243 + }, + { + "auxiliary_loss_clip": 0.01222647, + "auxiliary_loss_mlp": 0.01028146, + "balance_loss_clip": 1.05444598, + "balance_loss_mlp": 1.01946139, + "epoch": 0.372752960981182, + "flos": 30592156688640.0, + "grad_norm": 1.596638888824797, + "language_loss": 0.73854208, + "learning_rate": 2.889261055236992e-06, + "loss": 0.76104999, + "num_input_tokens_seen": 66617905, + "step": 3100, + "time_per_iteration": 2.7906506061553955 + }, + { + "auxiliary_loss_clip": 0.01208834, + "auxiliary_loss_mlp": 0.0102983, + "balance_loss_clip": 1.05421913, + "balance_loss_mlp": 1.02096725, + "epoch": 0.3728732038718211, + "flos": 25116749043840.0, + "grad_norm": 1.768075872693198, + "language_loss": 0.82635278, + "learning_rate": 2.8885632519404704e-06, + "loss": 0.84873939, + "num_input_tokens_seen": 66638175, + "step": 3101, + "time_per_iteration": 2.7195184230804443 + }, + { + "auxiliary_loss_clip": 0.01210057, + "auxiliary_loss_mlp": 0.01031101, + "balance_loss_clip": 1.05395579, + "balance_loss_mlp": 1.02263761, + "epoch": 0.37299344676246016, + "flos": 25302048330240.0, + "grad_norm": 2.0584879752442187, + "language_loss": 0.76107973, + "learning_rate": 2.8878653138496107e-06, + "loss": 0.78349137, + "num_input_tokens_seen": 66658670, + "step": 3102, + "time_per_iteration": 2.7574400901794434 + }, + { + "auxiliary_loss_clip": 0.01230655, + "auxiliary_loss_mlp": 0.01033908, + "balance_loss_clip": 1.04489827, + "balance_loss_mlp": 1.02520549, + "epoch": 0.37311368965309927, + "flos": 23842531002240.0, + "grad_norm": 2.459822606251986, + "language_loss": 0.76570308, + "learning_rate": 2.8871672410702878e-06, + "loss": 0.78834873, + "num_input_tokens_seen": 66676030, + "step": 3103, + "time_per_iteration": 2.863809585571289 + }, + { + "auxiliary_loss_clip": 0.01228352, + "auxiliary_loss_mlp": 0.01031568, + "balance_loss_clip": 1.05170703, + "balance_loss_mlp": 1.02279973, + "epoch": 0.3732339325437384, + "flos": 25811943845760.0, + "grad_norm": 1.8084828982006744, + "language_loss": 0.82417399, + "learning_rate": 2.8864690337084008e-06, + "loss": 0.84677321, + "num_input_tokens_seen": 66695305, + "step": 3104, + "time_per_iteration": 2.7619926929473877 + }, + { + "auxiliary_loss_clip": 0.01195676, + "auxiliary_loss_mlp": 0.01035825, + "balance_loss_clip": 1.05307758, + "balance_loss_mlp": 1.02656233, + "epoch": 0.37335417543437743, + "flos": 26208433146240.0, + "grad_norm": 1.7824800752140513, + "language_loss": 0.78462636, + "learning_rate": 2.885770691869866e-06, + "loss": 0.80694139, + "num_input_tokens_seen": 66716185, + "step": 3105, + "time_per_iteration": 3.715226173400879 + }, + { + "auxiliary_loss_clip": 0.01198957, + "auxiliary_loss_mlp": 0.01028519, + "balance_loss_clip": 1.05451405, + "balance_loss_mlp": 1.01982903, + "epoch": 0.37347441832501654, + "flos": 24023879792640.0, + "grad_norm": 2.1781561604156683, + "language_loss": 0.74496341, + "learning_rate": 2.8850722156606207e-06, + "loss": 0.7672382, + "num_input_tokens_seen": 66734575, + "step": 3106, + "time_per_iteration": 3.869501829147339 + }, + { + "auxiliary_loss_clip": 0.01192655, + "auxiliary_loss_mlp": 0.01029928, + "balance_loss_clip": 1.05139256, + "balance_loss_mlp": 1.0215832, + "epoch": 0.3735946612156556, + "flos": 19714922409600.0, + "grad_norm": 1.8209471601841976, + "language_loss": 0.67062837, + "learning_rate": 2.8843736051866252e-06, + "loss": 0.69285429, + "num_input_tokens_seen": 66753500, + "step": 3107, + "time_per_iteration": 3.510420799255371 + }, + { + "auxiliary_loss_clip": 0.01219073, + "auxiliary_loss_mlp": 0.0137959, + "balance_loss_clip": 1.04758143, + "balance_loss_mlp": 1.00030994, + "epoch": 0.3737149041062947, + "flos": 23039604334080.0, + "grad_norm": 1.7288311337745108, + "language_loss": 0.69305199, + "learning_rate": 2.8836748605538557e-06, + "loss": 0.71903861, + "num_input_tokens_seen": 66775140, + "step": 3108, + "time_per_iteration": 2.831760883331299 + }, + { + "auxiliary_loss_clip": 0.01215378, + "auxiliary_loss_mlp": 0.01029086, + "balance_loss_clip": 1.05433452, + "balance_loss_mlp": 1.02067018, + "epoch": 0.3738351469969338, + "flos": 34678108483200.0, + "grad_norm": 2.1724860438294584, + "language_loss": 0.63463771, + "learning_rate": 2.882975981868313e-06, + "loss": 0.65708232, + "num_input_tokens_seen": 66795525, + "step": 3109, + "time_per_iteration": 2.7979822158813477 + }, + { + "auxiliary_loss_clip": 0.0120357, + "auxiliary_loss_mlp": 0.01027255, + "balance_loss_clip": 1.05811191, + "balance_loss_mlp": 1.01845765, + "epoch": 0.3739553898875729, + "flos": 43507967448960.0, + "grad_norm": 2.478648766015239, + "language_loss": 0.68718255, + "learning_rate": 2.882276969236016e-06, + "loss": 0.70949078, + "num_input_tokens_seen": 66816885, + "step": 3110, + "time_per_iteration": 2.8665482997894287 + }, + { + "auxiliary_loss_clip": 0.01204424, + "auxiliary_loss_mlp": 0.01035232, + "balance_loss_clip": 1.05164301, + "balance_loss_mlp": 1.02654171, + "epoch": 0.374075632778212, + "flos": 12856487448960.0, + "grad_norm": 9.180712469145568, + "language_loss": 0.76220572, + "learning_rate": 2.881577822763005e-06, + "loss": 0.78460228, + "num_input_tokens_seen": 66834835, + "step": 3111, + "time_per_iteration": 2.6654880046844482 + }, + { + "auxiliary_loss_clip": 0.01201305, + "auxiliary_loss_mlp": 0.01033142, + "balance_loss_clip": 1.05369127, + "balance_loss_mlp": 1.0245595, + "epoch": 0.3741958756688511, + "flos": 26024031699840.0, + "grad_norm": 1.8570869207033884, + "language_loss": 0.87433445, + "learning_rate": 2.880878542555338e-06, + "loss": 0.89667886, + "num_input_tokens_seen": 66852600, + "step": 3112, + "time_per_iteration": 2.6336100101470947 + }, + { + "auxiliary_loss_clip": 0.01197498, + "auxiliary_loss_mlp": 0.01033723, + "balance_loss_clip": 1.05777335, + "balance_loss_mlp": 1.02416277, + "epoch": 0.37431611855949015, + "flos": 21433894652160.0, + "grad_norm": 2.0010059231787065, + "language_loss": 0.80088454, + "learning_rate": 2.8801791287190976e-06, + "loss": 0.82319677, + "num_input_tokens_seen": 66870595, + "step": 3113, + "time_per_iteration": 3.501154661178589 + }, + { + "auxiliary_loss_clip": 0.01209398, + "auxiliary_loss_mlp": 0.01029586, + "balance_loss_clip": 1.05770862, + "balance_loss_mlp": 1.02084756, + "epoch": 0.37443636145012926, + "flos": 24207096090240.0, + "grad_norm": 2.6829741536939364, + "language_loss": 0.86035353, + "learning_rate": 2.8794795813603817e-06, + "loss": 0.88274336, + "num_input_tokens_seen": 66886060, + "step": 3114, + "time_per_iteration": 2.6614937782287598 + }, + { + "auxiliary_loss_clip": 0.01208976, + "auxiliary_loss_mlp": 0.01031899, + "balance_loss_clip": 1.05686843, + "balance_loss_mlp": 1.02335179, + "epoch": 0.3745566043407684, + "flos": 15378601841280.0, + "grad_norm": 5.576975157785606, + "language_loss": 0.8182283, + "learning_rate": 2.878779900585314e-06, + "loss": 0.84063709, + "num_input_tokens_seen": 66903900, + "step": 3115, + "time_per_iteration": 2.639523983001709 + }, + { + "auxiliary_loss_clip": 0.01216217, + "auxiliary_loss_mlp": 0.01033439, + "balance_loss_clip": 1.05473042, + "balance_loss_mlp": 1.02493966, + "epoch": 0.37467684723140743, + "flos": 24608218245120.0, + "grad_norm": 1.514716375624781, + "language_loss": 0.75411528, + "learning_rate": 2.8780800865000336e-06, + "loss": 0.7766118, + "num_input_tokens_seen": 66925210, + "step": 3116, + "time_per_iteration": 2.67244029045105 + }, + { + "auxiliary_loss_clip": 0.01100075, + "auxiliary_loss_mlp": 0.0100252, + "balance_loss_clip": 1.01897502, + "balance_loss_mlp": 1.00102985, + "epoch": 0.37479709012204654, + "flos": 64377491610240.0, + "grad_norm": 0.9760432027096038, + "language_loss": 0.59202957, + "learning_rate": 2.877380139210702e-06, + "loss": 0.61305559, + "num_input_tokens_seen": 66983880, + "step": 3117, + "time_per_iteration": 3.182576894760132 + }, + { + "auxiliary_loss_clip": 0.01222803, + "auxiliary_loss_mlp": 0.01033288, + "balance_loss_clip": 1.05366158, + "balance_loss_mlp": 1.02361965, + "epoch": 0.37491733301268565, + "flos": 23803962773760.0, + "grad_norm": 1.7121819419397366, + "language_loss": 0.76550019, + "learning_rate": 2.876680058823501e-06, + "loss": 0.78806114, + "num_input_tokens_seen": 67004280, + "step": 3118, + "time_per_iteration": 2.7582902908325195 + }, + { + "auxiliary_loss_clip": 0.0119936, + "auxiliary_loss_mlp": 0.01030174, + "balance_loss_clip": 1.05209219, + "balance_loss_mlp": 1.0207212, + "epoch": 0.3750375759033247, + "flos": 32160950167680.0, + "grad_norm": 1.9459158872258102, + "language_loss": 0.65707123, + "learning_rate": 2.8759798454446314e-06, + "loss": 0.67936659, + "num_input_tokens_seen": 67027445, + "step": 3119, + "time_per_iteration": 2.7654337882995605 + }, + { + "auxiliary_loss_clip": 0.0120638, + "auxiliary_loss_mlp": 0.01028621, + "balance_loss_clip": 1.05597425, + "balance_loss_mlp": 1.02029467, + "epoch": 0.3751578187939638, + "flos": 23367791923200.0, + "grad_norm": 4.71782796176093, + "language_loss": 0.81307042, + "learning_rate": 2.8752794991803173e-06, + "loss": 0.83542049, + "num_input_tokens_seen": 67045130, + "step": 3120, + "time_per_iteration": 2.648648738861084 + }, + { + "auxiliary_loss_clip": 0.01210864, + "auxiliary_loss_mlp": 0.01031088, + "balance_loss_clip": 1.05647039, + "balance_loss_mlp": 1.02270746, + "epoch": 0.37527806168460287, + "flos": 14605731878400.0, + "grad_norm": 2.069433186939293, + "language_loss": 0.7538287, + "learning_rate": 2.8745790201367976e-06, + "loss": 0.77624822, + "num_input_tokens_seen": 67060885, + "step": 3121, + "time_per_iteration": 2.610501766204834 + }, + { + "auxiliary_loss_clip": 0.01196549, + "auxiliary_loss_mlp": 0.01033997, + "balance_loss_clip": 1.05655146, + "balance_loss_mlp": 1.02457976, + "epoch": 0.375398304575242, + "flos": 26390823431040.0, + "grad_norm": 2.247816545669217, + "language_loss": 0.84379262, + "learning_rate": 2.8738784084203373e-06, + "loss": 0.86609805, + "num_input_tokens_seen": 67080960, + "step": 3122, + "time_per_iteration": 2.6805648803710938 + }, + { + "auxiliary_loss_clip": 0.01201483, + "auxiliary_loss_mlp": 0.01031417, + "balance_loss_clip": 1.04966807, + "balance_loss_mlp": 1.02282846, + "epoch": 0.3755185474658811, + "flos": 22236605838720.0, + "grad_norm": 1.738128288009217, + "language_loss": 0.78852361, + "learning_rate": 2.873177664137216e-06, + "loss": 0.81085265, + "num_input_tokens_seen": 67101890, + "step": 3123, + "time_per_iteration": 2.669207811355591 + }, + { + "auxiliary_loss_clip": 0.01216333, + "auxiliary_loss_mlp": 0.0103509, + "balance_loss_clip": 1.05284619, + "balance_loss_mlp": 1.02590537, + "epoch": 0.37563879035652015, + "flos": 30812935633920.0, + "grad_norm": 1.5847876892460917, + "language_loss": 0.69231045, + "learning_rate": 2.8724767873937384e-06, + "loss": 0.7148248, + "num_input_tokens_seen": 67126010, + "step": 3124, + "time_per_iteration": 2.811251640319824 + }, + { + "auxiliary_loss_clip": 0.01214677, + "auxiliary_loss_mlp": 0.01030596, + "balance_loss_clip": 1.05637932, + "balance_loss_mlp": 1.02256083, + "epoch": 0.37575903324715926, + "flos": 20773533064320.0, + "grad_norm": 2.0359896884730637, + "language_loss": 0.87532675, + "learning_rate": 2.871775778296225e-06, + "loss": 0.89777946, + "num_input_tokens_seen": 67143100, + "step": 3125, + "time_per_iteration": 2.6499879360198975 + }, + { + "auxiliary_loss_clip": 0.01204204, + "auxiliary_loss_mlp": 0.01032858, + "balance_loss_clip": 1.05566859, + "balance_loss_mlp": 1.02302873, + "epoch": 0.37587927613779837, + "flos": 18697681244160.0, + "grad_norm": 2.6104085966409833, + "language_loss": 0.78350765, + "learning_rate": 2.8710746369510196e-06, + "loss": 0.80587822, + "num_input_tokens_seen": 67161085, + "step": 3126, + "time_per_iteration": 2.7203426361083984 + }, + { + "auxiliary_loss_clip": 0.01202894, + "auxiliary_loss_mlp": 0.01030434, + "balance_loss_clip": 1.05259371, + "balance_loss_mlp": 1.02150559, + "epoch": 0.3759995190284374, + "flos": 13624796384640.0, + "grad_norm": 3.6138466264448894, + "language_loss": 0.83536482, + "learning_rate": 2.8703733634644846e-06, + "loss": 0.85769814, + "num_input_tokens_seen": 67175840, + "step": 3127, + "time_per_iteration": 2.6543707847595215 + }, + { + "auxiliary_loss_clip": 0.01191655, + "auxiliary_loss_mlp": 0.01028186, + "balance_loss_clip": 1.05569744, + "balance_loss_mlp": 1.0199542, + "epoch": 0.37611976191907653, + "flos": 20484847457280.0, + "grad_norm": 1.8459766654393455, + "language_loss": 0.79192293, + "learning_rate": 2.869671957943002e-06, + "loss": 0.81412137, + "num_input_tokens_seen": 67194995, + "step": 3128, + "time_per_iteration": 2.60544490814209 + }, + { + "auxiliary_loss_clip": 0.01208253, + "auxiliary_loss_mlp": 0.01028725, + "balance_loss_clip": 1.05768025, + "balance_loss_mlp": 1.02014184, + "epoch": 0.37624000480971564, + "flos": 21141797253120.0, + "grad_norm": 1.94136621967365, + "language_loss": 0.74225426, + "learning_rate": 2.8689704204929747e-06, + "loss": 0.76462406, + "num_input_tokens_seen": 67214175, + "step": 3129, + "time_per_iteration": 2.6431875228881836 + }, + { + "auxiliary_loss_clip": 0.01194593, + "auxiliary_loss_mlp": 0.01031622, + "balance_loss_clip": 1.05649126, + "balance_loss_mlp": 1.02283072, + "epoch": 0.3763602477003547, + "flos": 22564470205440.0, + "grad_norm": 1.950975449365749, + "language_loss": 0.81277096, + "learning_rate": 2.8682687512208253e-06, + "loss": 0.83503318, + "num_input_tokens_seen": 67233185, + "step": 3130, + "time_per_iteration": 2.5893442630767822 + }, + { + "auxiliary_loss_clip": 0.01210959, + "auxiliary_loss_mlp": 0.01036363, + "balance_loss_clip": 1.05722773, + "balance_loss_mlp": 1.02695131, + "epoch": 0.3764804905909938, + "flos": 27526857851520.0, + "grad_norm": 3.056413508144992, + "language_loss": 0.80530334, + "learning_rate": 2.8675669502329972e-06, + "loss": 0.82777655, + "num_input_tokens_seen": 67254715, + "step": 3131, + "time_per_iteration": 4.61063551902771 + }, + { + "auxiliary_loss_clip": 0.01201444, + "auxiliary_loss_mlp": 0.01379445, + "balance_loss_clip": 1.0550133, + "balance_loss_mlp": 1.00022101, + "epoch": 0.3766007334816329, + "flos": 22528092706560.0, + "grad_norm": 2.6056037289134277, + "language_loss": 0.85713291, + "learning_rate": 2.866865017635952e-06, + "loss": 0.88294172, + "num_input_tokens_seen": 67272535, + "step": 3132, + "time_per_iteration": 2.6938600540161133 + }, + { + "auxiliary_loss_clip": 0.01220041, + "auxiliary_loss_mlp": 0.01032569, + "balance_loss_clip": 1.0556004, + "balance_loss_mlp": 1.02367043, + "epoch": 0.376720976372272, + "flos": 25957166532480.0, + "grad_norm": 2.231885891651828, + "language_loss": 0.79245031, + "learning_rate": 2.866162953536174e-06, + "loss": 0.81497645, + "num_input_tokens_seen": 67293505, + "step": 3133, + "time_per_iteration": 3.6075172424316406 + }, + { + "auxiliary_loss_clip": 0.01209485, + "auxiliary_loss_mlp": 0.01379624, + "balance_loss_clip": 1.05315721, + "balance_loss_mlp": 1.00041211, + "epoch": 0.3768412192629111, + "flos": 18041162411520.0, + "grad_norm": 1.538373816086026, + "language_loss": 0.7514106, + "learning_rate": 2.8654607580401634e-06, + "loss": 0.77730167, + "num_input_tokens_seen": 67313240, + "step": 3134, + "time_per_iteration": 2.7600905895233154 + }, + { + "auxiliary_loss_clip": 0.01098368, + "auxiliary_loss_mlp": 0.01012389, + "balance_loss_clip": 1.01938272, + "balance_loss_mlp": 1.01094663, + "epoch": 0.3769614621535502, + "flos": 62989472304000.0, + "grad_norm": 5.561818624148623, + "language_loss": 0.65196109, + "learning_rate": 2.8647584312544446e-06, + "loss": 0.67306876, + "num_input_tokens_seen": 67378445, + "step": 3135, + "time_per_iteration": 3.212693691253662 + }, + { + "auxiliary_loss_clip": 0.01215926, + "auxiliary_loss_mlp": 0.01379929, + "balance_loss_clip": 1.05107629, + "balance_loss_mlp": 1.00038874, + "epoch": 0.37708170504418925, + "flos": 23661685002240.0, + "grad_norm": 1.4138188455837544, + "language_loss": 0.85341758, + "learning_rate": 2.864055973285559e-06, + "loss": 0.87937611, + "num_input_tokens_seen": 67400445, + "step": 3136, + "time_per_iteration": 2.740952253341675 + }, + { + "auxiliary_loss_clip": 0.01199693, + "auxiliary_loss_mlp": 0.01031365, + "balance_loss_clip": 1.05199516, + "balance_loss_mlp": 1.02232325, + "epoch": 0.37720194793482836, + "flos": 24423170353920.0, + "grad_norm": 1.8539364321072698, + "language_loss": 0.86252421, + "learning_rate": 2.8633533842400698e-06, + "loss": 0.88483483, + "num_input_tokens_seen": 67420645, + "step": 3137, + "time_per_iteration": 2.665154457092285 + }, + { + "auxiliary_loss_clip": 0.01203781, + "auxiliary_loss_mlp": 0.01380493, + "balance_loss_clip": 1.05671692, + "balance_loss_mlp": 1.00044394, + "epoch": 0.3773221908254674, + "flos": 20996502739200.0, + "grad_norm": 1.7934602947169462, + "language_loss": 0.7711184, + "learning_rate": 2.862650664224558e-06, + "loss": 0.79696113, + "num_input_tokens_seen": 67439495, + "step": 3138, + "time_per_iteration": 2.6373939514160156 + }, + { + "auxiliary_loss_clip": 0.01201413, + "auxiliary_loss_mlp": 0.01028547, + "balance_loss_clip": 1.05814195, + "balance_loss_mlp": 1.02032733, + "epoch": 0.37744243371610653, + "flos": 37631724958080.0, + "grad_norm": 1.414834534719193, + "language_loss": 0.69718689, + "learning_rate": 2.861947813345627e-06, + "loss": 0.71948647, + "num_input_tokens_seen": 67462195, + "step": 3139, + "time_per_iteration": 3.677868127822876 + }, + { + "auxiliary_loss_clip": 0.01197812, + "auxiliary_loss_mlp": 0.01379969, + "balance_loss_clip": 1.05910552, + "balance_loss_mlp": 1.00046051, + "epoch": 0.37756267660674564, + "flos": 26140526484480.0, + "grad_norm": 2.6899273655189013, + "language_loss": 0.72402143, + "learning_rate": 2.8612448317098974e-06, + "loss": 0.74979925, + "num_input_tokens_seen": 67482530, + "step": 3140, + "time_per_iteration": 2.600555896759033 + }, + { + "auxiliary_loss_clip": 0.01216717, + "auxiliary_loss_mlp": 0.01379933, + "balance_loss_clip": 1.05029738, + "balance_loss_mlp": 1.00024784, + "epoch": 0.3776829194973847, + "flos": 19427888828160.0, + "grad_norm": 2.6427145594738546, + "language_loss": 0.83379906, + "learning_rate": 2.8605417194240114e-06, + "loss": 0.85976553, + "num_input_tokens_seen": 67500890, + "step": 3141, + "time_per_iteration": 2.705012321472168 + }, + { + "auxiliary_loss_clip": 0.01194724, + "auxiliary_loss_mlp": 0.01033727, + "balance_loss_clip": 1.05215073, + "balance_loss_mlp": 1.02509046, + "epoch": 0.3778031623880238, + "flos": 17382309194880.0, + "grad_norm": 2.0293065919622437, + "language_loss": 0.78492188, + "learning_rate": 2.8598384765946315e-06, + "loss": 0.80720639, + "num_input_tokens_seen": 67519545, + "step": 3142, + "time_per_iteration": 2.613521099090576 + }, + { + "auxiliary_loss_clip": 0.011926, + "auxiliary_loss_mlp": 0.01024171, + "balance_loss_clip": 1.05527329, + "balance_loss_mlp": 1.01619625, + "epoch": 0.3779234052786629, + "flos": 27125843437440.0, + "grad_norm": 1.8055107327738753, + "language_loss": 0.71250618, + "learning_rate": 2.8591351033284377e-06, + "loss": 0.73467386, + "num_input_tokens_seen": 67539275, + "step": 3143, + "time_per_iteration": 2.677924871444702 + }, + { + "auxiliary_loss_clip": 0.01206742, + "auxiliary_loss_mlp": 0.01030247, + "balance_loss_clip": 1.05548275, + "balance_loss_mlp": 1.0215503, + "epoch": 0.37804364816930197, + "flos": 19682639061120.0, + "grad_norm": 1.9760809424355046, + "language_loss": 0.83971, + "learning_rate": 2.8584315997321325e-06, + "loss": 0.86207986, + "num_input_tokens_seen": 67558280, + "step": 3144, + "time_per_iteration": 2.657400369644165 + }, + { + "auxiliary_loss_clip": 0.01194474, + "auxiliary_loss_mlp": 0.01032027, + "balance_loss_clip": 1.05626285, + "balance_loss_mlp": 1.02349806, + "epoch": 0.3781638910599411, + "flos": 22702905221760.0, + "grad_norm": 4.192924265242427, + "language_loss": 0.77774203, + "learning_rate": 2.8577279659124356e-06, + "loss": 0.80000705, + "num_input_tokens_seen": 67575955, + "step": 3145, + "time_per_iteration": 2.6126039028167725 + }, + { + "auxiliary_loss_clip": 0.01198198, + "auxiliary_loss_mlp": 0.01033603, + "balance_loss_clip": 1.0546608, + "balance_loss_mlp": 1.02565813, + "epoch": 0.3782841339505802, + "flos": 14647604158080.0, + "grad_norm": 1.953765689496666, + "language_loss": 0.83266115, + "learning_rate": 2.857024201976089e-06, + "loss": 0.85497916, + "num_input_tokens_seen": 67593515, + "step": 3146, + "time_per_iteration": 2.610896587371826 + }, + { + "auxiliary_loss_clip": 0.01205892, + "auxiliary_loss_mlp": 0.01032111, + "balance_loss_clip": 1.05247974, + "balance_loss_mlp": 1.02302146, + "epoch": 0.37840437684121925, + "flos": 32818223185920.0, + "grad_norm": 3.1028356286646823, + "language_loss": 0.73465693, + "learning_rate": 2.8563203080298516e-06, + "loss": 0.75703692, + "num_input_tokens_seen": 67614290, + "step": 3147, + "time_per_iteration": 2.7572407722473145 + }, + { + "auxiliary_loss_clip": 0.01208683, + "auxiliary_loss_mlp": 0.01379558, + "balance_loss_clip": 1.05388916, + "balance_loss_mlp": 1.00023913, + "epoch": 0.37852461973185836, + "flos": 18369206346240.0, + "grad_norm": 2.191959732224058, + "language_loss": 0.89165872, + "learning_rate": 2.855616284180505e-06, + "loss": 0.91754109, + "num_input_tokens_seen": 67631340, + "step": 3148, + "time_per_iteration": 2.5552122592926025 + }, + { + "auxiliary_loss_clip": 0.0110221, + "auxiliary_loss_mlp": 0.0101115, + "balance_loss_clip": 1.02059698, + "balance_loss_mlp": 1.00963616, + "epoch": 0.37864486262249747, + "flos": 59500680117120.0, + "grad_norm": 0.8734811399765415, + "language_loss": 0.66117382, + "learning_rate": 2.8549121305348477e-06, + "loss": 0.68230736, + "num_input_tokens_seen": 67691125, + "step": 3149, + "time_per_iteration": 3.1394248008728027 + }, + { + "auxiliary_loss_clip": 0.01202171, + "auxiliary_loss_mlp": 0.01029399, + "balance_loss_clip": 1.05467772, + "balance_loss_mlp": 1.02138555, + "epoch": 0.3787651055131365, + "flos": 23363015414400.0, + "grad_norm": 2.4622088838432767, + "language_loss": 0.83996564, + "learning_rate": 2.8542078471997006e-06, + "loss": 0.86228132, + "num_input_tokens_seen": 67708740, + "step": 3150, + "time_per_iteration": 2.561774730682373 + }, + { + "auxiliary_loss_clip": 0.01202303, + "auxiliary_loss_mlp": 0.01029539, + "balance_loss_clip": 1.05345404, + "balance_loss_mlp": 1.02114129, + "epoch": 0.37888534840377563, + "flos": 24601394661120.0, + "grad_norm": 1.7928463410341993, + "language_loss": 0.75690877, + "learning_rate": 2.8535034342819013e-06, + "loss": 0.77922726, + "num_input_tokens_seen": 67726150, + "step": 3151, + "time_per_iteration": 2.771089553833008 + }, + { + "auxiliary_loss_clip": 0.01191564, + "auxiliary_loss_mlp": 0.01029237, + "balance_loss_clip": 1.05586874, + "balance_loss_mlp": 1.021065, + "epoch": 0.37900559129441475, + "flos": 23986891762560.0, + "grad_norm": 1.583594284860671, + "language_loss": 0.72656941, + "learning_rate": 2.85279889188831e-06, + "loss": 0.74877739, + "num_input_tokens_seen": 67746525, + "step": 3152, + "time_per_iteration": 2.6237926483154297 + }, + { + "auxiliary_loss_clip": 0.01216508, + "auxiliary_loss_mlp": 0.01027406, + "balance_loss_clip": 1.04881048, + "balance_loss_mlp": 1.0186795, + "epoch": 0.3791258341850538, + "flos": 24644667571200.0, + "grad_norm": 2.1155570297410535, + "language_loss": 0.81351244, + "learning_rate": 2.852094220125805e-06, + "loss": 0.83595163, + "num_input_tokens_seen": 67766035, + "step": 3153, + "time_per_iteration": 2.6943376064300537 + }, + { + "auxiliary_loss_clip": 0.01204581, + "auxiliary_loss_mlp": 0.01028986, + "balance_loss_clip": 1.05756676, + "balance_loss_mlp": 1.02005744, + "epoch": 0.3792460770756929, + "flos": 17420841509760.0, + "grad_norm": 2.0993227846837135, + "language_loss": 0.71260059, + "learning_rate": 2.8513894191012846e-06, + "loss": 0.73493624, + "num_input_tokens_seen": 67785015, + "step": 3154, + "time_per_iteration": 2.6393773555755615 + }, + { + "auxiliary_loss_clip": 0.0119367, + "auxiliary_loss_mlp": 0.01031808, + "balance_loss_clip": 1.05624104, + "balance_loss_mlp": 1.02353477, + "epoch": 0.37936631996633197, + "flos": 24206557386240.0, + "grad_norm": 1.4835061666435083, + "language_loss": 0.78840959, + "learning_rate": 2.8506844889216664e-06, + "loss": 0.81066436, + "num_input_tokens_seen": 67804400, + "step": 3155, + "time_per_iteration": 2.5931878089904785 + }, + { + "auxiliary_loss_clip": 0.0109452, + "auxiliary_loss_mlp": 0.01001579, + "balance_loss_clip": 1.01990151, + "balance_loss_mlp": 1.00004148, + "epoch": 0.3794865628569711, + "flos": 70297114752000.0, + "grad_norm": 0.8663140639368858, + "language_loss": 0.6284824, + "learning_rate": 2.849979429693887e-06, + "loss": 0.64944333, + "num_input_tokens_seen": 67865385, + "step": 3156, + "time_per_iteration": 3.217888116836548 + }, + { + "auxiliary_loss_clip": 0.01193262, + "auxiliary_loss_mlp": 0.01030293, + "balance_loss_clip": 1.05734515, + "balance_loss_mlp": 1.021644, + "epoch": 0.3796068057476102, + "flos": 15779364860160.0, + "grad_norm": 2.0160939000843876, + "language_loss": 0.74271756, + "learning_rate": 2.8492742415249042e-06, + "loss": 0.76495308, + "num_input_tokens_seen": 67883030, + "step": 3157, + "time_per_iteration": 4.304169416427612 + }, + { + "auxiliary_loss_clip": 0.01190095, + "auxiliary_loss_mlp": 0.01031902, + "balance_loss_clip": 1.0544039, + "balance_loss_mlp": 1.02373564, + "epoch": 0.37972704863824924, + "flos": 25191694771200.0, + "grad_norm": 1.6852454742546408, + "language_loss": 0.76483583, + "learning_rate": 2.848568924521694e-06, + "loss": 0.78705579, + "num_input_tokens_seen": 67903810, + "step": 3158, + "time_per_iteration": 2.6037352085113525 + }, + { + "auxiliary_loss_clip": 0.01192646, + "auxiliary_loss_mlp": 0.01029149, + "balance_loss_clip": 1.05137467, + "balance_loss_mlp": 1.02089334, + "epoch": 0.37984729152888835, + "flos": 26210372480640.0, + "grad_norm": 1.7578777042328684, + "language_loss": 0.73358309, + "learning_rate": 2.8478634787912526e-06, + "loss": 0.75580102, + "num_input_tokens_seen": 67921865, + "step": 3159, + "time_per_iteration": 3.563333034515381 + }, + { + "auxiliary_loss_clip": 0.01202018, + "auxiliary_loss_mlp": 0.01033441, + "balance_loss_clip": 1.05548477, + "balance_loss_mlp": 1.02507842, + "epoch": 0.37996753441952746, + "flos": 25629302165760.0, + "grad_norm": 2.3318200970766663, + "language_loss": 0.76352078, + "learning_rate": 2.847157904440596e-06, + "loss": 0.78587538, + "num_input_tokens_seen": 67941595, + "step": 3160, + "time_per_iteration": 2.701539993286133 + }, + { + "auxiliary_loss_clip": 0.01200637, + "auxiliary_loss_mlp": 0.0102726, + "balance_loss_clip": 1.05550385, + "balance_loss_mlp": 1.01932693, + "epoch": 0.3800877773101665, + "flos": 20118414862080.0, + "grad_norm": 1.4913689549499662, + "language_loss": 0.74379098, + "learning_rate": 2.846452201576759e-06, + "loss": 0.76606995, + "num_input_tokens_seen": 67960970, + "step": 3161, + "time_per_iteration": 2.627950429916382 + }, + { + "auxiliary_loss_clip": 0.01111322, + "auxiliary_loss_mlp": 0.01001881, + "balance_loss_clip": 1.01797903, + "balance_loss_mlp": 1.00034368, + "epoch": 0.38020802020080563, + "flos": 63053608037760.0, + "grad_norm": 0.8757437889306352, + "language_loss": 0.62759256, + "learning_rate": 2.845746370306795e-06, + "loss": 0.64872462, + "num_input_tokens_seen": 68026160, + "step": 3162, + "time_per_iteration": 3.3316476345062256 + }, + { + "auxiliary_loss_clip": 0.0120538, + "auxiliary_loss_mlp": 0.0102756, + "balance_loss_clip": 1.05684912, + "balance_loss_mlp": 1.01938236, + "epoch": 0.38032826309144474, + "flos": 21288420570240.0, + "grad_norm": 2.187968334718799, + "language_loss": 0.78193593, + "learning_rate": 2.84504041073778e-06, + "loss": 0.80426538, + "num_input_tokens_seen": 68044575, + "step": 3163, + "time_per_iteration": 2.6210978031158447 + }, + { + "auxiliary_loss_clip": 0.0120171, + "auxiliary_loss_mlp": 0.01032876, + "balance_loss_clip": 1.05349994, + "balance_loss_mlp": 1.0242511, + "epoch": 0.3804485059820838, + "flos": 18954119416320.0, + "grad_norm": 1.7266461287878223, + "language_loss": 0.79311603, + "learning_rate": 2.844334322976806e-06, + "loss": 0.81546187, + "num_input_tokens_seen": 68064790, + "step": 3164, + "time_per_iteration": 2.704601764678955 + }, + { + "auxiliary_loss_clip": 0.01231096, + "auxiliary_loss_mlp": 0.01035083, + "balance_loss_clip": 1.05079699, + "balance_loss_mlp": 1.02593923, + "epoch": 0.3805687488727229, + "flos": 21833759831040.0, + "grad_norm": 1.9730370569173552, + "language_loss": 0.83254349, + "learning_rate": 2.8436281071309866e-06, + "loss": 0.85520524, + "num_input_tokens_seen": 68083330, + "step": 3165, + "time_per_iteration": 3.6308488845825195 + }, + { + "auxiliary_loss_clip": 0.0113638, + "auxiliary_loss_mlp": 0.01006765, + "balance_loss_clip": 1.01704311, + "balance_loss_mlp": 1.00533485, + "epoch": 0.380688991763362, + "flos": 58546209968640.0, + "grad_norm": 0.7252555722900837, + "language_loss": 0.52966726, + "learning_rate": 2.842921763307455e-06, + "loss": 0.5510987, + "num_input_tokens_seen": 68146140, + "step": 3166, + "time_per_iteration": 3.2981486320495605 + }, + { + "auxiliary_loss_clip": 0.01198161, + "auxiliary_loss_mlp": 0.0102649, + "balance_loss_clip": 1.04953837, + "balance_loss_mlp": 1.0180676, + "epoch": 0.38080923465400107, + "flos": 23799509487360.0, + "grad_norm": 2.2701924250549017, + "language_loss": 0.8280881, + "learning_rate": 2.842215291613361e-06, + "loss": 0.85033458, + "num_input_tokens_seen": 68164520, + "step": 3167, + "time_per_iteration": 3.00215220451355 + }, + { + "auxiliary_loss_clip": 0.01148799, + "auxiliary_loss_mlp": 0.01001555, + "balance_loss_clip": 1.01791906, + "balance_loss_mlp": 1.0002135, + "epoch": 0.3809294775446402, + "flos": 54969866380800.0, + "grad_norm": 0.8269740961037931, + "language_loss": 0.59273547, + "learning_rate": 2.8415086921558774e-06, + "loss": 0.61423898, + "num_input_tokens_seen": 68227945, + "step": 3168, + "time_per_iteration": 3.5734503269195557 + }, + { + "auxiliary_loss_clip": 0.01190255, + "auxiliary_loss_mlp": 0.01028683, + "balance_loss_clip": 1.04691839, + "balance_loss_mlp": 1.02037442, + "epoch": 0.38104972043527924, + "flos": 24643697904000.0, + "grad_norm": 1.559747959341887, + "language_loss": 0.78814018, + "learning_rate": 2.840801965042194e-06, + "loss": 0.81032956, + "num_input_tokens_seen": 68247405, + "step": 3169, + "time_per_iteration": 2.934319496154785 + }, + { + "auxiliary_loss_clip": 0.01196014, + "auxiliary_loss_mlp": 0.01025377, + "balance_loss_clip": 1.04808021, + "balance_loss_mlp": 1.01697302, + "epoch": 0.38116996332591835, + "flos": 22856783086080.0, + "grad_norm": 2.056873017645397, + "language_loss": 0.83914423, + "learning_rate": 2.840095110379521e-06, + "loss": 0.86135817, + "num_input_tokens_seen": 68266925, + "step": 3170, + "time_per_iteration": 2.8233299255371094 + }, + { + "auxiliary_loss_clip": 0.011257, + "auxiliary_loss_mlp": 0.01004603, + "balance_loss_clip": 1.01480865, + "balance_loss_mlp": 1.00324368, + "epoch": 0.38129020621655746, + "flos": 60836160804480.0, + "grad_norm": 0.7383898389358534, + "language_loss": 0.53965569, + "learning_rate": 2.8393881282750884e-06, + "loss": 0.56095874, + "num_input_tokens_seen": 68329755, + "step": 3171, + "time_per_iteration": 3.178251266479492 + }, + { + "auxiliary_loss_clip": 0.01209218, + "auxiliary_loss_mlp": 0.01031727, + "balance_loss_clip": 1.05325079, + "balance_loss_mlp": 1.02242255, + "epoch": 0.3814104491071965, + "flos": 21648101408640.0, + "grad_norm": 2.1470228274280263, + "language_loss": 0.78715181, + "learning_rate": 2.838681018836144e-06, + "loss": 0.80956125, + "num_input_tokens_seen": 68347075, + "step": 3172, + "time_per_iteration": 2.739452362060547 + }, + { + "auxiliary_loss_clip": 0.01219714, + "auxiliary_loss_mlp": 0.01379465, + "balance_loss_clip": 1.05065346, + "balance_loss_mlp": 1.00030899, + "epoch": 0.3815306919978356, + "flos": 19099090707840.0, + "grad_norm": 2.423393738278015, + "language_loss": 0.78793764, + "learning_rate": 2.837973782169955e-06, + "loss": 0.81392944, + "num_input_tokens_seen": 68365450, + "step": 3173, + "time_per_iteration": 2.7069530487060547 + }, + { + "auxiliary_loss_clip": 0.01082518, + "auxiliary_loss_mlp": 0.01003741, + "balance_loss_clip": 1.01583147, + "balance_loss_mlp": 1.00229859, + "epoch": 0.38165093488847474, + "flos": 67067918156160.0, + "grad_norm": 0.811284654498776, + "language_loss": 0.59191084, + "learning_rate": 2.8372664183838096e-06, + "loss": 0.61277342, + "num_input_tokens_seen": 68428470, + "step": 3174, + "time_per_iteration": 3.232603073120117 + }, + { + "auxiliary_loss_clip": 0.01191356, + "auxiliary_loss_mlp": 0.01027347, + "balance_loss_clip": 1.05542076, + "balance_loss_mlp": 1.01928806, + "epoch": 0.3817711777791138, + "flos": 22341105480960.0, + "grad_norm": 6.841538241242867, + "language_loss": 0.68429482, + "learning_rate": 2.836558927585015e-06, + "loss": 0.70648193, + "num_input_tokens_seen": 68445440, + "step": 3175, + "time_per_iteration": 2.6473703384399414 + }, + { + "auxiliary_loss_clip": 0.01202004, + "auxiliary_loss_mlp": 0.01033157, + "balance_loss_clip": 1.05407012, + "balance_loss_mlp": 1.02469325, + "epoch": 0.3818914206697529, + "flos": 22820621068800.0, + "grad_norm": 1.844232826760469, + "language_loss": 0.8209638, + "learning_rate": 2.8358513098808957e-06, + "loss": 0.84331542, + "num_input_tokens_seen": 68465755, + "step": 3176, + "time_per_iteration": 2.616562604904175 + }, + { + "auxiliary_loss_clip": 0.01215691, + "auxiliary_loss_mlp": 0.01031778, + "balance_loss_clip": 1.04883695, + "balance_loss_mlp": 1.02334976, + "epoch": 0.382011663560392, + "flos": 24386074583040.0, + "grad_norm": 1.8804947656622166, + "language_loss": 0.76896477, + "learning_rate": 2.835143565378798e-06, + "loss": 0.79143947, + "num_input_tokens_seen": 68486220, + "step": 3177, + "time_per_iteration": 2.8452675342559814 + }, + { + "auxiliary_loss_clip": 0.01229589, + "auxiliary_loss_mlp": 0.01031234, + "balance_loss_clip": 1.04734516, + "balance_loss_mlp": 1.02272856, + "epoch": 0.38213190645103107, + "flos": 21981568296960.0, + "grad_norm": 2.0401627268811215, + "language_loss": 0.78247011, + "learning_rate": 2.8344356941860847e-06, + "loss": 0.80507839, + "num_input_tokens_seen": 68505850, + "step": 3178, + "time_per_iteration": 2.759265899658203 + }, + { + "auxiliary_loss_clip": 0.01211243, + "auxiliary_loss_mlp": 0.01032368, + "balance_loss_clip": 1.04999602, + "balance_loss_mlp": 1.02374887, + "epoch": 0.3822521493416702, + "flos": 35516945773440.0, + "grad_norm": 2.0378614554717407, + "language_loss": 0.65709454, + "learning_rate": 2.8337276964101403e-06, + "loss": 0.67953062, + "num_input_tokens_seen": 68526290, + "step": 3179, + "time_per_iteration": 2.854132652282715 + }, + { + "auxiliary_loss_clip": 0.01204598, + "auxiliary_loss_mlp": 0.01027218, + "balance_loss_clip": 1.05571175, + "balance_loss_mlp": 1.01886702, + "epoch": 0.3823723922323093, + "flos": 21069904181760.0, + "grad_norm": 2.497230111662095, + "language_loss": 0.76528966, + "learning_rate": 2.833019572158367e-06, + "loss": 0.78760785, + "num_input_tokens_seen": 68544725, + "step": 3180, + "time_per_iteration": 2.639859676361084 + }, + { + "auxiliary_loss_clip": 0.0120536, + "auxiliary_loss_mlp": 0.01032065, + "balance_loss_clip": 1.05195928, + "balance_loss_mlp": 1.02395892, + "epoch": 0.38249263512294834, + "flos": 19789149864960.0, + "grad_norm": 2.072809447419307, + "language_loss": 0.80022788, + "learning_rate": 2.8323113215381872e-06, + "loss": 0.82260209, + "num_input_tokens_seen": 68563070, + "step": 3181, + "time_per_iteration": 2.818930149078369 + }, + { + "auxiliary_loss_clip": 0.01212266, + "auxiliary_loss_mlp": 0.01029129, + "balance_loss_clip": 1.04888797, + "balance_loss_mlp": 1.01977086, + "epoch": 0.38261287801358745, + "flos": 21433930565760.0, + "grad_norm": 2.059289484123044, + "language_loss": 0.75981355, + "learning_rate": 2.831602944657042e-06, + "loss": 0.78222752, + "num_input_tokens_seen": 68581150, + "step": 3182, + "time_per_iteration": 2.662806749343872 + }, + { + "auxiliary_loss_clip": 0.01214827, + "auxiliary_loss_mlp": 0.01028678, + "balance_loss_clip": 1.05260026, + "balance_loss_mlp": 1.02076292, + "epoch": 0.38273312090422656, + "flos": 21981568296960.0, + "grad_norm": 3.1377393263155, + "language_loss": 0.74436617, + "learning_rate": 2.830894441622391e-06, + "loss": 0.76680124, + "num_input_tokens_seen": 68597800, + "step": 3183, + "time_per_iteration": 3.787585496902466 + }, + { + "auxiliary_loss_clip": 0.01213674, + "auxiliary_loss_mlp": 0.01379447, + "balance_loss_clip": 1.0478673, + "balance_loss_mlp": 1.00039005, + "epoch": 0.3828533637948656, + "flos": 24790895838720.0, + "grad_norm": 1.8264856343769535, + "language_loss": 0.79781753, + "learning_rate": 2.8301858125417134e-06, + "loss": 0.82374871, + "num_input_tokens_seen": 68617640, + "step": 3184, + "time_per_iteration": 4.0098912715911865 + }, + { + "auxiliary_loss_clip": 0.0120826, + "auxiliary_loss_mlp": 0.0102928, + "balance_loss_clip": 1.05379975, + "balance_loss_mlp": 1.02103639, + "epoch": 0.38297360668550473, + "flos": 22455445449600.0, + "grad_norm": 2.3886187173711684, + "language_loss": 0.74073696, + "learning_rate": 2.8294770575225082e-06, + "loss": 0.76311243, + "num_input_tokens_seen": 68637770, + "step": 3185, + "time_per_iteration": 2.706833839416504 + }, + { + "auxiliary_loss_clip": 0.01199486, + "auxiliary_loss_mlp": 0.01030525, + "balance_loss_clip": 1.05646348, + "balance_loss_mlp": 1.02238643, + "epoch": 0.3830938495761438, + "flos": 24896903852160.0, + "grad_norm": 1.6600668959392688, + "language_loss": 0.83731383, + "learning_rate": 2.828768176672293e-06, + "loss": 0.85961396, + "num_input_tokens_seen": 68656885, + "step": 3186, + "time_per_iteration": 2.6792562007904053 + }, + { + "auxiliary_loss_clip": 0.01213465, + "auxiliary_loss_mlp": 0.01030025, + "balance_loss_clip": 1.04899907, + "balance_loss_mlp": 1.02155471, + "epoch": 0.3832140924667829, + "flos": 33036236784000.0, + "grad_norm": 1.787787032298151, + "language_loss": 0.7170943, + "learning_rate": 2.8280591700986044e-06, + "loss": 0.73952925, + "num_input_tokens_seen": 68678750, + "step": 3187, + "time_per_iteration": 2.84818434715271 + }, + { + "auxiliary_loss_clip": 0.0121232, + "auxiliary_loss_mlp": 0.01025034, + "balance_loss_clip": 1.05065036, + "balance_loss_mlp": 1.01625991, + "epoch": 0.383334335357422, + "flos": 31903721896320.0, + "grad_norm": 1.8151329632163637, + "language_loss": 0.74704784, + "learning_rate": 2.827350037908999e-06, + "loss": 0.7694214, + "num_input_tokens_seen": 68698190, + "step": 3188, + "time_per_iteration": 2.768714666366577 + }, + { + "auxiliary_loss_clip": 0.01225094, + "auxiliary_loss_mlp": 0.01028987, + "balance_loss_clip": 1.05260158, + "balance_loss_mlp": 1.02019501, + "epoch": 0.38345457824806106, + "flos": 19791915212160.0, + "grad_norm": 2.1728035392132363, + "language_loss": 0.78993225, + "learning_rate": 2.8266407802110496e-06, + "loss": 0.81247306, + "num_input_tokens_seen": 68716445, + "step": 3189, + "time_per_iteration": 2.7386295795440674 + }, + { + "auxiliary_loss_clip": 0.01249192, + "auxiliary_loss_mlp": 0.0103783, + "balance_loss_clip": 1.04625368, + "balance_loss_mlp": 1.02869296, + "epoch": 0.3835748211387002, + "flos": 22419391173120.0, + "grad_norm": 2.236357488612576, + "language_loss": 0.75963795, + "learning_rate": 2.8259313971123515e-06, + "loss": 0.78250819, + "num_input_tokens_seen": 68737565, + "step": 3190, + "time_per_iteration": 2.9978091716766357 + }, + { + "auxiliary_loss_clip": 0.01196888, + "auxiliary_loss_mlp": 0.01031197, + "balance_loss_clip": 1.05531025, + "balance_loss_mlp": 1.02298331, + "epoch": 0.3836950640293393, + "flos": 25118436983040.0, + "grad_norm": 1.665522110936481, + "language_loss": 0.78135681, + "learning_rate": 2.8252218887205166e-06, + "loss": 0.80363762, + "num_input_tokens_seen": 68758255, + "step": 3191, + "time_per_iteration": 4.455960273742676 + }, + { + "auxiliary_loss_clip": 0.012386, + "auxiliary_loss_mlp": 0.01033298, + "balance_loss_clip": 1.05037737, + "balance_loss_mlp": 1.0251205, + "epoch": 0.38381530691997834, + "flos": 21799213925760.0, + "grad_norm": 1.984108133937, + "language_loss": 0.80752957, + "learning_rate": 2.824512255143178e-06, + "loss": 0.83024848, + "num_input_tokens_seen": 68777490, + "step": 3192, + "time_per_iteration": 2.896120548248291 + }, + { + "auxiliary_loss_clip": 0.01218585, + "auxiliary_loss_mlp": 0.01026259, + "balance_loss_clip": 1.05032563, + "balance_loss_mlp": 1.01805162, + "epoch": 0.38393554981061745, + "flos": 21252689516160.0, + "grad_norm": 1.6912208675529015, + "language_loss": 0.78989458, + "learning_rate": 2.8238024964879855e-06, + "loss": 0.812343, + "num_input_tokens_seen": 68798385, + "step": 3193, + "time_per_iteration": 2.664149522781372 + }, + { + "auxiliary_loss_clip": 0.01196294, + "auxiliary_loss_mlp": 0.01034347, + "balance_loss_clip": 1.05757403, + "balance_loss_mlp": 1.02551353, + "epoch": 0.38405579270125656, + "flos": 17019360218880.0, + "grad_norm": 2.2342387424303025, + "language_loss": 0.76676726, + "learning_rate": 2.8230926128626095e-06, + "loss": 0.78907371, + "num_input_tokens_seen": 68816880, + "step": 3194, + "time_per_iteration": 2.629382371902466 + }, + { + "auxiliary_loss_clip": 0.01199903, + "auxiliary_loss_mlp": 0.01028356, + "balance_loss_clip": 1.04988551, + "balance_loss_mlp": 1.01905167, + "epoch": 0.3841760355918956, + "flos": 21835375943040.0, + "grad_norm": 1.767216818413952, + "language_loss": 0.79197717, + "learning_rate": 2.822382604374738e-06, + "loss": 0.81425977, + "num_input_tokens_seen": 68835805, + "step": 3195, + "time_per_iteration": 2.6495144367218018 + }, + { + "auxiliary_loss_clip": 0.01211119, + "auxiliary_loss_mlp": 0.01027103, + "balance_loss_clip": 1.05655754, + "balance_loss_mlp": 1.01810229, + "epoch": 0.3842962784825347, + "flos": 25915114684800.0, + "grad_norm": 2.025789209947179, + "language_loss": 0.65266836, + "learning_rate": 2.8216724711320793e-06, + "loss": 0.6750505, + "num_input_tokens_seen": 68854930, + "step": 3196, + "time_per_iteration": 2.7095930576324463 + }, + { + "auxiliary_loss_clip": 0.01191176, + "auxiliary_loss_mlp": 0.0137887, + "balance_loss_clip": 1.05514359, + "balance_loss_mlp": 1.00026608, + "epoch": 0.38441652137317384, + "flos": 25337492075520.0, + "grad_norm": 1.4627721409004175, + "language_loss": 0.79494202, + "learning_rate": 2.820962213242361e-06, + "loss": 0.82064247, + "num_input_tokens_seen": 68874260, + "step": 3197, + "time_per_iteration": 2.6267213821411133 + }, + { + "auxiliary_loss_clip": 0.01200625, + "auxiliary_loss_mlp": 0.01039639, + "balance_loss_clip": 1.05789399, + "balance_loss_mlp": 1.03106833, + "epoch": 0.3845367642638129, + "flos": 18113486446080.0, + "grad_norm": 2.013414889220252, + "language_loss": 0.84190607, + "learning_rate": 2.8202518308133264e-06, + "loss": 0.86430871, + "num_input_tokens_seen": 68891535, + "step": 3198, + "time_per_iteration": 2.621551275253296 + }, + { + "auxiliary_loss_clip": 0.01194318, + "auxiliary_loss_mlp": 0.01028416, + "balance_loss_clip": 1.05560589, + "balance_loss_mlp": 1.01985633, + "epoch": 0.384657007154452, + "flos": 25228395492480.0, + "grad_norm": 1.8159084925693567, + "language_loss": 0.73649549, + "learning_rate": 2.8195413239527426e-06, + "loss": 0.75872284, + "num_input_tokens_seen": 68911275, + "step": 3199, + "time_per_iteration": 2.6561238765716553 + }, + { + "auxiliary_loss_clip": 0.01194667, + "auxiliary_loss_mlp": 0.01034318, + "balance_loss_clip": 1.05300641, + "balance_loss_mlp": 1.02592611, + "epoch": 0.38477725004509106, + "flos": 19865855358720.0, + "grad_norm": 2.0413419873240493, + "language_loss": 0.80326504, + "learning_rate": 2.8188306927683906e-06, + "loss": 0.82555485, + "num_input_tokens_seen": 68930745, + "step": 3200, + "time_per_iteration": 2.643436908721924 + }, + { + "auxiliary_loss_clip": 0.01211252, + "auxiliary_loss_mlp": 0.01024924, + "balance_loss_clip": 1.0539602, + "balance_loss_mlp": 1.01651955, + "epoch": 0.38489749293573017, + "flos": 18259391491200.0, + "grad_norm": 2.963243623989401, + "language_loss": 0.74563217, + "learning_rate": 2.818119937368074e-06, + "loss": 0.76799393, + "num_input_tokens_seen": 68949380, + "step": 3201, + "time_per_iteration": 2.7293686866760254 + }, + { + "auxiliary_loss_clip": 0.01210593, + "auxiliary_loss_mlp": 0.01035788, + "balance_loss_clip": 1.05663681, + "balance_loss_mlp": 1.02609062, + "epoch": 0.3850177358263693, + "flos": 24389163152640.0, + "grad_norm": 2.091538724574464, + "language_loss": 0.65694022, + "learning_rate": 2.817409057859613e-06, + "loss": 0.67940408, + "num_input_tokens_seen": 68968370, + "step": 3202, + "time_per_iteration": 2.697354316711426 + }, + { + "auxiliary_loss_clip": 0.01218259, + "auxiliary_loss_mlp": 0.01029064, + "balance_loss_clip": 1.04844332, + "balance_loss_mlp": 1.01989055, + "epoch": 0.38513797871700833, + "flos": 17671533505920.0, + "grad_norm": 2.7585979213597254, + "language_loss": 0.7920475, + "learning_rate": 2.8166980543508482e-06, + "loss": 0.81452072, + "num_input_tokens_seen": 68984260, + "step": 3203, + "time_per_iteration": 2.7641544342041016 + }, + { + "auxiliary_loss_clip": 0.01198128, + "auxiliary_loss_mlp": 0.01025694, + "balance_loss_clip": 1.06007683, + "balance_loss_mlp": 1.01658058, + "epoch": 0.38525822160764744, + "flos": 25739583897600.0, + "grad_norm": 2.098075823028686, + "language_loss": 0.79450035, + "learning_rate": 2.815986926949638e-06, + "loss": 0.81673861, + "num_input_tokens_seen": 69002760, + "step": 3204, + "time_per_iteration": 2.792104482650757 + }, + { + "auxiliary_loss_clip": 0.0119998, + "auxiliary_loss_mlp": 0.01032284, + "balance_loss_clip": 1.05588508, + "balance_loss_mlp": 1.02397513, + "epoch": 0.38537846449828655, + "flos": 20193647898240.0, + "grad_norm": 1.748262460079904, + "language_loss": 0.80283558, + "learning_rate": 2.8152756757638597e-06, + "loss": 0.82515812, + "num_input_tokens_seen": 69021260, + "step": 3205, + "time_per_iteration": 2.680603265762329 + }, + { + "auxiliary_loss_clip": 0.0120328, + "auxiliary_loss_mlp": 0.01033544, + "balance_loss_clip": 1.05743349, + "balance_loss_mlp": 1.02482414, + "epoch": 0.3854987073889256, + "flos": 23039352938880.0, + "grad_norm": 2.238783850913665, + "language_loss": 0.84704554, + "learning_rate": 2.8145643009014093e-06, + "loss": 0.86941385, + "num_input_tokens_seen": 69039755, + "step": 3206, + "time_per_iteration": 2.6111695766448975 + }, + { + "auxiliary_loss_clip": 0.01203245, + "auxiliary_loss_mlp": 0.01031969, + "balance_loss_clip": 1.05657792, + "balance_loss_mlp": 1.02355921, + "epoch": 0.3856189502795647, + "flos": 20190631155840.0, + "grad_norm": 2.07319432927694, + "language_loss": 0.79275346, + "learning_rate": 2.813852802470202e-06, + "loss": 0.81510556, + "num_input_tokens_seen": 69057650, + "step": 3207, + "time_per_iteration": 2.650968313217163 + }, + { + "auxiliary_loss_clip": 0.01207418, + "auxiliary_loss_mlp": 0.01032609, + "balance_loss_clip": 1.05416107, + "balance_loss_mlp": 1.02348351, + "epoch": 0.38573919317020383, + "flos": 25702631781120.0, + "grad_norm": 1.7023811177037513, + "language_loss": 0.72405553, + "learning_rate": 2.8131411805781717e-06, + "loss": 0.74645579, + "num_input_tokens_seen": 69077775, + "step": 3208, + "time_per_iteration": 2.6564629077911377 + }, + { + "auxiliary_loss_clip": 0.01213404, + "auxiliary_loss_mlp": 0.01028028, + "balance_loss_clip": 1.05596972, + "balance_loss_mlp": 1.01830685, + "epoch": 0.3858594360608429, + "flos": 29821405628160.0, + "grad_norm": 2.057703792297924, + "language_loss": 0.64054585, + "learning_rate": 2.8124294353332707e-06, + "loss": 0.66296017, + "num_input_tokens_seen": 69096450, + "step": 3209, + "time_per_iteration": 3.567134141921997 + }, + { + "auxiliary_loss_clip": 0.01227641, + "auxiliary_loss_mlp": 0.01030936, + "balance_loss_clip": 1.05548298, + "balance_loss_mlp": 1.02284789, + "epoch": 0.385979678951482, + "flos": 24790428961920.0, + "grad_norm": 1.7077860836981402, + "language_loss": 0.77600753, + "learning_rate": 2.8117175668434713e-06, + "loss": 0.79859328, + "num_input_tokens_seen": 69116110, + "step": 3210, + "time_per_iteration": 3.6338424682617188 + }, + { + "auxiliary_loss_clip": 0.01194626, + "auxiliary_loss_mlp": 0.01023674, + "balance_loss_clip": 1.05623722, + "balance_loss_mlp": 1.01439941, + "epoch": 0.3860999218421211, + "flos": 21287881866240.0, + "grad_norm": 2.420618322401946, + "language_loss": 0.70252788, + "learning_rate": 2.811005575216762e-06, + "loss": 0.72471088, + "num_input_tokens_seen": 69134825, + "step": 3211, + "time_per_iteration": 2.580296039581299 + }, + { + "auxiliary_loss_clip": 0.01210314, + "auxiliary_loss_mlp": 0.01031545, + "balance_loss_clip": 1.04910874, + "balance_loss_mlp": 1.022861, + "epoch": 0.38622016473276016, + "flos": 24536720223360.0, + "grad_norm": 1.4439889618909707, + "language_loss": 0.78956091, + "learning_rate": 2.8102934605611513e-06, + "loss": 0.81197953, + "num_input_tokens_seen": 69156460, + "step": 3212, + "time_per_iteration": 2.7879412174224854 + }, + { + "auxiliary_loss_clip": 0.01216495, + "auxiliary_loss_mlp": 0.01029713, + "balance_loss_clip": 1.05612445, + "balance_loss_mlp": 1.02146411, + "epoch": 0.3863404076233993, + "flos": 20558212986240.0, + "grad_norm": 2.451075530978201, + "language_loss": 0.67561495, + "learning_rate": 2.8095812229846665e-06, + "loss": 0.69807708, + "num_input_tokens_seen": 69176420, + "step": 3213, + "time_per_iteration": 2.6759042739868164 + }, + { + "auxiliary_loss_clip": 0.01211736, + "auxiliary_loss_mlp": 0.01031267, + "balance_loss_clip": 1.05389428, + "balance_loss_mlp": 1.022475, + "epoch": 0.3864606505140384, + "flos": 22346277039360.0, + "grad_norm": 3.791526342860516, + "language_loss": 0.69363755, + "learning_rate": 2.808868862595355e-06, + "loss": 0.71606755, + "num_input_tokens_seen": 69196665, + "step": 3214, + "time_per_iteration": 2.705984354019165 + }, + { + "auxiliary_loss_clip": 0.01207725, + "auxiliary_loss_mlp": 0.01025429, + "balance_loss_clip": 1.05701494, + "balance_loss_mlp": 1.01662564, + "epoch": 0.38658089340467744, + "flos": 25703601448320.0, + "grad_norm": 1.9876458363118084, + "language_loss": 0.79249346, + "learning_rate": 2.8081563795012795e-06, + "loss": 0.81482494, + "num_input_tokens_seen": 69216290, + "step": 3215, + "time_per_iteration": 2.6674327850341797 + }, + { + "auxiliary_loss_clip": 0.01221461, + "auxiliary_loss_mlp": 0.01029011, + "balance_loss_clip": 1.05584729, + "balance_loss_mlp": 1.01989126, + "epoch": 0.38670113629531655, + "flos": 33802534558080.0, + "grad_norm": 1.7755431006990139, + "language_loss": 0.74019599, + "learning_rate": 2.807443773810524e-06, + "loss": 0.76270068, + "num_input_tokens_seen": 69237550, + "step": 3216, + "time_per_iteration": 2.818878412246704 + }, + { + "auxiliary_loss_clip": 0.01221092, + "auxiliary_loss_mlp": 0.01035958, + "balance_loss_clip": 1.05486536, + "balance_loss_mlp": 1.02658498, + "epoch": 0.3868213791859556, + "flos": 23331522165120.0, + "grad_norm": 1.7136521616550948, + "language_loss": 0.89469945, + "learning_rate": 2.80673104563119e-06, + "loss": 0.91726995, + "num_input_tokens_seen": 69258175, + "step": 3217, + "time_per_iteration": 3.627624273300171 + }, + { + "auxiliary_loss_clip": 0.01201425, + "auxiliary_loss_mlp": 0.01026093, + "balance_loss_clip": 1.05745029, + "balance_loss_mlp": 1.01807296, + "epoch": 0.3869416220765947, + "flos": 18441530380800.0, + "grad_norm": 1.9577325864327446, + "language_loss": 0.78603983, + "learning_rate": 2.8060181950713976e-06, + "loss": 0.80831504, + "num_input_tokens_seen": 69274965, + "step": 3218, + "time_per_iteration": 2.6068930625915527 + }, + { + "auxiliary_loss_clip": 0.01216369, + "auxiliary_loss_mlp": 0.01037256, + "balance_loss_clip": 1.05099511, + "balance_loss_mlp": 1.02804148, + "epoch": 0.3870618649672338, + "flos": 15632992938240.0, + "grad_norm": 2.0208340464585524, + "language_loss": 0.81026685, + "learning_rate": 2.805305222239286e-06, + "loss": 0.83280313, + "num_input_tokens_seen": 69292220, + "step": 3219, + "time_per_iteration": 2.699192523956299 + }, + { + "auxiliary_loss_clip": 0.01210652, + "auxiliary_loss_mlp": 0.01031285, + "balance_loss_clip": 1.05520391, + "balance_loss_mlp": 1.02279162, + "epoch": 0.3871821078578729, + "flos": 23513804709120.0, + "grad_norm": 1.7365148024219286, + "language_loss": 0.7380659, + "learning_rate": 2.8045921272430118e-06, + "loss": 0.76048523, + "num_input_tokens_seen": 69311900, + "step": 3220, + "time_per_iteration": 2.6643636226654053 + }, + { + "auxiliary_loss_clip": 0.01210823, + "auxiliary_loss_mlp": 0.01037087, + "balance_loss_clip": 1.05605078, + "balance_loss_mlp": 1.02706742, + "epoch": 0.387302350748512, + "flos": 17778259791360.0, + "grad_norm": 2.2466061831301145, + "language_loss": 0.76769876, + "learning_rate": 2.803878910190753e-06, + "loss": 0.79017782, + "num_input_tokens_seen": 69328820, + "step": 3221, + "time_per_iteration": 2.651740312576294 + }, + { + "auxiliary_loss_clip": 0.01206483, + "auxiliary_loss_mlp": 0.01031047, + "balance_loss_clip": 1.05536461, + "balance_loss_mlp": 1.02260673, + "epoch": 0.3874225936391511, + "flos": 11503409097600.0, + "grad_norm": 2.43886461959067, + "language_loss": 0.81987023, + "learning_rate": 2.8031655711907017e-06, + "loss": 0.84224552, + "num_input_tokens_seen": 69342525, + "step": 3222, + "time_per_iteration": 2.5492191314697266 + }, + { + "auxiliary_loss_clip": 0.01206817, + "auxiliary_loss_mlp": 0.01031651, + "balance_loss_clip": 1.05957031, + "balance_loss_mlp": 1.02251959, + "epoch": 0.38754283652979016, + "flos": 21945154884480.0, + "grad_norm": 3.2311684481769682, + "language_loss": 0.80918193, + "learning_rate": 2.8024521103510723e-06, + "loss": 0.83156663, + "num_input_tokens_seen": 69359295, + "step": 3223, + "time_per_iteration": 2.674179792404175 + }, + { + "auxiliary_loss_clip": 0.012025, + "auxiliary_loss_mlp": 0.01034663, + "balance_loss_clip": 1.05443203, + "balance_loss_mlp": 1.02593064, + "epoch": 0.38766307942042927, + "flos": 21175984022400.0, + "grad_norm": 2.3190414462760356, + "language_loss": 0.75419068, + "learning_rate": 2.8017385277800952e-06, + "loss": 0.77656233, + "num_input_tokens_seen": 69377650, + "step": 3224, + "time_per_iteration": 2.6374351978302 + }, + { + "auxiliary_loss_clip": 0.01226689, + "auxiliary_loss_mlp": 0.01034267, + "balance_loss_clip": 1.05542123, + "balance_loss_mlp": 1.02482605, + "epoch": 0.3877833223110684, + "flos": 27417294391680.0, + "grad_norm": 1.8561148275478438, + "language_loss": 0.75528646, + "learning_rate": 2.8010248235860213e-06, + "loss": 0.77789605, + "num_input_tokens_seen": 69397765, + "step": 3225, + "time_per_iteration": 2.742804527282715 + }, + { + "auxiliary_loss_clip": 0.01114548, + "auxiliary_loss_mlp": 0.01375343, + "balance_loss_clip": 1.02141309, + "balance_loss_mlp": 1.00015092, + "epoch": 0.38790356520170743, + "flos": 64500019879680.0, + "grad_norm": 0.8474591592561935, + "language_loss": 0.62756097, + "learning_rate": 2.8003109978771192e-06, + "loss": 0.65245986, + "num_input_tokens_seen": 69458930, + "step": 3226, + "time_per_iteration": 3.3045835494995117 + }, + { + "auxiliary_loss_clip": 0.01208237, + "auxiliary_loss_mlp": 0.01034859, + "balance_loss_clip": 1.0475657, + "balance_loss_mlp": 1.02597189, + "epoch": 0.38802380809234654, + "flos": 22345415112960.0, + "grad_norm": 2.0462848574705856, + "language_loss": 0.78854501, + "learning_rate": 2.799597050761674e-06, + "loss": 0.81097603, + "num_input_tokens_seen": 69475135, + "step": 3227, + "time_per_iteration": 2.688462734222412 + }, + { + "auxiliary_loss_clip": 0.01196269, + "auxiliary_loss_mlp": 0.01034051, + "balance_loss_clip": 1.05850589, + "balance_loss_mlp": 1.02505088, + "epoch": 0.38814405098298566, + "flos": 25261361199360.0, + "grad_norm": 1.9018969736429037, + "language_loss": 0.7903809, + "learning_rate": 2.7988829823479924e-06, + "loss": 0.81268406, + "num_input_tokens_seen": 69493525, + "step": 3228, + "time_per_iteration": 2.676563262939453 + }, + { + "auxiliary_loss_clip": 0.01204912, + "auxiliary_loss_mlp": 0.01036794, + "balance_loss_clip": 1.05260837, + "balance_loss_mlp": 1.02714956, + "epoch": 0.3882642938736247, + "flos": 18841180078080.0, + "grad_norm": 1.944369433310576, + "language_loss": 0.63797355, + "learning_rate": 2.7981687927443976e-06, + "loss": 0.66039062, + "num_input_tokens_seen": 69510325, + "step": 3229, + "time_per_iteration": 2.692164897918701 + }, + { + "auxiliary_loss_clip": 0.0120136, + "auxiliary_loss_mlp": 0.01027766, + "balance_loss_clip": 1.05298352, + "balance_loss_mlp": 1.01850951, + "epoch": 0.3883845367642638, + "flos": 21652806090240.0, + "grad_norm": 1.9080684259096559, + "language_loss": 0.85817587, + "learning_rate": 2.797454482059231e-06, + "loss": 0.88046718, + "num_input_tokens_seen": 69530480, + "step": 3230, + "time_per_iteration": 2.7378509044647217 + }, + { + "auxiliary_loss_clip": 0.01198607, + "auxiliary_loss_mlp": 0.01029488, + "balance_loss_clip": 1.05870354, + "balance_loss_mlp": 1.01952767, + "epoch": 0.3885047796549029, + "flos": 20557530627840.0, + "grad_norm": 1.6643874151141327, + "language_loss": 0.84642804, + "learning_rate": 2.7967400504008537e-06, + "loss": 0.86870897, + "num_input_tokens_seen": 69549780, + "step": 3231, + "time_per_iteration": 2.5792360305786133 + }, + { + "auxiliary_loss_clip": 0.01129967, + "auxiliary_loss_mlp": 0.0100217, + "balance_loss_clip": 1.02064073, + "balance_loss_mlp": 1.0005486, + "epoch": 0.388625022545542, + "flos": 64325491695360.0, + "grad_norm": 0.79297687407426, + "language_loss": 0.57438195, + "learning_rate": 2.7960254978776456e-06, + "loss": 0.59570336, + "num_input_tokens_seen": 69611870, + "step": 3232, + "time_per_iteration": 3.2751176357269287 + }, + { + "auxiliary_loss_clip": 0.01195966, + "auxiliary_loss_mlp": 0.01030002, + "balance_loss_clip": 1.0582273, + "balance_loss_mlp": 1.02117443, + "epoch": 0.3887452654361811, + "flos": 18113881495680.0, + "grad_norm": 2.3950011885865394, + "language_loss": 0.82193005, + "learning_rate": 2.7953108245980006e-06, + "loss": 0.84418976, + "num_input_tokens_seen": 69630385, + "step": 3233, + "time_per_iteration": 2.588054656982422 + }, + { + "auxiliary_loss_clip": 0.01207329, + "auxiliary_loss_mlp": 0.01028438, + "balance_loss_clip": 1.05548954, + "balance_loss_mlp": 1.02003396, + "epoch": 0.38886550832682015, + "flos": 24975261371520.0, + "grad_norm": 1.7286990991177829, + "language_loss": 0.73919141, + "learning_rate": 2.7945960306703365e-06, + "loss": 0.76154906, + "num_input_tokens_seen": 69653370, + "step": 3234, + "time_per_iteration": 2.7221760749816895 + }, + { + "auxiliary_loss_clip": 0.01210138, + "auxiliary_loss_mlp": 0.01034483, + "balance_loss_clip": 1.05860734, + "balance_loss_mlp": 1.02574539, + "epoch": 0.38898575121745926, + "flos": 27199496275200.0, + "grad_norm": 1.7089539337576702, + "language_loss": 0.654441, + "learning_rate": 2.7938811162030865e-06, + "loss": 0.67688721, + "num_input_tokens_seen": 69673635, + "step": 3235, + "time_per_iteration": 4.46764612197876 + }, + { + "auxiliary_loss_clip": 0.01202916, + "auxiliary_loss_mlp": 0.0102741, + "balance_loss_clip": 1.05707467, + "balance_loss_mlp": 1.01867783, + "epoch": 0.3891059941080984, + "flos": 28763728727040.0, + "grad_norm": 1.6279496481361408, + "language_loss": 0.82141256, + "learning_rate": 2.793166081304702e-06, + "loss": 0.84371579, + "num_input_tokens_seen": 69694130, + "step": 3236, + "time_per_iteration": 3.5916318893432617 + }, + { + "auxiliary_loss_clip": 0.01226563, + "auxiliary_loss_mlp": 0.01034052, + "balance_loss_clip": 1.05318964, + "balance_loss_mlp": 1.0246588, + "epoch": 0.38922623699873743, + "flos": 22893447893760.0, + "grad_norm": 1.8892279330643529, + "language_loss": 0.8235082, + "learning_rate": 2.7924509260836543e-06, + "loss": 0.8461144, + "num_input_tokens_seen": 69713255, + "step": 3237, + "time_per_iteration": 2.651256799697876 + }, + { + "auxiliary_loss_clip": 0.01219132, + "auxiliary_loss_mlp": 0.01033643, + "balance_loss_clip": 1.05307257, + "balance_loss_mlp": 1.0248034, + "epoch": 0.38934647988937654, + "flos": 19792418002560.0, + "grad_norm": 1.6745611432139347, + "language_loss": 0.68575054, + "learning_rate": 2.791735650648431e-06, + "loss": 0.7082783, + "num_input_tokens_seen": 69732375, + "step": 3238, + "time_per_iteration": 2.711151123046875 + }, + { + "auxiliary_loss_clip": 0.01210648, + "auxiliary_loss_mlp": 0.01033681, + "balance_loss_clip": 1.05467713, + "balance_loss_mlp": 1.02547383, + "epoch": 0.38946672278001565, + "flos": 19202081978880.0, + "grad_norm": 1.887159213846744, + "language_loss": 0.74233139, + "learning_rate": 2.791020255107538e-06, + "loss": 0.76477462, + "num_input_tokens_seen": 69749745, + "step": 3239, + "time_per_iteration": 2.636932849884033 + }, + { + "auxiliary_loss_clip": 0.01211376, + "auxiliary_loss_mlp": 0.01026528, + "balance_loss_clip": 1.04771042, + "balance_loss_mlp": 1.01722956, + "epoch": 0.3895869656706547, + "flos": 24936477661440.0, + "grad_norm": 1.6227322208642898, + "language_loss": 0.80476224, + "learning_rate": 2.7903047395695023e-06, + "loss": 0.82714128, + "num_input_tokens_seen": 69769645, + "step": 3240, + "time_per_iteration": 2.776040554046631 + }, + { + "auxiliary_loss_clip": 0.01202155, + "auxiliary_loss_mlp": 0.01379995, + "balance_loss_clip": 1.05697846, + "balance_loss_mlp": 1.0003159, + "epoch": 0.3897072085612938, + "flos": 24133622820480.0, + "grad_norm": 2.2737085751863653, + "language_loss": 0.90155292, + "learning_rate": 2.789589104142865e-06, + "loss": 0.92737442, + "num_input_tokens_seen": 69787270, + "step": 3241, + "time_per_iteration": 2.6309332847595215 + }, + { + "auxiliary_loss_clip": 0.01222049, + "auxiliary_loss_mlp": 0.01028373, + "balance_loss_clip": 1.05213237, + "balance_loss_mlp": 1.02048779, + "epoch": 0.3898274514519329, + "flos": 17166342672000.0, + "grad_norm": 1.6922634966299825, + "language_loss": 0.76811802, + "learning_rate": 2.7888733489361895e-06, + "loss": 0.79062223, + "num_input_tokens_seen": 69805685, + "step": 3242, + "time_per_iteration": 2.770277976989746 + }, + { + "auxiliary_loss_clip": 0.01082153, + "auxiliary_loss_mlp": 0.01004048, + "balance_loss_clip": 1.01747584, + "balance_loss_mlp": 1.00265968, + "epoch": 0.389947694342572, + "flos": 66074807952000.0, + "grad_norm": 0.7300925743961986, + "language_loss": 0.58755171, + "learning_rate": 2.788157474058054e-06, + "loss": 0.6084137, + "num_input_tokens_seen": 69867960, + "step": 3243, + "time_per_iteration": 3.314314365386963 + }, + { + "auxiliary_loss_clip": 0.01192149, + "auxiliary_loss_mlp": 0.0102653, + "balance_loss_clip": 1.05685639, + "balance_loss_mlp": 1.01865649, + "epoch": 0.3900679372332111, + "flos": 25740912700800.0, + "grad_norm": 1.5806146376750099, + "language_loss": 0.70055115, + "learning_rate": 2.7874414796170555e-06, + "loss": 0.72273791, + "num_input_tokens_seen": 69889450, + "step": 3244, + "time_per_iteration": 3.5435543060302734 + }, + { + "auxiliary_loss_clip": 0.01195945, + "auxiliary_loss_mlp": 0.01034587, + "balance_loss_clip": 1.05298197, + "balance_loss_mlp": 1.02536023, + "epoch": 0.3901881801238502, + "flos": 11801611808640.0, + "grad_norm": 2.3950698424099515, + "language_loss": 0.83455443, + "learning_rate": 2.7867253657218113e-06, + "loss": 0.8568598, + "num_input_tokens_seen": 69903340, + "step": 3245, + "time_per_iteration": 2.6542699337005615 + }, + { + "auxiliary_loss_clip": 0.01210991, + "auxiliary_loss_mlp": 0.01379815, + "balance_loss_clip": 1.05380869, + "balance_loss_mlp": 1.00027084, + "epoch": 0.39030842301448926, + "flos": 27308951994240.0, + "grad_norm": 3.734932858846509, + "language_loss": 0.73038399, + "learning_rate": 2.7860091324809544e-06, + "loss": 0.7562921, + "num_input_tokens_seen": 69924400, + "step": 3246, + "time_per_iteration": 2.719177722930908 + }, + { + "auxiliary_loss_clip": 0.01199666, + "auxiliary_loss_mlp": 0.01024968, + "balance_loss_clip": 1.05703211, + "balance_loss_mlp": 1.01629555, + "epoch": 0.39042866590512837, + "flos": 27163334257920.0, + "grad_norm": 2.2699219653055627, + "language_loss": 0.81512702, + "learning_rate": 2.7852927800031377e-06, + "loss": 0.83737326, + "num_input_tokens_seen": 69944565, + "step": 3247, + "time_per_iteration": 2.7280690670013428 + }, + { + "auxiliary_loss_clip": 0.01211208, + "auxiliary_loss_mlp": 0.01028557, + "balance_loss_clip": 1.05414081, + "balance_loss_mlp": 1.02006912, + "epoch": 0.3905489087957674, + "flos": 29716115886720.0, + "grad_norm": 1.7829054850241017, + "language_loss": 0.82527125, + "learning_rate": 2.7845763083970298e-06, + "loss": 0.84766889, + "num_input_tokens_seen": 69964965, + "step": 3248, + "time_per_iteration": 2.7336385250091553 + }, + { + "auxiliary_loss_clip": 0.01189756, + "auxiliary_loss_mlp": 0.01022092, + "balance_loss_clip": 1.05195546, + "balance_loss_mlp": 1.01349056, + "epoch": 0.39066915168640653, + "flos": 24498618871680.0, + "grad_norm": 2.505559931987548, + "language_loss": 0.81849504, + "learning_rate": 2.7838597177713205e-06, + "loss": 0.84061348, + "num_input_tokens_seen": 69986055, + "step": 3249, + "time_per_iteration": 2.7110137939453125 + }, + { + "auxiliary_loss_clip": 0.01223369, + "auxiliary_loss_mlp": 0.01036141, + "balance_loss_clip": 1.04744756, + "balance_loss_mlp": 1.02668214, + "epoch": 0.39078939457704565, + "flos": 20558572122240.0, + "grad_norm": 1.6842873708104724, + "language_loss": 0.73604864, + "learning_rate": 2.7831430082347143e-06, + "loss": 0.75864375, + "num_input_tokens_seen": 70005260, + "step": 3250, + "time_per_iteration": 2.823148488998413 + }, + { + "auxiliary_loss_clip": 0.01203053, + "auxiliary_loss_mlp": 0.01379596, + "balance_loss_clip": 1.05601072, + "balance_loss_mlp": 1.00028205, + "epoch": 0.3909096374676847, + "flos": 22783417557120.0, + "grad_norm": 2.3673988857959785, + "language_loss": 0.82059705, + "learning_rate": 2.7824261798959373e-06, + "loss": 0.84642363, + "num_input_tokens_seen": 70023440, + "step": 3251, + "time_per_iteration": 2.626758098602295 + }, + { + "auxiliary_loss_clip": 0.01211069, + "auxiliary_loss_mlp": 0.01034726, + "balance_loss_clip": 1.05168605, + "balance_loss_mlp": 1.02559471, + "epoch": 0.3910298803583238, + "flos": 23003119094400.0, + "grad_norm": 3.893704190591045, + "language_loss": 0.7995922, + "learning_rate": 2.78170923286373e-06, + "loss": 0.82205009, + "num_input_tokens_seen": 70043040, + "step": 3252, + "time_per_iteration": 2.731325387954712 + }, + { + "auxiliary_loss_clip": 0.0123442, + "auxiliary_loss_mlp": 0.01033218, + "balance_loss_clip": 1.04826164, + "balance_loss_mlp": 1.02377045, + "epoch": 0.3911501232489629, + "flos": 24316264500480.0, + "grad_norm": 2.1176300427503945, + "language_loss": 0.8412708, + "learning_rate": 2.780992167246854e-06, + "loss": 0.86394721, + "num_input_tokens_seen": 70060565, + "step": 3253, + "time_per_iteration": 2.924933671951294 + }, + { + "auxiliary_loss_clip": 0.01107787, + "auxiliary_loss_mlp": 0.01003897, + "balance_loss_clip": 1.01585436, + "balance_loss_mlp": 1.00235879, + "epoch": 0.391270366139602, + "flos": 60869054684160.0, + "grad_norm": 0.9715455418942334, + "language_loss": 0.72094488, + "learning_rate": 2.7802749831540883e-06, + "loss": 0.74206173, + "num_input_tokens_seen": 70119465, + "step": 3254, + "time_per_iteration": 3.610095500946045 + }, + { + "auxiliary_loss_clip": 0.01233391, + "auxiliary_loss_mlp": 0.01033208, + "balance_loss_clip": 1.05269527, + "balance_loss_mlp": 1.02544129, + "epoch": 0.3913906090302411, + "flos": 21543494025600.0, + "grad_norm": 1.94380110816808, + "language_loss": 0.81875223, + "learning_rate": 2.7795576806942268e-06, + "loss": 0.84141821, + "num_input_tokens_seen": 70138270, + "step": 3255, + "time_per_iteration": 2.74568772315979 + }, + { + "auxiliary_loss_clip": 0.01115967, + "auxiliary_loss_mlp": 0.01000347, + "balance_loss_clip": 1.03285694, + "balance_loss_mlp": 0.9990713, + "epoch": 0.3915108519208802, + "flos": 49839953702400.0, + "grad_norm": 0.7583931958054668, + "language_loss": 0.54866564, + "learning_rate": 2.778840259976085e-06, + "loss": 0.56982875, + "num_input_tokens_seen": 70193500, + "step": 3256, + "time_per_iteration": 3.2691128253936768 + }, + { + "auxiliary_loss_clip": 0.01205205, + "auxiliary_loss_mlp": 0.01037553, + "balance_loss_clip": 1.05673313, + "balance_loss_mlp": 1.02855313, + "epoch": 0.39163109481151925, + "flos": 16506447960960.0, + "grad_norm": 2.022200921711003, + "language_loss": 0.77172863, + "learning_rate": 2.778122721108495e-06, + "loss": 0.79415619, + "num_input_tokens_seen": 70211730, + "step": 3257, + "time_per_iteration": 2.6934025287628174 + }, + { + "auxiliary_loss_clip": 0.01196573, + "auxiliary_loss_mlp": 0.01030169, + "balance_loss_clip": 1.05605304, + "balance_loss_mlp": 1.02122831, + "epoch": 0.39175133770215836, + "flos": 26067484177920.0, + "grad_norm": 1.851567760387232, + "language_loss": 0.88278961, + "learning_rate": 2.7774050642003076e-06, + "loss": 0.90505695, + "num_input_tokens_seen": 70232540, + "step": 3258, + "time_per_iteration": 2.649287700653076 + }, + { + "auxiliary_loss_clip": 0.01198145, + "auxiliary_loss_mlp": 0.01028096, + "balance_loss_clip": 1.05969059, + "balance_loss_mlp": 1.01920319, + "epoch": 0.3918715805927975, + "flos": 21872076664320.0, + "grad_norm": 3.395205991971927, + "language_loss": 0.93314958, + "learning_rate": 2.7766872893603896e-06, + "loss": 0.95541197, + "num_input_tokens_seen": 70252515, + "step": 3259, + "time_per_iteration": 2.681570053100586 + }, + { + "auxiliary_loss_clip": 0.01201983, + "auxiliary_loss_mlp": 0.01027459, + "balance_loss_clip": 1.05426633, + "balance_loss_mlp": 1.0191617, + "epoch": 0.39199182348343653, + "flos": 20376181837440.0, + "grad_norm": 1.5703160104702898, + "language_loss": 0.73492026, + "learning_rate": 2.7759693966976275e-06, + "loss": 0.75721467, + "num_input_tokens_seen": 70271020, + "step": 3260, + "time_per_iteration": 2.6273982524871826 + }, + { + "auxiliary_loss_clip": 0.0121446, + "auxiliary_loss_mlp": 0.01031119, + "balance_loss_clip": 1.05004001, + "balance_loss_mlp": 1.02123642, + "epoch": 0.39211206637407564, + "flos": 21683545153920.0, + "grad_norm": 1.9830031781049162, + "language_loss": 0.85021996, + "learning_rate": 2.7752513863209242e-06, + "loss": 0.87267578, + "num_input_tokens_seen": 70289600, + "step": 3261, + "time_per_iteration": 4.576899290084839 + }, + { + "auxiliary_loss_clip": 0.0120697, + "auxiliary_loss_mlp": 0.01379436, + "balance_loss_clip": 1.05626702, + "balance_loss_mlp": 1.00018358, + "epoch": 0.39223230926471475, + "flos": 21066276908160.0, + "grad_norm": 1.6630279942525372, + "language_loss": 0.84353989, + "learning_rate": 2.774533258339203e-06, + "loss": 0.86940396, + "num_input_tokens_seen": 70307060, + "step": 3262, + "time_per_iteration": 3.571319580078125 + }, + { + "auxiliary_loss_clip": 0.0122701, + "auxiliary_loss_mlp": 0.01035221, + "balance_loss_clip": 1.04784274, + "balance_loss_mlp": 1.02636361, + "epoch": 0.3923525521553538, + "flos": 17603016312960.0, + "grad_norm": 2.394761281548421, + "language_loss": 0.79570508, + "learning_rate": 2.7738150128614014e-06, + "loss": 0.81832731, + "num_input_tokens_seen": 70324465, + "step": 3263, + "time_per_iteration": 2.68729305267334 + }, + { + "auxiliary_loss_clip": 0.01207925, + "auxiliary_loss_mlp": 0.01029215, + "balance_loss_clip": 1.05170298, + "balance_loss_mlp": 1.02063179, + "epoch": 0.3924727950459929, + "flos": 20558284813440.0, + "grad_norm": 1.7212758849050929, + "language_loss": 0.89713609, + "learning_rate": 2.7730966499964777e-06, + "loss": 0.9195075, + "num_input_tokens_seen": 70341415, + "step": 3264, + "time_per_iteration": 2.6903035640716553 + }, + { + "auxiliary_loss_clip": 0.01194057, + "auxiliary_loss_mlp": 0.01031474, + "balance_loss_clip": 1.05523646, + "balance_loss_mlp": 1.02285528, + "epoch": 0.39259303793663197, + "flos": 16216110328320.0, + "grad_norm": 2.2934559407208392, + "language_loss": 0.80512464, + "learning_rate": 2.772378169853408e-06, + "loss": 0.82737994, + "num_input_tokens_seen": 70358985, + "step": 3265, + "time_per_iteration": 2.528808355331421 + }, + { + "auxiliary_loss_clip": 0.01222024, + "auxiliary_loss_mlp": 0.01033298, + "balance_loss_clip": 1.05458367, + "balance_loss_mlp": 1.02455974, + "epoch": 0.3927132808272711, + "flos": 16797001075200.0, + "grad_norm": 1.7225610057042937, + "language_loss": 0.74554884, + "learning_rate": 2.771659572541183e-06, + "loss": 0.76810205, + "num_input_tokens_seen": 70376915, + "step": 3266, + "time_per_iteration": 2.7215163707733154 + }, + { + "auxiliary_loss_clip": 0.01207964, + "auxiliary_loss_mlp": 0.01032972, + "balance_loss_clip": 1.05907774, + "balance_loss_mlp": 1.02477646, + "epoch": 0.3928335237179102, + "flos": 20267228908800.0, + "grad_norm": 2.2165282991155224, + "language_loss": 0.8682754, + "learning_rate": 2.7709408581688143e-06, + "loss": 0.89068472, + "num_input_tokens_seen": 70396900, + "step": 3267, + "time_per_iteration": 2.6937100887298584 + }, + { + "auxiliary_loss_clip": 0.01225287, + "auxiliary_loss_mlp": 0.01030438, + "balance_loss_clip": 1.05386448, + "balance_loss_mlp": 1.02209961, + "epoch": 0.39295376660854925, + "flos": 24973250209920.0, + "grad_norm": 1.5737636652225842, + "language_loss": 0.87941468, + "learning_rate": 2.7702220268453307e-06, + "loss": 0.90197194, + "num_input_tokens_seen": 70417260, + "step": 3268, + "time_per_iteration": 2.7207298278808594 + }, + { + "auxiliary_loss_clip": 0.01214624, + "auxiliary_loss_mlp": 0.01028155, + "balance_loss_clip": 1.05497265, + "balance_loss_mlp": 1.01930964, + "epoch": 0.39307400949918836, + "flos": 18697788984960.0, + "grad_norm": 2.0149056994562886, + "language_loss": 0.85261768, + "learning_rate": 2.7695030786797785e-06, + "loss": 0.87504548, + "num_input_tokens_seen": 70433155, + "step": 3269, + "time_per_iteration": 2.6588680744171143 + }, + { + "auxiliary_loss_clip": 0.01226453, + "auxiliary_loss_mlp": 0.01027103, + "balance_loss_clip": 1.05232048, + "balance_loss_mlp": 1.01810849, + "epoch": 0.39319425238982747, + "flos": 22415476590720.0, + "grad_norm": 2.190710563864258, + "language_loss": 0.74823022, + "learning_rate": 2.7687840137812206e-06, + "loss": 0.77076578, + "num_input_tokens_seen": 70451240, + "step": 3270, + "time_per_iteration": 3.6283226013183594 + }, + { + "auxiliary_loss_clip": 0.01088078, + "auxiliary_loss_mlp": 0.00999703, + "balance_loss_clip": 1.01586127, + "balance_loss_mlp": 0.99832648, + "epoch": 0.3933144952804665, + "flos": 66192954762240.0, + "grad_norm": 0.7960432344954764, + "language_loss": 0.62061656, + "learning_rate": 2.7680648322587395e-06, + "loss": 0.64149439, + "num_input_tokens_seen": 70516115, + "step": 3271, + "time_per_iteration": 3.232846975326538 + }, + { + "auxiliary_loss_clip": 0.01192135, + "auxiliary_loss_mlp": 0.01030749, + "balance_loss_clip": 1.05594039, + "balance_loss_mlp": 1.02155244, + "epoch": 0.39343473817110564, + "flos": 15487159720320.0, + "grad_norm": 1.9464672254886166, + "language_loss": 0.80997872, + "learning_rate": 2.7673455342214334e-06, + "loss": 0.83220756, + "num_input_tokens_seen": 70533105, + "step": 3272, + "time_per_iteration": 2.6093273162841797 + }, + { + "auxiliary_loss_clip": 0.01205361, + "auxiliary_loss_mlp": 0.01028894, + "balance_loss_clip": 1.05684829, + "balance_loss_mlp": 1.02093649, + "epoch": 0.39355498106174475, + "flos": 21324905809920.0, + "grad_norm": 2.2653910998824953, + "language_loss": 0.75847393, + "learning_rate": 2.7666261197784198e-06, + "loss": 0.78081644, + "num_input_tokens_seen": 70551920, + "step": 3273, + "time_per_iteration": 2.6299197673797607 + }, + { + "auxiliary_loss_clip": 0.01206374, + "auxiliary_loss_mlp": 0.0103076, + "balance_loss_clip": 1.05514026, + "balance_loss_mlp": 1.02262402, + "epoch": 0.3936752239523838, + "flos": 13296357400320.0, + "grad_norm": 2.0995152422676036, + "language_loss": 0.76844633, + "learning_rate": 2.7659065890388336e-06, + "loss": 0.79081768, + "num_input_tokens_seen": 70567920, + "step": 3274, + "time_per_iteration": 2.6619322299957275 + }, + { + "auxiliary_loss_clip": 0.01211194, + "auxiliary_loss_mlp": 0.01026311, + "balance_loss_clip": 1.05314684, + "balance_loss_mlp": 1.01822877, + "epoch": 0.3937954668430229, + "flos": 16800161472000.0, + "grad_norm": 1.816783936244758, + "language_loss": 0.8477509, + "learning_rate": 2.7651869421118266e-06, + "loss": 0.87012589, + "num_input_tokens_seen": 70584530, + "step": 3275, + "time_per_iteration": 2.634721040725708 + }, + { + "auxiliary_loss_clip": 0.01205724, + "auxiliary_loss_mlp": 0.0103776, + "balance_loss_clip": 1.05664659, + "balance_loss_mlp": 1.0290513, + "epoch": 0.393915709733662, + "flos": 21064229832960.0, + "grad_norm": 1.7107423441020653, + "language_loss": 0.82955259, + "learning_rate": 2.76446717910657e-06, + "loss": 0.85198736, + "num_input_tokens_seen": 70605235, + "step": 3276, + "time_per_iteration": 2.7053444385528564 + }, + { + "auxiliary_loss_clip": 0.01196956, + "auxiliary_loss_mlp": 0.01032401, + "balance_loss_clip": 1.05320323, + "balance_loss_mlp": 1.02430093, + "epoch": 0.3940359526243011, + "flos": 17165265264000.0, + "grad_norm": 2.0035761808672974, + "language_loss": 0.7671088, + "learning_rate": 2.763747300132249e-06, + "loss": 0.78940231, + "num_input_tokens_seen": 70622675, + "step": 3277, + "time_per_iteration": 2.5843214988708496 + }, + { + "auxiliary_loss_clip": 0.01193279, + "auxiliary_loss_mlp": 0.01032891, + "balance_loss_clip": 1.05615652, + "balance_loss_mlp": 1.0241406, + "epoch": 0.3941561955149402, + "flos": 20995856294400.0, + "grad_norm": 1.4925294617167175, + "language_loss": 0.86311591, + "learning_rate": 2.7630273052980704e-06, + "loss": 0.88537759, + "num_input_tokens_seen": 70643265, + "step": 3278, + "time_per_iteration": 2.622878074645996 + }, + { + "auxiliary_loss_clip": 0.01199391, + "auxiliary_loss_mlp": 0.01033255, + "balance_loss_clip": 1.05181503, + "balance_loss_mlp": 1.02389157, + "epoch": 0.39427643840557924, + "flos": 18843406721280.0, + "grad_norm": 1.9576007396947002, + "language_loss": 0.67076641, + "learning_rate": 2.7623071947132554e-06, + "loss": 0.69309294, + "num_input_tokens_seen": 70660295, + "step": 3279, + "time_per_iteration": 2.6397650241851807 + }, + { + "auxiliary_loss_clip": 0.01217087, + "auxiliary_loss_mlp": 0.0102877, + "balance_loss_clip": 1.05369806, + "balance_loss_mlp": 1.02049053, + "epoch": 0.39439668129621835, + "flos": 23258659426560.0, + "grad_norm": 2.418442902952816, + "language_loss": 0.78663099, + "learning_rate": 2.7615869684870458e-06, + "loss": 0.80908954, + "num_input_tokens_seen": 70679605, + "step": 3280, + "time_per_iteration": 2.7291457653045654 + }, + { + "auxiliary_loss_clip": 0.01200687, + "auxiliary_loss_mlp": 0.01031089, + "balance_loss_clip": 1.05598772, + "balance_loss_mlp": 1.02267838, + "epoch": 0.39451692418685746, + "flos": 26652289507200.0, + "grad_norm": 1.8186157429259928, + "language_loss": 0.84725213, + "learning_rate": 2.7608666267286986e-06, + "loss": 0.8695699, + "num_input_tokens_seen": 70699835, + "step": 3281, + "time_per_iteration": 2.657688617706299 + }, + { + "auxiliary_loss_clip": 0.01229082, + "auxiliary_loss_mlp": 0.01036398, + "balance_loss_clip": 1.04540753, + "balance_loss_mlp": 1.02705169, + "epoch": 0.3946371670774965, + "flos": 18258709132800.0, + "grad_norm": 2.5394088443284235, + "language_loss": 0.86897194, + "learning_rate": 2.760146169547489e-06, + "loss": 0.89162683, + "num_input_tokens_seen": 70716600, + "step": 3282, + "time_per_iteration": 2.8224151134490967 + }, + { + "auxiliary_loss_clip": 0.01216178, + "auxiliary_loss_mlp": 0.01029926, + "balance_loss_clip": 1.05798721, + "balance_loss_mlp": 1.02116418, + "epoch": 0.39475740996813563, + "flos": 24206126423040.0, + "grad_norm": 1.4539586838750163, + "language_loss": 0.76289988, + "learning_rate": 2.75942559705271e-06, + "loss": 0.78536093, + "num_input_tokens_seen": 70736335, + "step": 3283, + "time_per_iteration": 2.685945510864258 + }, + { + "auxiliary_loss_clip": 0.01198283, + "auxiliary_loss_mlp": 0.01027825, + "balance_loss_clip": 1.05294645, + "balance_loss_mlp": 1.01915264, + "epoch": 0.39487765285877474, + "flos": 19317858491520.0, + "grad_norm": 2.3387970320266196, + "language_loss": 0.8901819, + "learning_rate": 2.7587049093536713e-06, + "loss": 0.91244292, + "num_input_tokens_seen": 70752665, + "step": 3284, + "time_per_iteration": 2.6757798194885254 + }, + { + "auxiliary_loss_clip": 0.01208348, + "auxiliary_loss_mlp": 0.01034476, + "balance_loss_clip": 1.05706644, + "balance_loss_mlp": 1.0265311, + "epoch": 0.3949978957494138, + "flos": 17311744926720.0, + "grad_norm": 1.8459274390019014, + "language_loss": 0.80468535, + "learning_rate": 2.757984106559701e-06, + "loss": 0.82711357, + "num_input_tokens_seen": 70771650, + "step": 3285, + "time_per_iteration": 2.6225903034210205 + }, + { + "auxiliary_loss_clip": 0.01204052, + "auxiliary_loss_mlp": 0.01024812, + "balance_loss_clip": 1.05433297, + "balance_loss_mlp": 1.01669931, + "epoch": 0.3951181386400529, + "flos": 36317861280000.0, + "grad_norm": 2.112143571554361, + "language_loss": 0.71450925, + "learning_rate": 2.7572631887801446e-06, + "loss": 0.73679793, + "num_input_tokens_seen": 70793275, + "step": 3286, + "time_per_iteration": 2.7845373153686523 + }, + { + "auxiliary_loss_clip": 0.0120259, + "auxiliary_loss_mlp": 0.01030939, + "balance_loss_clip": 1.05423999, + "balance_loss_mlp": 1.02153397, + "epoch": 0.395238381530692, + "flos": 23110348170240.0, + "grad_norm": 1.7253886888894778, + "language_loss": 0.76656812, + "learning_rate": 2.7565421561243654e-06, + "loss": 0.78890336, + "num_input_tokens_seen": 70811440, + "step": 3287, + "time_per_iteration": 4.469307899475098 + }, + { + "auxiliary_loss_clip": 0.01210147, + "auxiliary_loss_mlp": 0.01030558, + "balance_loss_clip": 1.05119193, + "balance_loss_mlp": 1.02215993, + "epoch": 0.3953586244213311, + "flos": 24347614095360.0, + "grad_norm": 2.0364372944635676, + "language_loss": 0.8211863, + "learning_rate": 2.7558210087017413e-06, + "loss": 0.84359342, + "num_input_tokens_seen": 70831375, + "step": 3288, + "time_per_iteration": 3.6687800884246826 + }, + { + "auxiliary_loss_clip": 0.01215068, + "auxiliary_loss_mlp": 0.01031353, + "balance_loss_clip": 1.05449033, + "balance_loss_mlp": 1.02228105, + "epoch": 0.3954788673119702, + "flos": 23440080044160.0, + "grad_norm": 1.8821225530367587, + "language_loss": 0.73735607, + "learning_rate": 2.7550997466216724e-06, + "loss": 0.75982034, + "num_input_tokens_seen": 70849170, + "step": 3289, + "time_per_iteration": 2.7220547199249268 + }, + { + "auxiliary_loss_clip": 0.01208893, + "auxiliary_loss_mlp": 0.01028382, + "balance_loss_clip": 1.0565691, + "balance_loss_mlp": 1.01961398, + "epoch": 0.3955991102026093, + "flos": 17494063384320.0, + "grad_norm": 2.478722556457404, + "language_loss": 0.81140041, + "learning_rate": 2.7543783699935714e-06, + "loss": 0.83377314, + "num_input_tokens_seen": 70867200, + "step": 3290, + "time_per_iteration": 2.695200204849243 + }, + { + "auxiliary_loss_clip": 0.01202945, + "auxiliary_loss_mlp": 0.01028502, + "balance_loss_clip": 1.05809045, + "balance_loss_mlp": 1.01972246, + "epoch": 0.39571935309324835, + "flos": 18221326053120.0, + "grad_norm": 2.903869155579217, + "language_loss": 0.86070359, + "learning_rate": 2.753656878926872e-06, + "loss": 0.88301802, + "num_input_tokens_seen": 70883080, + "step": 3291, + "time_per_iteration": 2.5644285678863525 + }, + { + "auxiliary_loss_clip": 0.01194444, + "auxiliary_loss_mlp": 0.01027954, + "balance_loss_clip": 1.04918599, + "balance_loss_mlp": 1.01923943, + "epoch": 0.39583959598388746, + "flos": 17748813617280.0, + "grad_norm": 1.7766137679917768, + "language_loss": 0.74377006, + "learning_rate": 2.752935273531023e-06, + "loss": 0.76599407, + "num_input_tokens_seen": 70901230, + "step": 3292, + "time_per_iteration": 2.6942970752716064 + }, + { + "auxiliary_loss_clip": 0.01207507, + "auxiliary_loss_mlp": 0.01027662, + "balance_loss_clip": 1.05835199, + "balance_loss_mlp": 1.01823306, + "epoch": 0.39595983887452657, + "flos": 19352368483200.0, + "grad_norm": 1.8515965420190987, + "language_loss": 0.78930259, + "learning_rate": 2.752213553915492e-06, + "loss": 0.81165433, + "num_input_tokens_seen": 70919585, + "step": 3293, + "time_per_iteration": 2.5876359939575195 + }, + { + "auxiliary_loss_clip": 0.01100688, + "auxiliary_loss_mlp": 0.01007888, + "balance_loss_clip": 1.01289368, + "balance_loss_mlp": 1.0066061, + "epoch": 0.3960800817651656, + "flos": 60682282940160.0, + "grad_norm": 0.8216991347813991, + "language_loss": 0.66001761, + "learning_rate": 2.751491720189762e-06, + "loss": 0.68110335, + "num_input_tokens_seen": 70977695, + "step": 3294, + "time_per_iteration": 3.188378095626831 + }, + { + "auxiliary_loss_clip": 0.01210648, + "auxiliary_loss_mlp": 0.01379683, + "balance_loss_clip": 1.05434561, + "balance_loss_mlp": 1.00034189, + "epoch": 0.39620032465580474, + "flos": 16836718538880.0, + "grad_norm": 2.8020332146652063, + "language_loss": 0.91834581, + "learning_rate": 2.7507697724633364e-06, + "loss": 0.94424909, + "num_input_tokens_seen": 70994455, + "step": 3295, + "time_per_iteration": 2.6429812908172607 + }, + { + "auxiliary_loss_clip": 0.01127023, + "auxiliary_loss_mlp": 0.01001076, + "balance_loss_clip": 1.03199863, + "balance_loss_mlp": 0.99989623, + "epoch": 0.3963205675464438, + "flos": 69071445941760.0, + "grad_norm": 0.7754840901177019, + "language_loss": 0.54640013, + "learning_rate": 2.7500477108457327e-06, + "loss": 0.56768113, + "num_input_tokens_seen": 71046465, + "step": 3296, + "time_per_iteration": 3.944276809692383 + }, + { + "auxiliary_loss_clip": 0.01199835, + "auxiliary_loss_mlp": 0.0103506, + "balance_loss_clip": 1.05485988, + "balance_loss_mlp": 1.02676916, + "epoch": 0.3964408104370829, + "flos": 25667439431040.0, + "grad_norm": 1.833892895032049, + "language_loss": 0.80837989, + "learning_rate": 2.7493255354464877e-06, + "loss": 0.83072883, + "num_input_tokens_seen": 71064275, + "step": 3297, + "time_per_iteration": 2.706660747528076 + }, + { + "auxiliary_loss_clip": 0.01239286, + "auxiliary_loss_mlp": 0.0103379, + "balance_loss_clip": 1.03926027, + "balance_loss_mlp": 1.0251832, + "epoch": 0.396561053327722, + "flos": 24277480790400.0, + "grad_norm": 2.2082181409855086, + "language_loss": 0.76170492, + "learning_rate": 2.748603246375156e-06, + "loss": 0.78443575, + "num_input_tokens_seen": 71082290, + "step": 3298, + "time_per_iteration": 3.20184588432312 + }, + { + "auxiliary_loss_clip": 0.01193848, + "auxiliary_loss_mlp": 0.01034379, + "balance_loss_clip": 1.05717397, + "balance_loss_mlp": 1.02609396, + "epoch": 0.39668129621836107, + "flos": 20522302364160.0, + "grad_norm": 2.4381498096419887, + "language_loss": 0.69632536, + "learning_rate": 2.7478808437413055e-06, + "loss": 0.7186076, + "num_input_tokens_seen": 71101700, + "step": 3299, + "time_per_iteration": 2.902240514755249 + }, + { + "auxiliary_loss_clip": 0.01229438, + "auxiliary_loss_mlp": 0.01031801, + "balance_loss_clip": 1.05479932, + "balance_loss_mlp": 1.02341509, + "epoch": 0.3968015391090002, + "flos": 27052585649280.0, + "grad_norm": 2.417808204117015, + "language_loss": 0.6602354, + "learning_rate": 2.7471583276545263e-06, + "loss": 0.68284774, + "num_input_tokens_seen": 71122360, + "step": 3300, + "time_per_iteration": 2.8068692684173584 + }, + { + "auxiliary_loss_clip": 0.01211986, + "auxiliary_loss_mlp": 0.01035131, + "balance_loss_clip": 1.05345774, + "balance_loss_mlp": 1.02605939, + "epoch": 0.3969217819996393, + "flos": 12531819392640.0, + "grad_norm": 1.9125845676513624, + "language_loss": 0.70431918, + "learning_rate": 2.7464356982244224e-06, + "loss": 0.72679031, + "num_input_tokens_seen": 71140360, + "step": 3301, + "time_per_iteration": 2.631944417953491 + }, + { + "auxiliary_loss_clip": 0.01099069, + "auxiliary_loss_mlp": 0.01001292, + "balance_loss_clip": 1.02767467, + "balance_loss_mlp": 0.99993294, + "epoch": 0.39704202489027834, + "flos": 66241399230720.0, + "grad_norm": 0.774686330813944, + "language_loss": 0.61716628, + "learning_rate": 2.745712955560617e-06, + "loss": 0.63816988, + "num_input_tokens_seen": 71196565, + "step": 3302, + "time_per_iteration": 3.2028648853302 + }, + { + "auxiliary_loss_clip": 0.01235493, + "auxiliary_loss_mlp": 0.01036375, + "balance_loss_clip": 1.05019116, + "balance_loss_mlp": 1.02806294, + "epoch": 0.39716226778091746, + "flos": 16982982720000.0, + "grad_norm": 2.4364327163928436, + "language_loss": 0.76831526, + "learning_rate": 2.7449900997727496e-06, + "loss": 0.79103392, + "num_input_tokens_seen": 71214675, + "step": 3303, + "time_per_iteration": 2.7458724975585938 + }, + { + "auxiliary_loss_clip": 0.01212083, + "auxiliary_loss_mlp": 0.01032277, + "balance_loss_clip": 1.05739808, + "balance_loss_mlp": 1.02352679, + "epoch": 0.39728251067155657, + "flos": 23477139901440.0, + "grad_norm": 1.7112200377941496, + "language_loss": 0.84199834, + "learning_rate": 2.744267130970476e-06, + "loss": 0.86444187, + "num_input_tokens_seen": 71234400, + "step": 3304, + "time_per_iteration": 2.7253592014312744 + }, + { + "auxiliary_loss_clip": 0.0120309, + "auxiliary_loss_mlp": 0.01034373, + "balance_loss_clip": 1.05357504, + "balance_loss_mlp": 1.0255096, + "epoch": 0.3974027535621956, + "flos": 20704441253760.0, + "grad_norm": 1.8116375440802321, + "language_loss": 0.76962024, + "learning_rate": 2.7435440492634697e-06, + "loss": 0.79199487, + "num_input_tokens_seen": 71253725, + "step": 3305, + "time_per_iteration": 2.712003231048584 + }, + { + "auxiliary_loss_clip": 0.01211341, + "auxiliary_loss_mlp": 0.01030878, + "balance_loss_clip": 1.05344796, + "balance_loss_mlp": 1.02056694, + "epoch": 0.39752299645283473, + "flos": 21543278544000.0, + "grad_norm": 1.9621548606085406, + "language_loss": 0.66662335, + "learning_rate": 2.7428208547614228e-06, + "loss": 0.68904555, + "num_input_tokens_seen": 71273220, + "step": 3306, + "time_per_iteration": 2.6979494094848633 + }, + { + "auxiliary_loss_clip": 0.01201913, + "auxiliary_loss_mlp": 0.01030215, + "balance_loss_clip": 1.05710948, + "balance_loss_mlp": 1.02072573, + "epoch": 0.39764323934347384, + "flos": 19208295031680.0, + "grad_norm": 4.166210073227472, + "language_loss": 0.77457857, + "learning_rate": 2.742097547574043e-06, + "loss": 0.79689991, + "num_input_tokens_seen": 71291445, + "step": 3307, + "time_per_iteration": 2.6896846294403076 + }, + { + "auxiliary_loss_clip": 0.01217902, + "auxiliary_loss_mlp": 0.01380047, + "balance_loss_clip": 1.05446637, + "balance_loss_mlp": 1.00037777, + "epoch": 0.3977634822341129, + "flos": 20850202644480.0, + "grad_norm": 1.8540076089253044, + "language_loss": 0.77330226, + "learning_rate": 2.7413741278110544e-06, + "loss": 0.79928172, + "num_input_tokens_seen": 71310135, + "step": 3308, + "time_per_iteration": 2.6906967163085938 + }, + { + "auxiliary_loss_clip": 0.01214651, + "auxiliary_loss_mlp": 0.01034854, + "balance_loss_clip": 1.05517876, + "balance_loss_mlp": 1.02585387, + "epoch": 0.397883725124752, + "flos": 39786042038400.0, + "grad_norm": 2.1214546257905855, + "language_loss": 0.69129443, + "learning_rate": 2.7406505955822016e-06, + "loss": 0.71378946, + "num_input_tokens_seen": 71331160, + "step": 3309, + "time_per_iteration": 2.8096060752868652 + }, + { + "auxiliary_loss_clip": 0.01205259, + "auxiliary_loss_mlp": 0.01033149, + "balance_loss_clip": 1.05004954, + "balance_loss_mlp": 1.0243814, + "epoch": 0.39800396801539106, + "flos": 17379507934080.0, + "grad_norm": 3.2508700677277917, + "language_loss": 0.66181982, + "learning_rate": 2.7399269509972415e-06, + "loss": 0.68420386, + "num_input_tokens_seen": 71345315, + "step": 3310, + "time_per_iteration": 2.6246745586395264 + }, + { + "auxiliary_loss_clip": 0.01198709, + "auxiliary_loss_mlp": 0.0103654, + "balance_loss_clip": 1.04792595, + "balance_loss_mlp": 1.02756929, + "epoch": 0.3981242109060302, + "flos": 19202764337280.0, + "grad_norm": 2.1601236764489546, + "language_loss": 0.85200799, + "learning_rate": 2.7392031941659514e-06, + "loss": 0.87436044, + "num_input_tokens_seen": 71363160, + "step": 3311, + "time_per_iteration": 2.70896053314209 + }, + { + "auxiliary_loss_clip": 0.0121583, + "auxiliary_loss_mlp": 0.01035231, + "balance_loss_clip": 1.05930603, + "balance_loss_mlp": 1.02598619, + "epoch": 0.3982444537966693, + "flos": 24565124903040.0, + "grad_norm": 1.969159976152332, + "language_loss": 0.86117804, + "learning_rate": 2.7384793251981244e-06, + "loss": 0.88368869, + "num_input_tokens_seen": 71382145, + "step": 3312, + "time_per_iteration": 2.7041757106781006 + }, + { + "auxiliary_loss_clip": 0.01206657, + "auxiliary_loss_mlp": 0.01031116, + "balance_loss_clip": 1.05442178, + "balance_loss_mlp": 1.0225215, + "epoch": 0.39836469668730834, + "flos": 26213856099840.0, + "grad_norm": 1.798347556837734, + "language_loss": 0.80700487, + "learning_rate": 2.737755344203571e-06, + "loss": 0.8293826, + "num_input_tokens_seen": 71402095, + "step": 3313, + "time_per_iteration": 5.258257627487183 + }, + { + "auxiliary_loss_clip": 0.01205831, + "auxiliary_loss_mlp": 0.01040556, + "balance_loss_clip": 1.05766821, + "balance_loss_mlp": 1.03144288, + "epoch": 0.39848493957794745, + "flos": 27636134002560.0, + "grad_norm": 1.7356829979130446, + "language_loss": 0.79884458, + "learning_rate": 2.7370312512921186e-06, + "loss": 0.82130843, + "num_input_tokens_seen": 71423875, + "step": 3314, + "time_per_iteration": 2.7548041343688965 + }, + { + "auxiliary_loss_clip": 0.01214537, + "auxiliary_loss_mlp": 0.0103409, + "balance_loss_clip": 1.05132735, + "balance_loss_mlp": 1.0254178, + "epoch": 0.39860518246858656, + "flos": 12239326944000.0, + "grad_norm": 2.670019366342523, + "language_loss": 0.76481926, + "learning_rate": 2.736307046573611e-06, + "loss": 0.78730553, + "num_input_tokens_seen": 71439745, + "step": 3315, + "time_per_iteration": 3.525106430053711 + }, + { + "auxiliary_loss_clip": 0.0119081, + "auxiliary_loss_mlp": 0.01027937, + "balance_loss_clip": 1.05513394, + "balance_loss_mlp": 1.01931787, + "epoch": 0.3987254253592256, + "flos": 22379135005440.0, + "grad_norm": 1.5786614335780305, + "language_loss": 0.81548285, + "learning_rate": 2.73558273015791e-06, + "loss": 0.83767033, + "num_input_tokens_seen": 71459575, + "step": 3316, + "time_per_iteration": 2.63179349899292 + }, + { + "auxiliary_loss_clip": 0.01195562, + "auxiliary_loss_mlp": 0.01032661, + "balance_loss_clip": 1.05639982, + "balance_loss_mlp": 1.02316022, + "epoch": 0.3988456682498647, + "flos": 23514020190720.0, + "grad_norm": 2.3394186358133617, + "language_loss": 0.7058931, + "learning_rate": 2.734858302154894e-06, + "loss": 0.72817528, + "num_input_tokens_seen": 71481075, + "step": 3317, + "time_per_iteration": 2.6399645805358887 + }, + { + "auxiliary_loss_clip": 0.01203285, + "auxiliary_loss_mlp": 0.01028424, + "balance_loss_clip": 1.05220771, + "balance_loss_mlp": 1.01977491, + "epoch": 0.39896591114050384, + "flos": 19208761908480.0, + "grad_norm": 1.8346626346641304, + "language_loss": 0.76368171, + "learning_rate": 2.734133762674457e-06, + "loss": 0.78599876, + "num_input_tokens_seen": 71500665, + "step": 3318, + "time_per_iteration": 2.6291277408599854 + }, + { + "auxiliary_loss_clip": 0.01211191, + "auxiliary_loss_mlp": 0.01030814, + "balance_loss_clip": 1.05447257, + "balance_loss_mlp": 1.02117014, + "epoch": 0.3990861540311429, + "flos": 28401031146240.0, + "grad_norm": 16.41537794225347, + "language_loss": 0.70417351, + "learning_rate": 2.7334091118265124e-06, + "loss": 0.72659355, + "num_input_tokens_seen": 71522560, + "step": 3319, + "time_per_iteration": 2.717719554901123 + }, + { + "auxiliary_loss_clip": 0.01091481, + "auxiliary_loss_mlp": 0.01008258, + "balance_loss_clip": 1.01371741, + "balance_loss_mlp": 1.00689876, + "epoch": 0.399206396921782, + "flos": 61758563086080.0, + "grad_norm": 0.6786844052527655, + "language_loss": 0.57835007, + "learning_rate": 2.732684349720989e-06, + "loss": 0.59934747, + "num_input_tokens_seen": 71590520, + "step": 3320, + "time_per_iteration": 3.2087278366088867 + }, + { + "auxiliary_loss_clip": 0.01222581, + "auxiliary_loss_mlp": 0.01035567, + "balance_loss_clip": 1.0529058, + "balance_loss_mlp": 1.02671003, + "epoch": 0.3993266398124211, + "flos": 28074567409920.0, + "grad_norm": 1.591510665798012, + "language_loss": 0.75795084, + "learning_rate": 2.7319594764678318e-06, + "loss": 0.7805323, + "num_input_tokens_seen": 71612620, + "step": 3321, + "time_per_iteration": 3.6586902141571045 + }, + { + "auxiliary_loss_clip": 0.01229151, + "auxiliary_loss_mlp": 0.01030445, + "balance_loss_clip": 1.05137944, + "balance_loss_mlp": 1.02126575, + "epoch": 0.39944688270306017, + "flos": 23225083188480.0, + "grad_norm": 1.9945303276089918, + "language_loss": 0.83510149, + "learning_rate": 2.7312344921770044e-06, + "loss": 0.85769749, + "num_input_tokens_seen": 71634320, + "step": 3322, + "time_per_iteration": 2.782555103302002 + }, + { + "auxiliary_loss_clip": 0.01204521, + "auxiliary_loss_mlp": 0.01031379, + "balance_loss_clip": 1.0486412, + "balance_loss_mlp": 1.02298069, + "epoch": 0.3995671255936993, + "flos": 19390433921280.0, + "grad_norm": 1.7725366341778135, + "language_loss": 0.78255641, + "learning_rate": 2.7305093969584857e-06, + "loss": 0.80491537, + "num_input_tokens_seen": 71653145, + "step": 3323, + "time_per_iteration": 2.646358013153076 + }, + { + "auxiliary_loss_clip": 0.01193682, + "auxiliary_loss_mlp": 0.01035418, + "balance_loss_clip": 1.05288124, + "balance_loss_mlp": 1.02610779, + "epoch": 0.3996873684843384, + "flos": 23842638743040.0, + "grad_norm": 1.7958313681946116, + "language_loss": 0.79766256, + "learning_rate": 2.729784190922272e-06, + "loss": 0.81995356, + "num_input_tokens_seen": 71674580, + "step": 3324, + "time_per_iteration": 2.6760330200195312 + }, + { + "auxiliary_loss_clip": 0.01102166, + "auxiliary_loss_mlp": 0.01006969, + "balance_loss_clip": 1.0144031, + "balance_loss_mlp": 1.00561035, + "epoch": 0.39980761137497745, + "flos": 66576877280640.0, + "grad_norm": 0.9368724949289357, + "language_loss": 0.57151532, + "learning_rate": 2.729058874178378e-06, + "loss": 0.59260666, + "num_input_tokens_seen": 71745260, + "step": 3325, + "time_per_iteration": 3.2911148071289062 + }, + { + "auxiliary_loss_clip": 0.01217354, + "auxiliary_loss_mlp": 0.01028228, + "balance_loss_clip": 1.05723107, + "balance_loss_mlp": 1.01962078, + "epoch": 0.39992785426561656, + "flos": 28549162834560.0, + "grad_norm": 1.8864820292032582, + "language_loss": 0.69172508, + "learning_rate": 2.7283334468368315e-06, + "loss": 0.71418089, + "num_input_tokens_seen": 71766540, + "step": 3326, + "time_per_iteration": 2.758838653564453 + }, + { + "auxiliary_loss_clip": 0.01244793, + "auxiliary_loss_mlp": 0.01036788, + "balance_loss_clip": 1.03997016, + "balance_loss_mlp": 1.02709675, + "epoch": 0.4000480971562556, + "flos": 15049408671360.0, + "grad_norm": 2.050112853159813, + "language_loss": 0.7288847, + "learning_rate": 2.72760790900768e-06, + "loss": 0.75170046, + "num_input_tokens_seen": 71783125, + "step": 3327, + "time_per_iteration": 3.0247185230255127 + }, + { + "auxiliary_loss_clip": 0.01196723, + "auxiliary_loss_mlp": 0.01028779, + "balance_loss_clip": 1.06009984, + "balance_loss_mlp": 1.02007091, + "epoch": 0.4001683400468947, + "flos": 23915609222400.0, + "grad_norm": 1.8601791731772446, + "language_loss": 0.79090631, + "learning_rate": 2.7268822608009875e-06, + "loss": 0.81316137, + "num_input_tokens_seen": 71802500, + "step": 3328, + "time_per_iteration": 2.755751371383667 + }, + { + "auxiliary_loss_clip": 0.01222965, + "auxiliary_loss_mlp": 0.01033717, + "balance_loss_clip": 1.05229592, + "balance_loss_mlp": 1.02431166, + "epoch": 0.40028858293753383, + "flos": 24352677912960.0, + "grad_norm": 1.9480832805783652, + "language_loss": 0.77914762, + "learning_rate": 2.726156502326834e-06, + "loss": 0.80171442, + "num_input_tokens_seen": 71823800, + "step": 3329, + "time_per_iteration": 276.41149520874023 + }, + { + "auxiliary_loss_clip": 0.01151311, + "auxiliary_loss_mlp": 0.01000569, + "balance_loss_clip": 1.02532673, + "balance_loss_mlp": 0.99932355, + "epoch": 0.4004088258281729, + "flos": 66787025800320.0, + "grad_norm": 0.6955215586685437, + "language_loss": 0.60275429, + "learning_rate": 2.725430633695316e-06, + "loss": 0.62427306, + "num_input_tokens_seen": 71886880, + "step": 3330, + "time_per_iteration": 3.424323081970215 + }, + { + "auxiliary_loss_clip": 0.0107938, + "auxiliary_loss_mlp": 0.01002349, + "balance_loss_clip": 1.0156765, + "balance_loss_mlp": 1.00093067, + "epoch": 0.400529068718812, + "flos": 58598386473600.0, + "grad_norm": 0.8867626434740397, + "language_loss": 0.57890213, + "learning_rate": 2.7247046550165485e-06, + "loss": 0.59971941, + "num_input_tokens_seen": 71939005, + "step": 3331, + "time_per_iteration": 4.012409448623657 + }, + { + "auxiliary_loss_clip": 0.01196588, + "auxiliary_loss_mlp": 0.01033596, + "balance_loss_clip": 1.05930042, + "balance_loss_mlp": 1.02480435, + "epoch": 0.4006493116094511, + "flos": 25377460934400.0, + "grad_norm": 1.3971379711620489, + "language_loss": 0.75592417, + "learning_rate": 2.7239785664006606e-06, + "loss": 0.77822602, + "num_input_tokens_seen": 71962545, + "step": 3332, + "time_per_iteration": 2.696873188018799 + }, + { + "auxiliary_loss_clip": 0.01092048, + "auxiliary_loss_mlp": 0.01002013, + "balance_loss_clip": 1.01435292, + "balance_loss_mlp": 1.00055313, + "epoch": 0.40076955450009016, + "flos": 60280729822080.0, + "grad_norm": 0.7887488625449716, + "language_loss": 0.6178664, + "learning_rate": 2.7232523679578002e-06, + "loss": 0.63880694, + "num_input_tokens_seen": 72025625, + "step": 3333, + "time_per_iteration": 3.261148691177368 + }, + { + "auxiliary_loss_clip": 0.01200087, + "auxiliary_loss_mlp": 0.01035231, + "balance_loss_clip": 1.05587077, + "balance_loss_mlp": 1.02726758, + "epoch": 0.4008897973907293, + "flos": 16617268396800.0, + "grad_norm": 1.9045608939100733, + "language_loss": 0.79282475, + "learning_rate": 2.7225260597981295e-06, + "loss": 0.81517792, + "num_input_tokens_seen": 72043330, + "step": 3334, + "time_per_iteration": 2.6102724075317383 + }, + { + "auxiliary_loss_clip": 0.01216829, + "auxiliary_loss_mlp": 0.01380519, + "balance_loss_clip": 1.05399728, + "balance_loss_mlp": 1.00067306, + "epoch": 0.4010100402813684, + "flos": 15377344865280.0, + "grad_norm": 2.292956446006871, + "language_loss": 0.78529775, + "learning_rate": 2.721799642031831e-06, + "loss": 0.81127125, + "num_input_tokens_seen": 72059500, + "step": 3335, + "time_per_iteration": 2.754291296005249 + }, + { + "auxiliary_loss_clip": 0.01216621, + "auxiliary_loss_mlp": 0.01029448, + "balance_loss_clip": 1.05220258, + "balance_loss_mlp": 1.02085876, + "epoch": 0.40113028317200744, + "flos": 13298835438720.0, + "grad_norm": 3.0777368796339903, + "language_loss": 0.77691132, + "learning_rate": 2.721073114769101e-06, + "loss": 0.79937202, + "num_input_tokens_seen": 72077175, + "step": 3336, + "time_per_iteration": 2.9084596633911133 + }, + { + "auxiliary_loss_clip": 0.01213993, + "auxiliary_loss_mlp": 0.01029168, + "balance_loss_clip": 1.05132174, + "balance_loss_mlp": 1.02026939, + "epoch": 0.40125052606264655, + "flos": 20668027841280.0, + "grad_norm": 1.9591681894272706, + "language_loss": 0.75191295, + "learning_rate": 2.7203464781201523e-06, + "loss": 0.77434456, + "num_input_tokens_seen": 72096490, + "step": 3337, + "time_per_iteration": 2.723128318786621 + }, + { + "auxiliary_loss_clip": 0.01196872, + "auxiliary_loss_mlp": 0.01030924, + "balance_loss_clip": 1.05924726, + "balance_loss_mlp": 1.0224005, + "epoch": 0.40137076895328566, + "flos": 24607679541120.0, + "grad_norm": 2.0153715958844542, + "language_loss": 0.7797935, + "learning_rate": 2.719619732195215e-06, + "loss": 0.80207151, + "num_input_tokens_seen": 72118130, + "step": 3338, + "time_per_iteration": 3.572361469268799 + }, + { + "auxiliary_loss_clip": 0.01219905, + "auxiliary_loss_mlp": 0.01032034, + "balance_loss_clip": 1.05206108, + "balance_loss_mlp": 1.0230341, + "epoch": 0.4014910118439247, + "flos": 24206593299840.0, + "grad_norm": 1.4438582065728793, + "language_loss": 0.72467816, + "learning_rate": 2.7188928771045377e-06, + "loss": 0.74719763, + "num_input_tokens_seen": 72139450, + "step": 3339, + "time_per_iteration": 3.68753981590271 + }, + { + "auxiliary_loss_clip": 0.01212374, + "auxiliary_loss_mlp": 0.01033596, + "balance_loss_clip": 1.05148637, + "balance_loss_mlp": 1.02467942, + "epoch": 0.4016112547345638, + "flos": 26725080418560.0, + "grad_norm": 1.616355046683335, + "language_loss": 0.79782826, + "learning_rate": 2.7181659129583815e-06, + "loss": 0.82028794, + "num_input_tokens_seen": 72159040, + "step": 3340, + "time_per_iteration": 3.668099880218506 + }, + { + "auxiliary_loss_clip": 0.01197426, + "auxiliary_loss_mlp": 0.01028109, + "balance_loss_clip": 1.04719651, + "balance_loss_mlp": 1.01893032, + "epoch": 0.4017314976252029, + "flos": 21288025520640.0, + "grad_norm": 1.7104884109184806, + "language_loss": 0.7572242, + "learning_rate": 2.7174388398670276e-06, + "loss": 0.77947962, + "num_input_tokens_seen": 72178220, + "step": 3341, + "time_per_iteration": 2.76462984085083 + }, + { + "auxiliary_loss_clip": 0.01192966, + "auxiliary_loss_mlp": 0.01029764, + "balance_loss_clip": 1.0535177, + "balance_loss_mlp": 1.02110326, + "epoch": 0.401851740515842, + "flos": 25484690010240.0, + "grad_norm": 1.8552834333303814, + "language_loss": 0.92346448, + "learning_rate": 2.716711657940773e-06, + "loss": 0.94569176, + "num_input_tokens_seen": 72199230, + "step": 3342, + "time_per_iteration": 2.6172187328338623 + }, + { + "auxiliary_loss_clip": 0.01117809, + "auxiliary_loss_mlp": 0.01000979, + "balance_loss_clip": 1.01521778, + "balance_loss_mlp": 0.99966127, + "epoch": 0.4019719834064811, + "flos": 55395334978560.0, + "grad_norm": 0.8496792350614548, + "language_loss": 0.56447637, + "learning_rate": 2.7159843672899284e-06, + "loss": 0.58566427, + "num_input_tokens_seen": 72263430, + "step": 3343, + "time_per_iteration": 3.380760908126831 + }, + { + "auxiliary_loss_clip": 0.01202839, + "auxiliary_loss_mlp": 0.01032322, + "balance_loss_clip": 1.05764723, + "balance_loss_mlp": 1.02333939, + "epoch": 0.40209222629712016, + "flos": 18180100218240.0, + "grad_norm": 1.9790228293753205, + "language_loss": 0.81255507, + "learning_rate": 2.715256968024825e-06, + "loss": 0.8349067, + "num_input_tokens_seen": 72280505, + "step": 3344, + "time_per_iteration": 2.662628650665283 + }, + { + "auxiliary_loss_clip": 0.01217731, + "auxiliary_loss_mlp": 0.01032693, + "balance_loss_clip": 1.05326581, + "balance_loss_mlp": 1.0240922, + "epoch": 0.40221246918775927, + "flos": 25961009287680.0, + "grad_norm": 1.531243681774412, + "language_loss": 0.82344544, + "learning_rate": 2.7145294602558083e-06, + "loss": 0.84594965, + "num_input_tokens_seen": 72301215, + "step": 3345, + "time_per_iteration": 2.701873779296875 + }, + { + "auxiliary_loss_clip": 0.01199019, + "auxiliary_loss_mlp": 0.01034758, + "balance_loss_clip": 1.05326152, + "balance_loss_mlp": 1.02529252, + "epoch": 0.4023327120783984, + "flos": 33838912056960.0, + "grad_norm": 1.7058657417238312, + "language_loss": 0.70612818, + "learning_rate": 2.713801844093241e-06, + "loss": 0.72846591, + "num_input_tokens_seen": 72322365, + "step": 3346, + "time_per_iteration": 2.9667651653289795 + }, + { + "auxiliary_loss_clip": 0.01205315, + "auxiliary_loss_mlp": 0.01033348, + "balance_loss_clip": 1.05663991, + "balance_loss_mlp": 1.02450871, + "epoch": 0.40245295496903744, + "flos": 26900252069760.0, + "grad_norm": 2.4201767329952095, + "language_loss": 0.88449138, + "learning_rate": 2.7130741196475014e-06, + "loss": 0.90687799, + "num_input_tokens_seen": 72340495, + "step": 3347, + "time_per_iteration": 3.5816171169281006 + }, + { + "auxiliary_loss_clip": 0.0121227, + "auxiliary_loss_mlp": 0.01033892, + "balance_loss_clip": 1.05427027, + "balance_loss_mlp": 1.02484989, + "epoch": 0.40257319785967655, + "flos": 36902738436480.0, + "grad_norm": 2.0851951266850337, + "language_loss": 0.78841245, + "learning_rate": 2.7123462870289848e-06, + "loss": 0.81087404, + "num_input_tokens_seen": 72360545, + "step": 3348, + "time_per_iteration": 2.7985000610351562 + }, + { + "auxiliary_loss_clip": 0.01209531, + "auxiliary_loss_mlp": 0.01031048, + "balance_loss_clip": 1.05032492, + "balance_loss_mlp": 1.02273953, + "epoch": 0.40269344075031566, + "flos": 24353180703360.0, + "grad_norm": 1.834414462708624, + "language_loss": 0.80962133, + "learning_rate": 2.711618346348102e-06, + "loss": 0.8320272, + "num_input_tokens_seen": 72381070, + "step": 3349, + "time_per_iteration": 2.7616093158721924 + }, + { + "auxiliary_loss_clip": 0.01203442, + "auxiliary_loss_mlp": 0.01030762, + "balance_loss_clip": 1.05465984, + "balance_loss_mlp": 1.02287054, + "epoch": 0.4028136836409547, + "flos": 14389657614720.0, + "grad_norm": 1.5212862743375024, + "language_loss": 0.63387066, + "learning_rate": 2.7108902977152825e-06, + "loss": 0.65621269, + "num_input_tokens_seen": 72398970, + "step": 3350, + "time_per_iteration": 2.66650390625 + }, + { + "auxiliary_loss_clip": 0.01193956, + "auxiliary_loss_mlp": 0.01031503, + "balance_loss_clip": 1.05168438, + "balance_loss_mlp": 1.02195978, + "epoch": 0.4029339265315938, + "flos": 26136037284480.0, + "grad_norm": 2.120244546491041, + "language_loss": 0.74627656, + "learning_rate": 2.7101621412409704e-06, + "loss": 0.76853108, + "num_input_tokens_seen": 72418455, + "step": 3351, + "time_per_iteration": 2.6918280124664307 + }, + { + "auxiliary_loss_clip": 0.01191121, + "auxiliary_loss_mlp": 0.0103374, + "balance_loss_clip": 1.05445826, + "balance_loss_mlp": 1.02451944, + "epoch": 0.40305416942223293, + "flos": 23256325042560.0, + "grad_norm": 2.0228001947633554, + "language_loss": 0.85826814, + "learning_rate": 2.7094338770356256e-06, + "loss": 0.88051671, + "num_input_tokens_seen": 72437540, + "step": 3352, + "time_per_iteration": 2.6503713130950928 + }, + { + "auxiliary_loss_clip": 0.01204518, + "auxiliary_loss_mlp": 0.01027744, + "balance_loss_clip": 1.05105388, + "balance_loss_mlp": 1.01951265, + "epoch": 0.403174412312872, + "flos": 27089645506560.0, + "grad_norm": 2.1327625954536353, + "language_loss": 0.64067173, + "learning_rate": 2.708705505209726e-06, + "loss": 0.66299438, + "num_input_tokens_seen": 72458315, + "step": 3353, + "time_per_iteration": 2.8094046115875244 + }, + { + "auxiliary_loss_clip": 0.01217969, + "auxiliary_loss_mlp": 0.01032762, + "balance_loss_clip": 1.04753113, + "balance_loss_mlp": 1.0243336, + "epoch": 0.4032946552035111, + "flos": 21756336065280.0, + "grad_norm": 2.379635397106627, + "language_loss": 0.91723377, + "learning_rate": 2.7079770258737646e-06, + "loss": 0.93974102, + "num_input_tokens_seen": 72476225, + "step": 3354, + "time_per_iteration": 2.855120897293091 + }, + { + "auxiliary_loss_clip": 0.01208333, + "auxiliary_loss_mlp": 0.01034024, + "balance_loss_clip": 1.04621506, + "balance_loss_mlp": 1.02458847, + "epoch": 0.4034148980941502, + "flos": 17343956448000.0, + "grad_norm": 2.2765345006934345, + "language_loss": 0.75318485, + "learning_rate": 2.707248439138251e-06, + "loss": 0.77560842, + "num_input_tokens_seen": 72492460, + "step": 3355, + "time_per_iteration": 2.7141947746276855 + }, + { + "auxiliary_loss_clip": 0.01206894, + "auxiliary_loss_mlp": 0.01029977, + "balance_loss_clip": 1.05707979, + "balance_loss_mlp": 1.02080989, + "epoch": 0.40353514098478926, + "flos": 22017838055040.0, + "grad_norm": 8.340950010849046, + "language_loss": 0.65561998, + "learning_rate": 2.7065197451137114e-06, + "loss": 0.67798865, + "num_input_tokens_seen": 72513840, + "step": 3356, + "time_per_iteration": 2.7991058826446533 + }, + { + "auxiliary_loss_clip": 0.01206788, + "auxiliary_loss_mlp": 0.01028714, + "balance_loss_clip": 1.05307734, + "balance_loss_mlp": 1.01948142, + "epoch": 0.4036553838754284, + "flos": 14246446089600.0, + "grad_norm": 2.229758051946604, + "language_loss": 0.67945939, + "learning_rate": 2.7057909439106894e-06, + "loss": 0.70181441, + "num_input_tokens_seen": 72531695, + "step": 3357, + "time_per_iteration": 2.7707719802856445 + }, + { + "auxiliary_loss_clip": 0.0118805, + "auxiliary_loss_mlp": 0.01379922, + "balance_loss_clip": 1.05101848, + "balance_loss_mlp": 1.00054622, + "epoch": 0.40377562676606743, + "flos": 24790644443520.0, + "grad_norm": 2.2668737845931006, + "language_loss": 0.78215069, + "learning_rate": 2.7050620356397417e-06, + "loss": 0.80783045, + "num_input_tokens_seen": 72550645, + "step": 3358, + "time_per_iteration": 2.693938970565796 + }, + { + "auxiliary_loss_clip": 0.01193831, + "auxiliary_loss_mlp": 0.01029017, + "balance_loss_clip": 1.05949903, + "balance_loss_mlp": 1.02104771, + "epoch": 0.40389586965670654, + "flos": 24061226958720.0, + "grad_norm": 1.7074304287582875, + "language_loss": 0.72373855, + "learning_rate": 2.7043330204114437e-06, + "loss": 0.74596703, + "num_input_tokens_seen": 72569355, + "step": 3359, + "time_per_iteration": 2.6758944988250732 + }, + { + "auxiliary_loss_clip": 0.0118751, + "auxiliary_loss_mlp": 0.01026388, + "balance_loss_clip": 1.05312812, + "balance_loss_mlp": 1.01762664, + "epoch": 0.40401611254734565, + "flos": 16399613934720.0, + "grad_norm": 1.797298224662474, + "language_loss": 0.8571232, + "learning_rate": 2.7036038983363862e-06, + "loss": 0.87926215, + "num_input_tokens_seen": 72585960, + "step": 3360, + "time_per_iteration": 2.633817434310913 + }, + { + "auxiliary_loss_clip": 0.01194239, + "auxiliary_loss_mlp": 0.01031339, + "balance_loss_clip": 1.05284119, + "balance_loss_mlp": 1.02278543, + "epoch": 0.4041363554379847, + "flos": 23988220565760.0, + "grad_norm": 1.6661885131171386, + "language_loss": 0.84504324, + "learning_rate": 2.702874669525177e-06, + "loss": 0.86729896, + "num_input_tokens_seen": 72604440, + "step": 3361, + "time_per_iteration": 2.6904256343841553 + }, + { + "auxiliary_loss_clip": 0.01220547, + "auxiliary_loss_mlp": 0.01028436, + "balance_loss_clip": 1.05491865, + "balance_loss_mlp": 1.02065134, + "epoch": 0.4042565983286238, + "flos": 28401964899840.0, + "grad_norm": 1.9798898295106984, + "language_loss": 0.69770795, + "learning_rate": 2.7021453340884394e-06, + "loss": 0.7201978, + "num_input_tokens_seen": 72622165, + "step": 3362, + "time_per_iteration": 2.820643663406372 + }, + { + "auxiliary_loss_clip": 0.01192827, + "auxiliary_loss_mlp": 0.01379317, + "balance_loss_clip": 1.05030942, + "balance_loss_mlp": 1.00037813, + "epoch": 0.40437684121926293, + "flos": 17710963660800.0, + "grad_norm": 2.4238818313019594, + "language_loss": 0.73196495, + "learning_rate": 2.7014158921368125e-06, + "loss": 0.75768638, + "num_input_tokens_seen": 72640490, + "step": 3363, + "time_per_iteration": 2.823014259338379 + }, + { + "auxiliary_loss_clip": 0.01192637, + "auxiliary_loss_mlp": 0.01033962, + "balance_loss_clip": 1.05674541, + "balance_loss_mlp": 1.02559948, + "epoch": 0.404497084109902, + "flos": 24018959629440.0, + "grad_norm": 1.7991922824765798, + "language_loss": 0.85510814, + "learning_rate": 2.700686343780953e-06, + "loss": 0.87737417, + "num_input_tokens_seen": 72660360, + "step": 3364, + "time_per_iteration": 3.531083345413208 + }, + { + "auxiliary_loss_clip": 0.01208052, + "auxiliary_loss_mlp": 0.01036318, + "balance_loss_clip": 1.05196333, + "balance_loss_mlp": 1.02707958, + "epoch": 0.4046173270005411, + "flos": 22929861306240.0, + "grad_norm": 1.6983098090086715, + "language_loss": 0.87794125, + "learning_rate": 2.699956689131532e-06, + "loss": 0.9003849, + "num_input_tokens_seen": 72680345, + "step": 3365, + "time_per_iteration": 3.5681352615356445 + }, + { + "auxiliary_loss_clip": 0.01213579, + "auxiliary_loss_mlp": 0.01032992, + "balance_loss_clip": 1.05471659, + "balance_loss_mlp": 1.02437317, + "epoch": 0.4047375698911802, + "flos": 20668135582080.0, + "grad_norm": 2.234134179967459, + "language_loss": 0.85046315, + "learning_rate": 2.699226928299238e-06, + "loss": 0.87292892, + "num_input_tokens_seen": 72698365, + "step": 3366, + "time_per_iteration": 3.4922165870666504 + }, + { + "auxiliary_loss_clip": 0.01203234, + "auxiliary_loss_mlp": 0.01028432, + "balance_loss_clip": 1.05547118, + "balance_loss_mlp": 1.02020073, + "epoch": 0.40485781278181926, + "flos": 28912865996160.0, + "grad_norm": 2.363242545810815, + "language_loss": 0.78706479, + "learning_rate": 2.698497061394774e-06, + "loss": 0.80938148, + "num_input_tokens_seen": 72716850, + "step": 3367, + "time_per_iteration": 2.7031784057617188 + }, + { + "auxiliary_loss_clip": 0.01226383, + "auxiliary_loss_mlp": 0.01380315, + "balance_loss_clip": 1.05564702, + "balance_loss_mlp": 1.0004214, + "epoch": 0.40497805567245837, + "flos": 23148377694720.0, + "grad_norm": 1.5916435246686254, + "language_loss": 0.80858761, + "learning_rate": 2.6977670885288627e-06, + "loss": 0.83465457, + "num_input_tokens_seen": 72738250, + "step": 3368, + "time_per_iteration": 2.8026647567749023 + }, + { + "auxiliary_loss_clip": 0.01192931, + "auxiliary_loss_mlp": 0.01030619, + "balance_loss_clip": 1.04961157, + "balance_loss_mlp": 1.02185154, + "epoch": 0.4050982985630975, + "flos": 16289404030080.0, + "grad_norm": 1.7829072727447324, + "language_loss": 0.75295568, + "learning_rate": 2.6970370098122378e-06, + "loss": 0.77519119, + "num_input_tokens_seen": 72755235, + "step": 3369, + "time_per_iteration": 2.774059772491455 + }, + { + "auxiliary_loss_clip": 0.0119065, + "auxiliary_loss_mlp": 0.01031425, + "balance_loss_clip": 1.05524015, + "balance_loss_mlp": 1.02278793, + "epoch": 0.40521854145373654, + "flos": 34459484353920.0, + "grad_norm": 1.5427093727067038, + "language_loss": 0.86672044, + "learning_rate": 2.6963068253556535e-06, + "loss": 0.88894117, + "num_input_tokens_seen": 72776620, + "step": 3370, + "time_per_iteration": 2.8353428840637207 + }, + { + "auxiliary_loss_clip": 0.01208885, + "auxiliary_loss_mlp": 0.01027622, + "balance_loss_clip": 1.05492616, + "balance_loss_mlp": 1.01827645, + "epoch": 0.40533878434437565, + "flos": 25331099454720.0, + "grad_norm": 1.7851518671099327, + "language_loss": 0.8561272, + "learning_rate": 2.6955765352698763e-06, + "loss": 0.87849224, + "num_input_tokens_seen": 72796765, + "step": 3371, + "time_per_iteration": 2.6880624294281006 + }, + { + "auxiliary_loss_clip": 0.01195971, + "auxiliary_loss_mlp": 0.01037807, + "balance_loss_clip": 1.0572902, + "balance_loss_mlp": 1.02872872, + "epoch": 0.40545902723501476, + "flos": 15012061505280.0, + "grad_norm": 1.8799847453989014, + "language_loss": 0.731902, + "learning_rate": 2.6948461396656923e-06, + "loss": 0.7542398, + "num_input_tokens_seen": 72814175, + "step": 3372, + "time_per_iteration": 2.7399420738220215 + }, + { + "auxiliary_loss_clip": 0.01205094, + "auxiliary_loss_mlp": 0.01028391, + "balance_loss_clip": 1.05477107, + "balance_loss_mlp": 1.01951623, + "epoch": 0.4055792701256538, + "flos": 25521103422720.0, + "grad_norm": 2.061129388296727, + "language_loss": 0.74365658, + "learning_rate": 2.6941156386539013e-06, + "loss": 0.76599139, + "num_input_tokens_seen": 72834125, + "step": 3373, + "time_per_iteration": 3.5775911808013916 + }, + { + "auxiliary_loss_clip": 0.01206589, + "auxiliary_loss_mlp": 0.0103175, + "balance_loss_clip": 1.05716968, + "balance_loss_mlp": 1.02257037, + "epoch": 0.4056995130162929, + "flos": 19574583972480.0, + "grad_norm": 2.0090901276145394, + "language_loss": 0.80936009, + "learning_rate": 2.6933850323453203e-06, + "loss": 0.83174348, + "num_input_tokens_seen": 72852570, + "step": 3374, + "time_per_iteration": 2.6924936771392822 + }, + { + "auxiliary_loss_clip": 0.01191505, + "auxiliary_loss_mlp": 0.01029292, + "balance_loss_clip": 1.05588293, + "balance_loss_mlp": 1.02068567, + "epoch": 0.405819755906932, + "flos": 15413794191360.0, + "grad_norm": 1.849621550360638, + "language_loss": 0.75035918, + "learning_rate": 2.6926543208507806e-06, + "loss": 0.77256721, + "num_input_tokens_seen": 72871250, + "step": 3375, + "time_per_iteration": 2.545780897140503 + }, + { + "auxiliary_loss_clip": 0.01201075, + "auxiliary_loss_mlp": 0.0102697, + "balance_loss_clip": 1.05579424, + "balance_loss_mlp": 1.01819634, + "epoch": 0.4059399987975711, + "flos": 21433930565760.0, + "grad_norm": 2.514618569287911, + "language_loss": 0.80002594, + "learning_rate": 2.6919235042811316e-06, + "loss": 0.82230633, + "num_input_tokens_seen": 72890035, + "step": 3376, + "time_per_iteration": 2.632854700088501 + }, + { + "auxiliary_loss_clip": 0.01215553, + "auxiliary_loss_mlp": 0.01035965, + "balance_loss_clip": 1.05235147, + "balance_loss_mlp": 1.02666688, + "epoch": 0.4060602416882102, + "flos": 25556942217600.0, + "grad_norm": 1.9794415800511427, + "language_loss": 0.76646674, + "learning_rate": 2.691192582747237e-06, + "loss": 0.78898191, + "num_input_tokens_seen": 72909665, + "step": 3377, + "time_per_iteration": 2.822512626647949 + }, + { + "auxiliary_loss_clip": 0.01192889, + "auxiliary_loss_mlp": 0.01037669, + "balance_loss_clip": 1.05601454, + "balance_loss_mlp": 1.02841818, + "epoch": 0.40618048457884925, + "flos": 23766759262080.0, + "grad_norm": 1.8025264538450279, + "language_loss": 0.7383343, + "learning_rate": 2.6904615563599765e-06, + "loss": 0.76063985, + "num_input_tokens_seen": 72929465, + "step": 3378, + "time_per_iteration": 2.60967755317688 + }, + { + "auxiliary_loss_clip": 0.01209505, + "auxiliary_loss_mlp": 0.01037642, + "balance_loss_clip": 1.04876363, + "balance_loss_mlp": 1.02926159, + "epoch": 0.40630072746948837, + "flos": 17639681120640.0, + "grad_norm": 1.6356032604063566, + "language_loss": 0.83378267, + "learning_rate": 2.6897304252302477e-06, + "loss": 0.85625416, + "num_input_tokens_seen": 72946785, + "step": 3379, + "time_per_iteration": 2.706331968307495 + }, + { + "auxiliary_loss_clip": 0.01107944, + "auxiliary_loss_mlp": 0.0100852, + "balance_loss_clip": 1.01024234, + "balance_loss_mlp": 1.00713086, + "epoch": 0.4064209703601275, + "flos": 60836053063680.0, + "grad_norm": 0.7921635086012736, + "language_loss": 0.54810739, + "learning_rate": 2.688999189468962e-06, + "loss": 0.56927204, + "num_input_tokens_seen": 73003215, + "step": 3380, + "time_per_iteration": 3.123375177383423 + }, + { + "auxiliary_loss_clip": 0.01201802, + "auxiliary_loss_mlp": 0.01029478, + "balance_loss_clip": 1.05656719, + "balance_loss_mlp": 1.02078819, + "epoch": 0.40654121325076653, + "flos": 24024346669440.0, + "grad_norm": 2.126265591592166, + "language_loss": 0.76348937, + "learning_rate": 2.6882678491870464e-06, + "loss": 0.78580219, + "num_input_tokens_seen": 73023650, + "step": 3381, + "time_per_iteration": 2.8346452713012695 + }, + { + "auxiliary_loss_clip": 0.01205334, + "auxiliary_loss_mlp": 0.01030594, + "balance_loss_clip": 1.05584264, + "balance_loss_mlp": 1.02147484, + "epoch": 0.40666145614140564, + "flos": 27344252085120.0, + "grad_norm": 1.9144935238410492, + "language_loss": 0.71086645, + "learning_rate": 2.6875364044954453e-06, + "loss": 0.7332257, + "num_input_tokens_seen": 73043880, + "step": 3382, + "time_per_iteration": 2.716498374938965 + }, + { + "auxiliary_loss_clip": 0.01204404, + "auxiliary_loss_mlp": 0.01030676, + "balance_loss_clip": 1.04899025, + "balance_loss_mlp": 1.02184296, + "epoch": 0.40678169903204475, + "flos": 26176724415360.0, + "grad_norm": 1.4901355720319698, + "language_loss": 0.8236084, + "learning_rate": 2.6868048555051185e-06, + "loss": 0.84595925, + "num_input_tokens_seen": 73065410, + "step": 3383, + "time_per_iteration": 2.7747128009796143 + }, + { + "auxiliary_loss_clip": 0.01208793, + "auxiliary_loss_mlp": 0.01031955, + "balance_loss_clip": 1.04840124, + "balance_loss_mlp": 1.02313328, + "epoch": 0.4069019419226838, + "flos": 28622420622720.0, + "grad_norm": 3.0458773782484965, + "language_loss": 0.85945928, + "learning_rate": 2.686073202327041e-06, + "loss": 0.88186669, + "num_input_tokens_seen": 73084410, + "step": 3384, + "time_per_iteration": 2.723633050918579 + }, + { + "auxiliary_loss_clip": 0.01191778, + "auxiliary_loss_mlp": 0.01031533, + "balance_loss_clip": 1.04733431, + "balance_loss_mlp": 1.02303338, + "epoch": 0.4070221848133229, + "flos": 25229006023680.0, + "grad_norm": 1.7370553055368565, + "language_loss": 0.73070991, + "learning_rate": 2.6853414450722043e-06, + "loss": 0.75294304, + "num_input_tokens_seen": 73104075, + "step": 3385, + "time_per_iteration": 2.8412036895751953 + }, + { + "auxiliary_loss_clip": 0.01196323, + "auxiliary_loss_mlp": 0.01027644, + "balance_loss_clip": 1.05270648, + "balance_loss_mlp": 1.01957357, + "epoch": 0.40714242770396203, + "flos": 18405224709120.0, + "grad_norm": 1.613441212124173, + "language_loss": 0.85249817, + "learning_rate": 2.684609583851616e-06, + "loss": 0.87473786, + "num_input_tokens_seen": 73122250, + "step": 3386, + "time_per_iteration": 2.5978946685791016 + }, + { + "auxiliary_loss_clip": 0.01223782, + "auxiliary_loss_mlp": 0.01033833, + "balance_loss_clip": 1.04989505, + "balance_loss_mlp": 1.0250355, + "epoch": 0.4072626705946011, + "flos": 30228920403840.0, + "grad_norm": 1.5977007143966537, + "language_loss": 0.80788952, + "learning_rate": 2.683877618776297e-06, + "loss": 0.83046567, + "num_input_tokens_seen": 73144505, + "step": 3387, + "time_per_iteration": 2.894223213195801 + }, + { + "auxiliary_loss_clip": 0.01196731, + "auxiliary_loss_mlp": 0.01035201, + "balance_loss_clip": 1.04649138, + "balance_loss_mlp": 1.02525854, + "epoch": 0.4073829134852402, + "flos": 21834549930240.0, + "grad_norm": 2.3704626838442344, + "language_loss": 0.74284083, + "learning_rate": 2.6831455499572876e-06, + "loss": 0.76516014, + "num_input_tokens_seen": 73162440, + "step": 3388, + "time_per_iteration": 2.7789173126220703 + }, + { + "auxiliary_loss_clip": 0.01190102, + "auxiliary_loss_mlp": 0.01027095, + "balance_loss_clip": 1.05297363, + "balance_loss_mlp": 1.01837516, + "epoch": 0.40750315637587925, + "flos": 25260211964160.0, + "grad_norm": 3.161519408869815, + "language_loss": 0.77410233, + "learning_rate": 2.682413377505641e-06, + "loss": 0.79627436, + "num_input_tokens_seen": 73181245, + "step": 3389, + "time_per_iteration": 2.813887357711792 + }, + { + "auxiliary_loss_clip": 0.01199962, + "auxiliary_loss_mlp": 0.01029074, + "balance_loss_clip": 1.05269647, + "balance_loss_mlp": 1.02086616, + "epoch": 0.40762339926651836, + "flos": 19712767593600.0, + "grad_norm": 1.8099893761371968, + "language_loss": 0.76545274, + "learning_rate": 2.6816811015324284e-06, + "loss": 0.78774315, + "num_input_tokens_seen": 73199295, + "step": 3390, + "time_per_iteration": 3.6786630153656006 + }, + { + "auxiliary_loss_clip": 0.01075997, + "auxiliary_loss_mlp": 0.01000977, + "balance_loss_clip": 1.01316929, + "balance_loss_mlp": 0.99953479, + "epoch": 0.40774364215715747, + "flos": 71449307314560.0, + "grad_norm": 0.7995689452684275, + "language_loss": 0.56696904, + "learning_rate": 2.6809487221487343e-06, + "loss": 0.58773875, + "num_input_tokens_seen": 73258780, + "step": 3391, + "time_per_iteration": 4.0804431438446045 + }, + { + "auxiliary_loss_clip": 0.01186874, + "auxiliary_loss_mlp": 0.0102783, + "balance_loss_clip": 1.05061793, + "balance_loss_mlp": 1.01919293, + "epoch": 0.4078638850477965, + "flos": 15084134144640.0, + "grad_norm": 3.2970928757018925, + "language_loss": 0.82097125, + "learning_rate": 2.6802162394656605e-06, + "loss": 0.84311825, + "num_input_tokens_seen": 73275490, + "step": 3392, + "time_per_iteration": 2.6688954830169678 + }, + { + "auxiliary_loss_clip": 0.01200542, + "auxiliary_loss_mlp": 0.01030082, + "balance_loss_clip": 1.04885042, + "balance_loss_mlp": 1.02180886, + "epoch": 0.40798412793843564, + "flos": 23842890138240.0, + "grad_norm": 1.6497966233011037, + "language_loss": 0.7178672, + "learning_rate": 2.679483653594324e-06, + "loss": 0.74017346, + "num_input_tokens_seen": 73297260, + "step": 3393, + "time_per_iteration": 2.9294092655181885 + }, + { + "auxiliary_loss_clip": 0.01202837, + "auxiliary_loss_mlp": 0.01021002, + "balance_loss_clip": 1.05509079, + "balance_loss_mlp": 1.01306868, + "epoch": 0.40810437082907475, + "flos": 21065774117760.0, + "grad_norm": 2.014587384756688, + "language_loss": 0.76986015, + "learning_rate": 2.678750964645857e-06, + "loss": 0.79209858, + "num_input_tokens_seen": 73316340, + "step": 3394, + "time_per_iteration": 2.6368095874786377 + }, + { + "auxiliary_loss_clip": 0.01202284, + "auxiliary_loss_mlp": 0.010286, + "balance_loss_clip": 1.05767655, + "balance_loss_mlp": 1.01973629, + "epoch": 0.4082246137197138, + "flos": 11321377948800.0, + "grad_norm": 2.3310251387394874, + "language_loss": 0.83332366, + "learning_rate": 2.6780181727314094e-06, + "loss": 0.85563242, + "num_input_tokens_seen": 73331245, + "step": 3395, + "time_per_iteration": 2.7031641006469727 + }, + { + "auxiliary_loss_clip": 0.01217959, + "auxiliary_loss_mlp": 0.01379705, + "balance_loss_clip": 1.05050278, + "balance_loss_mlp": 1.00042593, + "epoch": 0.4083448566103529, + "flos": 19062569554560.0, + "grad_norm": 1.8272786486085397, + "language_loss": 0.78336018, + "learning_rate": 2.6772852779621435e-06, + "loss": 0.80933678, + "num_input_tokens_seen": 73349105, + "step": 3396, + "time_per_iteration": 2.731226682662964 + }, + { + "auxiliary_loss_clip": 0.01195631, + "auxiliary_loss_mlp": 0.01379498, + "balance_loss_clip": 1.05660415, + "balance_loss_mlp": 1.00046194, + "epoch": 0.408465099500992, + "flos": 23550254035200.0, + "grad_norm": 8.209908415193652, + "language_loss": 0.86751187, + "learning_rate": 2.676552280449239e-06, + "loss": 0.89326316, + "num_input_tokens_seen": 73368990, + "step": 3397, + "time_per_iteration": 2.757155656814575 + }, + { + "auxiliary_loss_clip": 0.01187342, + "auxiliary_loss_mlp": 0.0103339, + "balance_loss_clip": 1.05157566, + "balance_loss_mlp": 1.02545094, + "epoch": 0.4085853423916311, + "flos": 12750012558720.0, + "grad_norm": 2.7932425299604096, + "language_loss": 0.7526468, + "learning_rate": 2.6758191803038917e-06, + "loss": 0.77485412, + "num_input_tokens_seen": 73387485, + "step": 3398, + "time_per_iteration": 3.7207162380218506 + }, + { + "auxiliary_loss_clip": 0.01222668, + "auxiliary_loss_mlp": 0.01032182, + "balance_loss_clip": 1.04808259, + "balance_loss_mlp": 1.02346742, + "epoch": 0.4087055852822702, + "flos": 24353072962560.0, + "grad_norm": 1.5953093642068787, + "language_loss": 0.82992262, + "learning_rate": 2.6750859776373125e-06, + "loss": 0.85247111, + "num_input_tokens_seen": 73406940, + "step": 3399, + "time_per_iteration": 3.0622386932373047 + }, + { + "auxiliary_loss_clip": 0.01140934, + "auxiliary_loss_mlp": 0.00999984, + "balance_loss_clip": 1.01325703, + "balance_loss_mlp": 0.99860138, + "epoch": 0.4088258281729093, + "flos": 66387950720640.0, + "grad_norm": 0.771711846140246, + "language_loss": 0.60415214, + "learning_rate": 2.674352672560727e-06, + "loss": 0.62556136, + "num_input_tokens_seen": 73468385, + "step": 3400, + "time_per_iteration": 3.6253716945648193 + }, + { + "auxiliary_loss_clip": 0.01213284, + "auxiliary_loss_mlp": 0.01029028, + "balance_loss_clip": 1.05071425, + "balance_loss_mlp": 1.02080226, + "epoch": 0.40894607106354836, + "flos": 20449260057600.0, + "grad_norm": 1.6366878121295656, + "language_loss": 0.77107882, + "learning_rate": 2.673619265185377e-06, + "loss": 0.79350197, + "num_input_tokens_seen": 73488225, + "step": 3401, + "time_per_iteration": 3.1404645442962646 + }, + { + "auxiliary_loss_clip": 0.0120013, + "auxiliary_loss_mlp": 0.01024883, + "balance_loss_clip": 1.05275726, + "balance_loss_mlp": 1.01608539, + "epoch": 0.40906631395418747, + "flos": 27053627143680.0, + "grad_norm": 1.600111228447382, + "language_loss": 0.77969933, + "learning_rate": 2.672885755622521e-06, + "loss": 0.8019495, + "num_input_tokens_seen": 73510640, + "step": 3402, + "time_per_iteration": 2.7783877849578857 + }, + { + "auxiliary_loss_clip": 0.01221695, + "auxiliary_loss_mlp": 0.01031178, + "balance_loss_clip": 1.04859948, + "balance_loss_mlp": 1.0222671, + "epoch": 0.4091865568448266, + "flos": 25484151306240.0, + "grad_norm": 1.9664156411304472, + "language_loss": 0.70003545, + "learning_rate": 2.67215214398343e-06, + "loss": 0.72256416, + "num_input_tokens_seen": 73530655, + "step": 3403, + "time_per_iteration": 2.8297157287597656 + }, + { + "auxiliary_loss_clip": 0.01224254, + "auxiliary_loss_mlp": 0.0102801, + "balance_loss_clip": 1.04570639, + "balance_loss_mlp": 1.01950419, + "epoch": 0.40930679973546563, + "flos": 28657864368000.0, + "grad_norm": 2.2191004342458465, + "language_loss": 0.78623056, + "learning_rate": 2.671418430379393e-06, + "loss": 0.80875319, + "num_input_tokens_seen": 73549340, + "step": 3404, + "time_per_iteration": 2.8024959564208984 + }, + { + "auxiliary_loss_clip": 0.01187176, + "auxiliary_loss_mlp": 0.01026325, + "balance_loss_clip": 1.05255103, + "balance_loss_mlp": 1.01815879, + "epoch": 0.40942704262610474, + "flos": 20886292834560.0, + "grad_norm": 2.1902916125700247, + "language_loss": 0.83630848, + "learning_rate": 2.670684614921715e-06, + "loss": 0.8584435, + "num_input_tokens_seen": 73568315, + "step": 3405, + "time_per_iteration": 2.672760486602783 + }, + { + "auxiliary_loss_clip": 0.01205428, + "auxiliary_loss_mlp": 0.01026971, + "balance_loss_clip": 1.05011213, + "balance_loss_mlp": 1.01864982, + "epoch": 0.4095472855167438, + "flos": 21618080616960.0, + "grad_norm": 2.8227002785582713, + "language_loss": 0.69963342, + "learning_rate": 2.6699506977217128e-06, + "loss": 0.72195745, + "num_input_tokens_seen": 73588490, + "step": 3406, + "time_per_iteration": 2.717479705810547 + }, + { + "auxiliary_loss_clip": 0.01196595, + "auxiliary_loss_mlp": 0.01023261, + "balance_loss_clip": 1.05640757, + "balance_loss_mlp": 1.01490462, + "epoch": 0.4096675284073829, + "flos": 27926112499200.0, + "grad_norm": 2.0715016344311126, + "language_loss": 0.69953036, + "learning_rate": 2.6692166788907233e-06, + "loss": 0.72172892, + "num_input_tokens_seen": 73608685, + "step": 3407, + "time_per_iteration": 2.7516117095947266 + }, + { + "auxiliary_loss_clip": 0.01207894, + "auxiliary_loss_mlp": 0.01028154, + "balance_loss_clip": 1.05088973, + "balance_loss_mlp": 1.01948786, + "epoch": 0.409787771298022, + "flos": 19206607092480.0, + "grad_norm": 1.7833523599339705, + "language_loss": 0.76921016, + "learning_rate": 2.6684825585400957e-06, + "loss": 0.79157066, + "num_input_tokens_seen": 73627630, + "step": 3408, + "time_per_iteration": 2.6913795471191406 + }, + { + "auxiliary_loss_clip": 0.01098189, + "auxiliary_loss_mlp": 0.01007995, + "balance_loss_clip": 1.01425469, + "balance_loss_mlp": 1.00665414, + "epoch": 0.4099080141886611, + "flos": 59269234832640.0, + "grad_norm": 0.8151350283304957, + "language_loss": 0.65154117, + "learning_rate": 2.6677483367811947e-06, + "loss": 0.67260307, + "num_input_tokens_seen": 73687670, + "step": 3409, + "time_per_iteration": 3.356499433517456 + }, + { + "auxiliary_loss_clip": 0.01203189, + "auxiliary_loss_mlp": 0.010281, + "balance_loss_clip": 1.05377233, + "balance_loss_mlp": 1.0202086, + "epoch": 0.4100282570793002, + "flos": 21906443001600.0, + "grad_norm": 1.8656119963825792, + "language_loss": 0.75397396, + "learning_rate": 2.6670140137254028e-06, + "loss": 0.77628684, + "num_input_tokens_seen": 73707145, + "step": 3410, + "time_per_iteration": 2.693554162979126 + }, + { + "auxiliary_loss_clip": 0.01220039, + "auxiliary_loss_mlp": 0.0103005, + "balance_loss_clip": 1.04828691, + "balance_loss_mlp": 1.02217042, + "epoch": 0.4101484999699393, + "flos": 18551596631040.0, + "grad_norm": 2.2934094735908794, + "language_loss": 0.89269489, + "learning_rate": 2.666279589484115e-06, + "loss": 0.91519582, + "num_input_tokens_seen": 73725045, + "step": 3411, + "time_per_iteration": 2.8099889755249023 + }, + { + "auxiliary_loss_clip": 0.01220086, + "auxiliary_loss_mlp": 0.01028241, + "balance_loss_clip": 1.04721045, + "balance_loss_mlp": 1.02041519, + "epoch": 0.41026874286057835, + "flos": 19094529680640.0, + "grad_norm": 2.1479260817159562, + "language_loss": 0.81362677, + "learning_rate": 2.6655450641687435e-06, + "loss": 0.83611, + "num_input_tokens_seen": 73742610, + "step": 3412, + "time_per_iteration": 3.0079526901245117 + }, + { + "auxiliary_loss_clip": 0.01188675, + "auxiliary_loss_mlp": 0.01024044, + "balance_loss_clip": 1.05698776, + "balance_loss_mlp": 1.01574755, + "epoch": 0.41038898575121746, + "flos": 31209568588800.0, + "grad_norm": 1.6822411722061557, + "language_loss": 0.69498205, + "learning_rate": 2.664810437890715e-06, + "loss": 0.71710932, + "num_input_tokens_seen": 73764280, + "step": 3413, + "time_per_iteration": 2.71053409576416 + }, + { + "auxiliary_loss_clip": 0.01223027, + "auxiliary_loss_mlp": 0.0103263, + "balance_loss_clip": 1.05072224, + "balance_loss_mlp": 1.0245955, + "epoch": 0.41050922864185657, + "flos": 14355865895040.0, + "grad_norm": 2.2188841681589166, + "language_loss": 0.79835403, + "learning_rate": 2.6640757107614714e-06, + "loss": 0.82091069, + "num_input_tokens_seen": 73782375, + "step": 3414, + "time_per_iteration": 2.735241174697876 + }, + { + "auxiliary_loss_clip": 0.01206091, + "auxiliary_loss_mlp": 0.01029001, + "balance_loss_clip": 1.04947436, + "balance_loss_mlp": 1.0208652, + "epoch": 0.4106294715324956, + "flos": 30956290813440.0, + "grad_norm": 10.203036176170796, + "language_loss": 0.69655704, + "learning_rate": 2.6633408828924697e-06, + "loss": 0.71890795, + "num_input_tokens_seen": 73801240, + "step": 3415, + "time_per_iteration": 2.8856749534606934 + }, + { + "auxiliary_loss_clip": 0.01223812, + "auxiliary_loss_mlp": 0.01026561, + "balance_loss_clip": 1.05307317, + "balance_loss_mlp": 1.01835346, + "epoch": 0.41074971442313474, + "flos": 24457321209600.0, + "grad_norm": 1.533603710194307, + "language_loss": 0.70018935, + "learning_rate": 2.662605954395185e-06, + "loss": 0.72269309, + "num_input_tokens_seen": 73821200, + "step": 3416, + "time_per_iteration": 2.821831464767456 + }, + { + "auxiliary_loss_clip": 0.01199542, + "auxiliary_loss_mlp": 0.01028275, + "balance_loss_clip": 1.05274999, + "balance_loss_mlp": 1.02058005, + "epoch": 0.41086995731377385, + "flos": 21542991235200.0, + "grad_norm": 1.64364474027011, + "language_loss": 0.83629751, + "learning_rate": 2.6618709253811027e-06, + "loss": 0.8585757, + "num_input_tokens_seen": 73840655, + "step": 3417, + "time_per_iteration": 5.512585878372192 + }, + { + "auxiliary_loss_clip": 0.01184277, + "auxiliary_loss_mlp": 0.0103091, + "balance_loss_clip": 1.05452394, + "balance_loss_mlp": 1.02338159, + "epoch": 0.4109902002044129, + "flos": 20702753314560.0, + "grad_norm": 5.0688756196892095, + "language_loss": 0.87594807, + "learning_rate": 2.6611357959617277e-06, + "loss": 0.8980999, + "num_input_tokens_seen": 73860275, + "step": 3418, + "time_per_iteration": 2.764042615890503 + }, + { + "auxiliary_loss_clip": 0.01205398, + "auxiliary_loss_mlp": 0.01031092, + "balance_loss_clip": 1.0486623, + "balance_loss_mlp": 1.0227598, + "epoch": 0.411110443095052, + "flos": 18179992477440.0, + "grad_norm": 3.1170419369852698, + "language_loss": 0.91220236, + "learning_rate": 2.660400566248578e-06, + "loss": 0.93456721, + "num_input_tokens_seen": 73878400, + "step": 3419, + "time_per_iteration": 3.208944320678711 + }, + { + "auxiliary_loss_clip": 0.01215484, + "auxiliary_loss_mlp": 0.01030962, + "balance_loss_clip": 1.05085874, + "balance_loss_mlp": 1.02218771, + "epoch": 0.41123068598569107, + "flos": 14575244209920.0, + "grad_norm": 2.669164252059264, + "language_loss": 0.66878355, + "learning_rate": 2.6596652363531876e-06, + "loss": 0.691248, + "num_input_tokens_seen": 73894275, + "step": 3420, + "time_per_iteration": 2.7094786167144775 + }, + { + "auxiliary_loss_clip": 0.01187703, + "auxiliary_loss_mlp": 0.01026208, + "balance_loss_clip": 1.05467641, + "balance_loss_mlp": 1.01841164, + "epoch": 0.4113509288763302, + "flos": 21177995184000.0, + "grad_norm": 1.479731842075914, + "language_loss": 0.78175032, + "learning_rate": 2.6589298063871055e-06, + "loss": 0.80388945, + "num_input_tokens_seen": 73914450, + "step": 3421, + "time_per_iteration": 2.712232828140259 + }, + { + "auxiliary_loss_clip": 0.01187431, + "auxiliary_loss_mlp": 0.01031112, + "balance_loss_clip": 1.05564308, + "balance_loss_mlp": 1.02305984, + "epoch": 0.4114711717669693, + "flos": 18442212739200.0, + "grad_norm": 1.864308548369914, + "language_loss": 0.69835603, + "learning_rate": 2.658194276461895e-06, + "loss": 0.72054142, + "num_input_tokens_seen": 73932375, + "step": 3422, + "time_per_iteration": 2.872340202331543 + }, + { + "auxiliary_loss_clip": 0.01202584, + "auxiliary_loss_mlp": 0.01032979, + "balance_loss_clip": 1.04800701, + "balance_loss_mlp": 1.02394938, + "epoch": 0.41159141465760835, + "flos": 27233395735680.0, + "grad_norm": 2.078496262809547, + "language_loss": 0.66749018, + "learning_rate": 2.6574586466891368e-06, + "loss": 0.6898458, + "num_input_tokens_seen": 73952850, + "step": 3423, + "time_per_iteration": 2.9180591106414795 + }, + { + "auxiliary_loss_clip": 0.01207351, + "auxiliary_loss_mlp": 0.01379074, + "balance_loss_clip": 1.05115938, + "balance_loss_mlp": 1.00041676, + "epoch": 0.41171165754824746, + "flos": 20006876154240.0, + "grad_norm": 2.110125714177547, + "language_loss": 0.64789975, + "learning_rate": 2.6567229171804247e-06, + "loss": 0.67376399, + "num_input_tokens_seen": 73970735, + "step": 3424, + "time_per_iteration": 3.691019296646118 + }, + { + "auxiliary_loss_clip": 0.01218521, + "auxiliary_loss_mlp": 0.01027609, + "balance_loss_clip": 1.04635704, + "balance_loss_mlp": 1.01918674, + "epoch": 0.41183190043888657, + "flos": 18004318035840.0, + "grad_norm": 2.5552476239584765, + "language_loss": 0.87840778, + "learning_rate": 2.655987088047368e-06, + "loss": 0.90086913, + "num_input_tokens_seen": 73989080, + "step": 3425, + "time_per_iteration": 2.813671112060547 + }, + { + "auxiliary_loss_clip": 0.01200566, + "auxiliary_loss_mlp": 0.0102853, + "balance_loss_clip": 1.0508604, + "balance_loss_mlp": 1.02025056, + "epoch": 0.4119521433295256, + "flos": 27163370171520.0, + "grad_norm": 2.01156157001667, + "language_loss": 0.78910196, + "learning_rate": 2.6552511594015912e-06, + "loss": 0.8113929, + "num_input_tokens_seen": 74009470, + "step": 3426, + "time_per_iteration": 2.8940415382385254 + }, + { + "auxiliary_loss_clip": 0.01201006, + "auxiliary_loss_mlp": 0.0102736, + "balance_loss_clip": 1.04812455, + "balance_loss_mlp": 1.01831186, + "epoch": 0.41207238622016473, + "flos": 15122020014720.0, + "grad_norm": 2.485940813283007, + "language_loss": 0.8521955, + "learning_rate": 2.654515131354735e-06, + "loss": 0.87447917, + "num_input_tokens_seen": 74027735, + "step": 3427, + "time_per_iteration": 2.651275873184204 + }, + { + "auxiliary_loss_clip": 0.01218547, + "auxiliary_loss_mlp": 0.01032274, + "balance_loss_clip": 1.05309153, + "balance_loss_mlp": 1.02379823, + "epoch": 0.41219262911080384, + "flos": 27052872958080.0, + "grad_norm": 1.800905393913781, + "language_loss": 0.85454088, + "learning_rate": 2.653779004018453e-06, + "loss": 0.87704903, + "num_input_tokens_seen": 74048300, + "step": 3428, + "time_per_iteration": 2.752397060394287 + }, + { + "auxiliary_loss_clip": 0.01195705, + "auxiliary_loss_mlp": 0.01026327, + "balance_loss_clip": 1.04985809, + "balance_loss_mlp": 1.018507, + "epoch": 0.4123128720014429, + "flos": 24686360282880.0, + "grad_norm": 1.8159327681801933, + "language_loss": 0.81988585, + "learning_rate": 2.653042777504417e-06, + "loss": 0.84210622, + "num_input_tokens_seen": 74070890, + "step": 3429, + "time_per_iteration": 2.7150356769561768 + }, + { + "auxiliary_loss_clip": 0.01213578, + "auxiliary_loss_mlp": 0.01033254, + "balance_loss_clip": 1.05191028, + "balance_loss_mlp": 1.02437329, + "epoch": 0.412433114892082, + "flos": 26244774731520.0, + "grad_norm": 1.6974683326309137, + "language_loss": 0.79685104, + "learning_rate": 2.6523064519243105e-06, + "loss": 0.81931937, + "num_input_tokens_seen": 74090460, + "step": 3430, + "time_per_iteration": 2.700160026550293 + }, + { + "auxiliary_loss_clip": 0.01198792, + "auxiliary_loss_mlp": 0.01033421, + "balance_loss_clip": 1.05453014, + "balance_loss_mlp": 1.02524328, + "epoch": 0.4125533577827211, + "flos": 21361031913600.0, + "grad_norm": 1.9894355173751834, + "language_loss": 0.79275322, + "learning_rate": 2.6515700273898333e-06, + "loss": 0.8150754, + "num_input_tokens_seen": 74108335, + "step": 3431, + "time_per_iteration": 2.6152024269104004 + }, + { + "auxiliary_loss_clip": 0.01195146, + "auxiliary_loss_mlp": 0.01031083, + "balance_loss_clip": 1.0529139, + "balance_loss_mlp": 1.02269101, + "epoch": 0.4126736006733602, + "flos": 26067556005120.0, + "grad_norm": 1.9967506329501028, + "language_loss": 0.68912137, + "learning_rate": 2.6508335040127018e-06, + "loss": 0.71138364, + "num_input_tokens_seen": 74128030, + "step": 3432, + "time_per_iteration": 2.735490322113037 + }, + { + "auxiliary_loss_clip": 0.01201589, + "auxiliary_loss_mlp": 0.01029905, + "balance_loss_clip": 1.05377507, + "balance_loss_mlp": 1.02191234, + "epoch": 0.4127938435639993, + "flos": 25666146541440.0, + "grad_norm": 1.5277391704053904, + "language_loss": 0.77370036, + "learning_rate": 2.6500968819046446e-06, + "loss": 0.79601538, + "num_input_tokens_seen": 74148330, + "step": 3433, + "time_per_iteration": 2.722261667251587 + }, + { + "auxiliary_loss_clip": 0.01197295, + "auxiliary_loss_mlp": 0.01029055, + "balance_loss_clip": 1.04540455, + "balance_loss_mlp": 1.02166426, + "epoch": 0.4129140864546384, + "flos": 17995914253440.0, + "grad_norm": 2.5056609853565557, + "language_loss": 0.59298784, + "learning_rate": 2.649360161177408e-06, + "loss": 0.6152513, + "num_input_tokens_seen": 74163390, + "step": 3434, + "time_per_iteration": 2.6706948280334473 + }, + { + "auxiliary_loss_clip": 0.01201838, + "auxiliary_loss_mlp": 0.01026091, + "balance_loss_clip": 1.05225134, + "balance_loss_mlp": 1.0173527, + "epoch": 0.41303432934527745, + "flos": 23732895715200.0, + "grad_norm": 1.7277683483470918, + "language_loss": 0.73422283, + "learning_rate": 2.6486233419427504e-06, + "loss": 0.75650215, + "num_input_tokens_seen": 74183205, + "step": 3435, + "time_per_iteration": 2.672712564468384 + }, + { + "auxiliary_loss_clip": 0.01201944, + "auxiliary_loss_mlp": 0.01033295, + "balance_loss_clip": 1.05087376, + "balance_loss_mlp": 1.02462256, + "epoch": 0.41315457223591656, + "flos": 19755286318080.0, + "grad_norm": 2.065553439936676, + "language_loss": 0.75140262, + "learning_rate": 2.6478864243124484e-06, + "loss": 0.77375495, + "num_input_tokens_seen": 74202870, + "step": 3436, + "time_per_iteration": 2.704228162765503 + }, + { + "auxiliary_loss_clip": 0.01194815, + "auxiliary_loss_mlp": 0.01029409, + "balance_loss_clip": 1.05014014, + "balance_loss_mlp": 1.02102256, + "epoch": 0.4132748151265556, + "flos": 20923316778240.0, + "grad_norm": 2.2366867513589246, + "language_loss": 0.85391927, + "learning_rate": 2.6471494083982903e-06, + "loss": 0.87616146, + "num_input_tokens_seen": 74222255, + "step": 3437, + "time_per_iteration": 2.7486050128936768 + }, + { + "auxiliary_loss_clip": 0.01215804, + "auxiliary_loss_mlp": 0.01027112, + "balance_loss_clip": 1.04853678, + "balance_loss_mlp": 1.01904774, + "epoch": 0.4133950580171947, + "flos": 32232520016640.0, + "grad_norm": 1.720131492170453, + "language_loss": 0.74724275, + "learning_rate": 2.6464122943120818e-06, + "loss": 0.76967192, + "num_input_tokens_seen": 74242480, + "step": 3438, + "time_per_iteration": 2.811661958694458 + }, + { + "auxiliary_loss_clip": 0.01212741, + "auxiliary_loss_mlp": 0.01032748, + "balance_loss_clip": 1.051126, + "balance_loss_mlp": 1.02460039, + "epoch": 0.41351530090783384, + "flos": 23292487059840.0, + "grad_norm": 3.707660950141741, + "language_loss": 0.826141, + "learning_rate": 2.645675082165642e-06, + "loss": 0.84859592, + "num_input_tokens_seen": 74258690, + "step": 3439, + "time_per_iteration": 2.8651278018951416 + }, + { + "auxiliary_loss_clip": 0.01205336, + "auxiliary_loss_mlp": 0.01040165, + "balance_loss_clip": 1.05167758, + "balance_loss_mlp": 1.0320586, + "epoch": 0.4136355437984729, + "flos": 25593571111680.0, + "grad_norm": 2.041983846223122, + "language_loss": 0.75659931, + "learning_rate": 2.644937772070806e-06, + "loss": 0.77905428, + "num_input_tokens_seen": 74277135, + "step": 3440, + "time_per_iteration": 2.7659056186676025 + }, + { + "auxiliary_loss_clip": 0.01190835, + "auxiliary_loss_mlp": 0.0102666, + "balance_loss_clip": 1.05651021, + "balance_loss_mlp": 1.01854753, + "epoch": 0.413755786689112, + "flos": 19828615933440.0, + "grad_norm": 2.500977137684729, + "language_loss": 0.83750218, + "learning_rate": 2.6442003641394225e-06, + "loss": 0.85967714, + "num_input_tokens_seen": 74294730, + "step": 3441, + "time_per_iteration": 2.6319055557250977 + }, + { + "auxiliary_loss_clip": 0.01200607, + "auxiliary_loss_mlp": 0.01029925, + "balance_loss_clip": 1.04950821, + "balance_loss_mlp": 1.02241516, + "epoch": 0.4138760295797511, + "flos": 26870446759680.0, + "grad_norm": 1.4350297262070715, + "language_loss": 0.84200335, + "learning_rate": 2.643462858483356e-06, + "loss": 0.86430871, + "num_input_tokens_seen": 74315015, + "step": 3442, + "time_per_iteration": 3.6295430660247803 + }, + { + "auxiliary_loss_clip": 0.0121645, + "auxiliary_loss_mlp": 0.01030552, + "balance_loss_clip": 1.0472703, + "balance_loss_mlp": 1.02182591, + "epoch": 0.41399627247039017, + "flos": 16399254798720.0, + "grad_norm": 1.9287557261580761, + "language_loss": 0.72770005, + "learning_rate": 2.6427252552144856e-06, + "loss": 0.75017011, + "num_input_tokens_seen": 74333665, + "step": 3443, + "time_per_iteration": 4.716650009155273 + }, + { + "auxiliary_loss_clip": 0.011886, + "auxiliary_loss_mlp": 0.01030342, + "balance_loss_clip": 1.05444098, + "balance_loss_mlp": 1.02127659, + "epoch": 0.4141165153610293, + "flos": 22930220442240.0, + "grad_norm": 2.0710385311808923, + "language_loss": 0.75170314, + "learning_rate": 2.6419875544447044e-06, + "loss": 0.77389252, + "num_input_tokens_seen": 74355065, + "step": 3444, + "time_per_iteration": 2.6641647815704346 + }, + { + "auxiliary_loss_clip": 0.01189486, + "auxiliary_loss_mlp": 0.01027561, + "balance_loss_clip": 1.05464447, + "balance_loss_mlp": 1.01947832, + "epoch": 0.4142367582516684, + "flos": 25192556697600.0, + "grad_norm": 1.7337846194868032, + "language_loss": 0.7172122, + "learning_rate": 2.6412497562859218e-06, + "loss": 0.73938268, + "num_input_tokens_seen": 74376345, + "step": 3445, + "time_per_iteration": 2.6600332260131836 + }, + { + "auxiliary_loss_clip": 0.0120083, + "auxiliary_loss_mlp": 0.01031199, + "balance_loss_clip": 1.05177927, + "balance_loss_mlp": 1.02311099, + "epoch": 0.41435700114230745, + "flos": 21690476478720.0, + "grad_norm": 2.168725442612817, + "language_loss": 0.75880009, + "learning_rate": 2.6405118608500617e-06, + "loss": 0.78112036, + "num_input_tokens_seen": 74395170, + "step": 3446, + "time_per_iteration": 2.6451661586761475 + }, + { + "auxiliary_loss_clip": 0.01209731, + "auxiliary_loss_mlp": 0.01028719, + "balance_loss_clip": 1.05397654, + "balance_loss_mlp": 1.02102995, + "epoch": 0.41447724403294656, + "flos": 25995160143360.0, + "grad_norm": 1.7399846118843478, + "language_loss": 0.8153553, + "learning_rate": 2.6397738682490613e-06, + "loss": 0.83773983, + "num_input_tokens_seen": 74416070, + "step": 3447, + "time_per_iteration": 2.7872979640960693 + }, + { + "auxiliary_loss_clip": 0.0118705, + "auxiliary_loss_mlp": 0.01024019, + "balance_loss_clip": 1.05295181, + "balance_loss_mlp": 1.01584423, + "epoch": 0.41459748692358567, + "flos": 18259678800000.0, + "grad_norm": 1.681168694219274, + "language_loss": 0.75343156, + "learning_rate": 2.6390357785948734e-06, + "loss": 0.77554226, + "num_input_tokens_seen": 74433185, + "step": 3448, + "time_per_iteration": 2.5828216075897217 + }, + { + "auxiliary_loss_clip": 0.01198059, + "auxiliary_loss_mlp": 0.0103218, + "balance_loss_clip": 1.05345523, + "balance_loss_mlp": 1.02348351, + "epoch": 0.4147177298142247, + "flos": 24168456034560.0, + "grad_norm": 1.682749603571415, + "language_loss": 0.80404335, + "learning_rate": 2.6382975919994667e-06, + "loss": 0.82634568, + "num_input_tokens_seen": 74453760, + "step": 3449, + "time_per_iteration": 2.757143020629883 + }, + { + "auxiliary_loss_clip": 0.01204617, + "auxiliary_loss_mlp": 0.01026351, + "balance_loss_clip": 1.04925013, + "balance_loss_mlp": 1.01926446, + "epoch": 0.41483797270486383, + "flos": 20084659056000.0, + "grad_norm": 1.658252795177755, + "language_loss": 0.73068678, + "learning_rate": 2.637559308574822e-06, + "loss": 0.75299644, + "num_input_tokens_seen": 74473505, + "step": 3450, + "time_per_iteration": 3.5558228492736816 + }, + { + "auxiliary_loss_clip": 0.01187357, + "auxiliary_loss_mlp": 0.01026076, + "balance_loss_clip": 1.05448747, + "balance_loss_mlp": 1.01835179, + "epoch": 0.4149582155955029, + "flos": 30081040110720.0, + "grad_norm": 4.746911007180487, + "language_loss": 0.71355438, + "learning_rate": 2.6368209284329376e-06, + "loss": 0.73568881, + "num_input_tokens_seen": 74494135, + "step": 3451, + "time_per_iteration": 2.6828298568725586 + }, + { + "auxiliary_loss_clip": 0.01194601, + "auxiliary_loss_mlp": 0.01030407, + "balance_loss_clip": 1.05138326, + "balance_loss_mlp": 1.02267587, + "epoch": 0.415078458486142, + "flos": 16764394504320.0, + "grad_norm": 1.8638737549216946, + "language_loss": 0.75619918, + "learning_rate": 2.636082451685825e-06, + "loss": 0.77844924, + "num_input_tokens_seen": 74512335, + "step": 3452, + "time_per_iteration": 2.622297525405884 + }, + { + "auxiliary_loss_clip": 0.01212351, + "auxiliary_loss_mlp": 0.01026323, + "balance_loss_clip": 1.05540073, + "balance_loss_mlp": 1.01861048, + "epoch": 0.4151987013767811, + "flos": 26033692458240.0, + "grad_norm": 1.500676047819425, + "language_loss": 0.86631602, + "learning_rate": 2.6353438784455094e-06, + "loss": 0.88870275, + "num_input_tokens_seen": 74535620, + "step": 3453, + "time_per_iteration": 2.75594425201416 + }, + { + "auxiliary_loss_clip": 0.01201087, + "auxiliary_loss_mlp": 0.01028407, + "balance_loss_clip": 1.05155003, + "balance_loss_mlp": 1.01962125, + "epoch": 0.41531894426742016, + "flos": 24608002763520.0, + "grad_norm": 2.898982427893938, + "language_loss": 0.71559393, + "learning_rate": 2.6346052088240326e-06, + "loss": 0.73788893, + "num_input_tokens_seen": 74555140, + "step": 3454, + "time_per_iteration": 2.753369092941284 + }, + { + "auxiliary_loss_clip": 0.01206397, + "auxiliary_loss_mlp": 0.01028683, + "balance_loss_clip": 1.05131292, + "balance_loss_mlp": 1.01997519, + "epoch": 0.4154391871580593, + "flos": 14975791747200.0, + "grad_norm": 2.438404899284516, + "language_loss": 0.7708357, + "learning_rate": 2.63386644293345e-06, + "loss": 0.79318655, + "num_input_tokens_seen": 74571485, + "step": 3455, + "time_per_iteration": 2.630821704864502 + }, + { + "auxiliary_loss_clip": 0.01206158, + "auxiliary_loss_mlp": 0.01029379, + "balance_loss_clip": 1.04759932, + "balance_loss_mlp": 1.02150488, + "epoch": 0.4155594300486984, + "flos": 14647173194880.0, + "grad_norm": 2.181352954359362, + "language_loss": 0.82779402, + "learning_rate": 2.633127580885833e-06, + "loss": 0.85014939, + "num_input_tokens_seen": 74585985, + "step": 3456, + "time_per_iteration": 2.7460341453552246 + }, + { + "auxiliary_loss_clip": 0.01187253, + "auxiliary_loss_mlp": 0.01030379, + "balance_loss_clip": 1.05615401, + "balance_loss_mlp": 1.02265763, + "epoch": 0.41567967293933744, + "flos": 29497276275840.0, + "grad_norm": 4.406912470410802, + "language_loss": 0.64860368, + "learning_rate": 2.632388622793265e-06, + "loss": 0.67078006, + "num_input_tokens_seen": 74605140, + "step": 3457, + "time_per_iteration": 2.7521517276763916 + }, + { + "auxiliary_loss_clip": 0.01196127, + "auxiliary_loss_mlp": 0.01034232, + "balance_loss_clip": 1.05502987, + "balance_loss_mlp": 1.02605772, + "epoch": 0.41579991582997655, + "flos": 19238387650560.0, + "grad_norm": 1.7088214291503527, + "language_loss": 0.68173695, + "learning_rate": 2.6316495687678457e-06, + "loss": 0.70404059, + "num_input_tokens_seen": 74623790, + "step": 3458, + "time_per_iteration": 2.6478748321533203 + }, + { + "auxiliary_loss_clip": 0.01214423, + "auxiliary_loss_mlp": 0.01030636, + "balance_loss_clip": 1.04643881, + "balance_loss_mlp": 1.02252984, + "epoch": 0.41592015872061566, + "flos": 24462061804800.0, + "grad_norm": 2.226290221428258, + "language_loss": 0.7658577, + "learning_rate": 2.6309104189216887e-06, + "loss": 0.78830826, + "num_input_tokens_seen": 74641355, + "step": 3459, + "time_per_iteration": 2.744542121887207 + }, + { + "auxiliary_loss_clip": 0.01198943, + "auxiliary_loss_mlp": 0.0137941, + "balance_loss_clip": 1.04849482, + "balance_loss_mlp": 1.00032842, + "epoch": 0.4160404016112547, + "flos": 20775651966720.0, + "grad_norm": 4.130141062434331, + "language_loss": 0.74354899, + "learning_rate": 2.630171173366923e-06, + "loss": 0.76933253, + "num_input_tokens_seen": 74657155, + "step": 3460, + "time_per_iteration": 2.6991186141967773 + }, + { + "auxiliary_loss_clip": 0.01221803, + "auxiliary_loss_mlp": 0.01029038, + "balance_loss_clip": 1.04942381, + "balance_loss_mlp": 1.02079487, + "epoch": 0.41616064450189383, + "flos": 13916462820480.0, + "grad_norm": 2.501213498360814, + "language_loss": 0.74332726, + "learning_rate": 2.629431832215691e-06, + "loss": 0.7658357, + "num_input_tokens_seen": 74671960, + "step": 3461, + "time_per_iteration": 2.690575361251831 + }, + { + "auxiliary_loss_clip": 0.01200693, + "auxiliary_loss_mlp": 0.0103315, + "balance_loss_clip": 1.05036747, + "balance_loss_mlp": 1.02506208, + "epoch": 0.41628088739253294, + "flos": 20010826650240.0, + "grad_norm": 1.6153273028566024, + "language_loss": 0.86842799, + "learning_rate": 2.628692395580151e-06, + "loss": 0.89076638, + "num_input_tokens_seen": 74692050, + "step": 3462, + "time_per_iteration": 2.6694610118865967 + }, + { + "auxiliary_loss_clip": 0.01208138, + "auxiliary_loss_mlp": 0.01026929, + "balance_loss_clip": 1.0443821, + "balance_loss_mlp": 1.01832771, + "epoch": 0.416401130283172, + "flos": 29168801377920.0, + "grad_norm": 1.6375555415946759, + "language_loss": 0.79305333, + "learning_rate": 2.6279528635724747e-06, + "loss": 0.815404, + "num_input_tokens_seen": 74712205, + "step": 3463, + "time_per_iteration": 2.821949005126953 + }, + { + "auxiliary_loss_clip": 0.01193574, + "auxiliary_loss_mlp": 0.01028286, + "balance_loss_clip": 1.04967713, + "balance_loss_mlp": 1.01948237, + "epoch": 0.4165213731738111, + "flos": 16246813478400.0, + "grad_norm": 2.428302421844926, + "language_loss": 0.78976142, + "learning_rate": 2.627213236304848e-06, + "loss": 0.81198001, + "num_input_tokens_seen": 74729005, + "step": 3464, + "time_per_iteration": 2.6347429752349854 + }, + { + "auxiliary_loss_clip": 0.01198601, + "auxiliary_loss_mlp": 0.0103186, + "balance_loss_clip": 1.05336428, + "balance_loss_mlp": 1.02381921, + "epoch": 0.4166416160644502, + "flos": 33765438787200.0, + "grad_norm": 2.2583500536290484, + "language_loss": 0.70779884, + "learning_rate": 2.626473513889472e-06, + "loss": 0.73010343, + "num_input_tokens_seen": 74751385, + "step": 3465, + "time_per_iteration": 2.751121759414673 + }, + { + "auxiliary_loss_clip": 0.011863, + "auxiliary_loss_mlp": 0.01028316, + "balance_loss_clip": 1.05073261, + "balance_loss_mlp": 1.02067208, + "epoch": 0.41676185895508927, + "flos": 20917498775040.0, + "grad_norm": 1.7960620086749977, + "language_loss": 0.82488525, + "learning_rate": 2.625733696438562e-06, + "loss": 0.84703147, + "num_input_tokens_seen": 74768890, + "step": 3466, + "time_per_iteration": 2.7381417751312256 + }, + { + "auxiliary_loss_clip": 0.0120254, + "auxiliary_loss_mlp": 0.01037456, + "balance_loss_clip": 1.05186391, + "balance_loss_mlp": 1.02921903, + "epoch": 0.4168821018457284, + "flos": 18406122549120.0, + "grad_norm": 1.639370510428344, + "language_loss": 0.752446, + "learning_rate": 2.6249937840643476e-06, + "loss": 0.77484602, + "num_input_tokens_seen": 74787195, + "step": 3467, + "time_per_iteration": 2.717050313949585 + }, + { + "auxiliary_loss_clip": 0.0119021, + "auxiliary_loss_mlp": 0.01379345, + "balance_loss_clip": 1.05773103, + "balance_loss_mlp": 1.00038493, + "epoch": 0.41700234473636744, + "flos": 18698399516160.0, + "grad_norm": 2.0300511707126736, + "language_loss": 0.66843677, + "learning_rate": 2.6242537768790733e-06, + "loss": 0.69413227, + "num_input_tokens_seen": 74806350, + "step": 3468, + "time_per_iteration": 2.578343391418457 + }, + { + "auxiliary_loss_clip": 0.01201239, + "auxiliary_loss_mlp": 0.01031018, + "balance_loss_clip": 1.05632854, + "balance_loss_mlp": 1.02239943, + "epoch": 0.41712258762700655, + "flos": 31033283616000.0, + "grad_norm": 2.743228931542533, + "language_loss": 0.68381059, + "learning_rate": 2.6235136749949975e-06, + "loss": 0.70613313, + "num_input_tokens_seen": 74829800, + "step": 3469, + "time_per_iteration": 4.631229400634766 + }, + { + "auxiliary_loss_clip": 0.0118772, + "auxiliary_loss_mlp": 0.01025754, + "balance_loss_clip": 1.05559134, + "balance_loss_mlp": 1.01785338, + "epoch": 0.41724283051764566, + "flos": 35914763877120.0, + "grad_norm": 1.9843961646621155, + "language_loss": 0.6158818, + "learning_rate": 2.6227734785243924e-06, + "loss": 0.63801652, + "num_input_tokens_seen": 74849760, + "step": 3470, + "time_per_iteration": 2.7451395988464355 + }, + { + "auxiliary_loss_clip": 0.01219148, + "auxiliary_loss_mlp": 0.01027724, + "balance_loss_clip": 1.04405332, + "balance_loss_mlp": 1.01994538, + "epoch": 0.4173630734082847, + "flos": 25333649320320.0, + "grad_norm": 2.2520871019389586, + "language_loss": 0.79062271, + "learning_rate": 2.6220331875795466e-06, + "loss": 0.81309146, + "num_input_tokens_seen": 74869110, + "step": 3471, + "time_per_iteration": 2.883927822113037 + }, + { + "auxiliary_loss_clip": 0.01194073, + "auxiliary_loss_mlp": 0.01035223, + "balance_loss_clip": 1.05519366, + "balance_loss_mlp": 1.02754557, + "epoch": 0.4174833162989238, + "flos": 26685398868480.0, + "grad_norm": 1.605602535011797, + "language_loss": 0.75286555, + "learning_rate": 2.62129280227276e-06, + "loss": 0.77515846, + "num_input_tokens_seen": 74889110, + "step": 3472, + "time_per_iteration": 2.7340004444122314 + }, + { + "auxiliary_loss_clip": 0.01201359, + "auxiliary_loss_mlp": 0.01026602, + "balance_loss_clip": 1.05443358, + "balance_loss_mlp": 1.01834118, + "epoch": 0.41760355918956293, + "flos": 74739584010240.0, + "grad_norm": 2.056692663716243, + "language_loss": 0.68778783, + "learning_rate": 2.62055232271635e-06, + "loss": 0.71006739, + "num_input_tokens_seen": 74916260, + "step": 3473, + "time_per_iteration": 3.0999252796173096 + }, + { + "auxiliary_loss_clip": 0.01200524, + "auxiliary_loss_mlp": 0.01029183, + "balance_loss_clip": 1.04834521, + "balance_loss_mlp": 1.02096343, + "epoch": 0.417723802080202, + "flos": 14317513148160.0, + "grad_norm": 2.2364820391696374, + "language_loss": 0.87897861, + "learning_rate": 2.619811749022646e-06, + "loss": 0.90127563, + "num_input_tokens_seen": 74931570, + "step": 3474, + "time_per_iteration": 2.7314436435699463 + }, + { + "auxiliary_loss_clip": 0.01201545, + "auxiliary_loss_mlp": 0.0103332, + "balance_loss_clip": 1.0562489, + "balance_loss_mlp": 1.02442718, + "epoch": 0.4178440449708411, + "flos": 14643797316480.0, + "grad_norm": 2.7433844245072634, + "language_loss": 0.71298343, + "learning_rate": 2.6190710813039917e-06, + "loss": 0.73533213, + "num_input_tokens_seen": 74944695, + "step": 3475, + "time_per_iteration": 2.5906424522399902 + }, + { + "auxiliary_loss_clip": 0.01210928, + "auxiliary_loss_mlp": 0.01379415, + "balance_loss_clip": 1.04340768, + "balance_loss_mlp": 1.00026131, + "epoch": 0.4179642878614802, + "flos": 21507296094720.0, + "grad_norm": 2.7700759201379164, + "language_loss": 0.83771038, + "learning_rate": 2.618330319672747e-06, + "loss": 0.86361378, + "num_input_tokens_seen": 74964115, + "step": 3476, + "time_per_iteration": 3.7049691677093506 + }, + { + "auxiliary_loss_clip": 0.01188905, + "auxiliary_loss_mlp": 0.01030558, + "balance_loss_clip": 1.05598474, + "balance_loss_mlp": 1.02264869, + "epoch": 0.41808453075211927, + "flos": 18441997257600.0, + "grad_norm": 2.1151414290897836, + "language_loss": 0.91899395, + "learning_rate": 2.617589464241284e-06, + "loss": 0.94118857, + "num_input_tokens_seen": 74978515, + "step": 3477, + "time_per_iteration": 2.650562286376953 + }, + { + "auxiliary_loss_clip": 0.01217552, + "auxiliary_loss_mlp": 0.01034283, + "balance_loss_clip": 1.05140662, + "balance_loss_mlp": 1.02665985, + "epoch": 0.4182047736427584, + "flos": 20301020628480.0, + "grad_norm": 3.106994180365151, + "language_loss": 0.74496382, + "learning_rate": 2.6168485151219914e-06, + "loss": 0.76748216, + "num_input_tokens_seen": 74998135, + "step": 3478, + "time_per_iteration": 2.7051150798797607 + }, + { + "auxiliary_loss_clip": 0.01197994, + "auxiliary_loss_mlp": 0.01026541, + "balance_loss_clip": 1.05473685, + "balance_loss_mlp": 1.01856029, + "epoch": 0.4183250165333975, + "flos": 18876623823360.0, + "grad_norm": 2.125479910635409, + "language_loss": 0.71784645, + "learning_rate": 2.616107472427269e-06, + "loss": 0.7400918, + "num_input_tokens_seen": 75012830, + "step": 3479, + "time_per_iteration": 2.681218147277832 + }, + { + "auxiliary_loss_clip": 0.01200446, + "auxiliary_loss_mlp": 0.01034846, + "balance_loss_clip": 1.05275393, + "balance_loss_mlp": 1.02656078, + "epoch": 0.41844525942403654, + "flos": 17740050698880.0, + "grad_norm": 2.491303690695048, + "language_loss": 0.76996982, + "learning_rate": 2.615366336269533e-06, + "loss": 0.79232275, + "num_input_tokens_seen": 75026495, + "step": 3480, + "time_per_iteration": 2.646898031234741 + }, + { + "auxiliary_loss_clip": 0.01192041, + "auxiliary_loss_mlp": 0.0102916, + "balance_loss_clip": 1.05706179, + "balance_loss_mlp": 1.02030218, + "epoch": 0.41856550231467565, + "flos": 18361377181440.0, + "grad_norm": 2.533834134086794, + "language_loss": 0.80119807, + "learning_rate": 2.6146251067612126e-06, + "loss": 0.82341003, + "num_input_tokens_seen": 75041970, + "step": 3481, + "time_per_iteration": 2.6037960052490234 + }, + { + "auxiliary_loss_clip": 0.01198304, + "auxiliary_loss_mlp": 0.01031687, + "balance_loss_clip": 1.05618453, + "balance_loss_mlp": 1.02374816, + "epoch": 0.41868574520531476, + "flos": 22781801445120.0, + "grad_norm": 1.6273591980821482, + "language_loss": 0.82825863, + "learning_rate": 2.6138837840147525e-06, + "loss": 0.85055852, + "num_input_tokens_seen": 75061005, + "step": 3482, + "time_per_iteration": 2.6585676670074463 + }, + { + "auxiliary_loss_clip": 0.01212301, + "auxiliary_loss_mlp": 0.01029706, + "balance_loss_clip": 1.05050611, + "balance_loss_mlp": 1.02158761, + "epoch": 0.4188059880959538, + "flos": 13699167494400.0, + "grad_norm": 1.8822138406921272, + "language_loss": 0.76084185, + "learning_rate": 2.6131423681426103e-06, + "loss": 0.7832619, + "num_input_tokens_seen": 75076920, + "step": 3483, + "time_per_iteration": 2.785022497177124 + }, + { + "auxiliary_loss_clip": 0.01189799, + "auxiliary_loss_mlp": 0.01031181, + "balance_loss_clip": 1.05646384, + "balance_loss_mlp": 1.02340603, + "epoch": 0.41892623098659293, + "flos": 37818281220480.0, + "grad_norm": 1.9661154616519458, + "language_loss": 0.72923505, + "learning_rate": 2.6124008592572587e-06, + "loss": 0.75144494, + "num_input_tokens_seen": 75100905, + "step": 3484, + "time_per_iteration": 2.737288236618042 + }, + { + "auxiliary_loss_clip": 0.0119296, + "auxiliary_loss_mlp": 0.01026502, + "balance_loss_clip": 1.05625606, + "balance_loss_mlp": 1.01807415, + "epoch": 0.419046473877232, + "flos": 23258874908160.0, + "grad_norm": 2.090643650228593, + "language_loss": 0.81396431, + "learning_rate": 2.6116592574711835e-06, + "loss": 0.83615887, + "num_input_tokens_seen": 75119205, + "step": 3485, + "time_per_iteration": 2.6897850036621094 + }, + { + "auxiliary_loss_clip": 0.01193453, + "auxiliary_loss_mlp": 0.01036516, + "balance_loss_clip": 1.05720782, + "balance_loss_mlp": 1.02780771, + "epoch": 0.4191667167678711, + "flos": 20741034234240.0, + "grad_norm": 1.8921819505926456, + "language_loss": 0.84149122, + "learning_rate": 2.6109175628968853e-06, + "loss": 0.86379093, + "num_input_tokens_seen": 75138970, + "step": 3486, + "time_per_iteration": 2.6033337116241455 + }, + { + "auxiliary_loss_clip": 0.01184305, + "auxiliary_loss_mlp": 0.01026368, + "balance_loss_clip": 1.05087805, + "balance_loss_mlp": 1.01873291, + "epoch": 0.4192869596585102, + "flos": 23586416052480.0, + "grad_norm": 2.151233451107798, + "language_loss": 0.8290481, + "learning_rate": 2.610175775646878e-06, + "loss": 0.8511548, + "num_input_tokens_seen": 75157550, + "step": 3487, + "time_per_iteration": 2.6254007816314697 + }, + { + "auxiliary_loss_clip": 0.01200178, + "auxiliary_loss_mlp": 0.01025594, + "balance_loss_clip": 1.05024219, + "balance_loss_mlp": 1.01755309, + "epoch": 0.41940720254914926, + "flos": 25081269384960.0, + "grad_norm": 3.209412357297403, + "language_loss": 0.73307312, + "learning_rate": 2.6094338958336907e-06, + "loss": 0.7553308, + "num_input_tokens_seen": 75176220, + "step": 3488, + "time_per_iteration": 2.6838722229003906 + }, + { + "auxiliary_loss_clip": 0.01208614, + "auxiliary_loss_mlp": 0.01020928, + "balance_loss_clip": 1.05546474, + "balance_loss_mlp": 1.01379323, + "epoch": 0.41952744543978837, + "flos": 15554132628480.0, + "grad_norm": 3.5102985982160906, + "language_loss": 0.82220995, + "learning_rate": 2.608691923569867e-06, + "loss": 0.84450543, + "num_input_tokens_seen": 75193095, + "step": 3489, + "time_per_iteration": 2.7094247341156006 + }, + { + "auxiliary_loss_clip": 0.01204992, + "auxiliary_loss_mlp": 0.01033907, + "balance_loss_clip": 1.05734062, + "balance_loss_mlp": 1.02538347, + "epoch": 0.4196476883304275, + "flos": 24644775312000.0, + "grad_norm": 1.620844492924879, + "language_loss": 0.75575918, + "learning_rate": 2.6079498589679616e-06, + "loss": 0.77814817, + "num_input_tokens_seen": 75214185, + "step": 3490, + "time_per_iteration": 2.8356502056121826 + }, + { + "auxiliary_loss_clip": 0.01224266, + "auxiliary_loss_mlp": 0.0102742, + "balance_loss_clip": 1.04365218, + "balance_loss_mlp": 1.01821065, + "epoch": 0.41976793122106654, + "flos": 24531333183360.0, + "grad_norm": 5.335176757425504, + "language_loss": 0.76118797, + "learning_rate": 2.6072077021405465e-06, + "loss": 0.78370482, + "num_input_tokens_seen": 75233020, + "step": 3491, + "time_per_iteration": 2.7840073108673096 + }, + { + "auxiliary_loss_clip": 0.01218075, + "auxiliary_loss_mlp": 0.01033954, + "balance_loss_clip": 1.04824376, + "balance_loss_mlp": 1.02560318, + "epoch": 0.41988817411170565, + "flos": 21175301664000.0, + "grad_norm": 1.9067751303502605, + "language_loss": 0.6871711, + "learning_rate": 2.6064654532002054e-06, + "loss": 0.70969141, + "num_input_tokens_seen": 75252030, + "step": 3492, + "time_per_iteration": 2.745948314666748 + }, + { + "auxiliary_loss_clip": 0.01190753, + "auxiliary_loss_mlp": 0.01032191, + "balance_loss_clip": 1.05628562, + "balance_loss_mlp": 1.02380431, + "epoch": 0.42000841700234476, + "flos": 31649402626560.0, + "grad_norm": 1.6150785771393534, + "language_loss": 0.7596131, + "learning_rate": 2.6057231122595375e-06, + "loss": 0.78184253, + "num_input_tokens_seen": 75273340, + "step": 3493, + "time_per_iteration": 2.663006067276001 + }, + { + "auxiliary_loss_clip": 0.01205695, + "auxiliary_loss_mlp": 0.01023597, + "balance_loss_clip": 1.05103016, + "balance_loss_mlp": 1.01535916, + "epoch": 0.4201286598929838, + "flos": 21281525159040.0, + "grad_norm": 1.7460274334716492, + "language_loss": 0.73040175, + "learning_rate": 2.604980679431154e-06, + "loss": 0.75269473, + "num_input_tokens_seen": 75291580, + "step": 3494, + "time_per_iteration": 2.7108848094940186 + }, + { + "auxiliary_loss_clip": 0.0119816, + "auxiliary_loss_mlp": 0.01036155, + "balance_loss_clip": 1.05248952, + "balance_loss_mlp": 1.02848411, + "epoch": 0.4202489027836229, + "flos": 18546532813440.0, + "grad_norm": 1.964714462292399, + "language_loss": 0.74644655, + "learning_rate": 2.604238154827684e-06, + "loss": 0.76878977, + "num_input_tokens_seen": 75308205, + "step": 3495, + "time_per_iteration": 4.429099082946777 + }, + { + "auxiliary_loss_clip": 0.01199648, + "auxiliary_loss_mlp": 0.01026953, + "balance_loss_clip": 1.05418444, + "balance_loss_mlp": 1.01974416, + "epoch": 0.42036914567426203, + "flos": 19317643009920.0, + "grad_norm": 1.658497987176286, + "language_loss": 0.72636235, + "learning_rate": 2.6034955385617656e-06, + "loss": 0.74862832, + "num_input_tokens_seen": 75326535, + "step": 3496, + "time_per_iteration": 3.490813732147217 + }, + { + "auxiliary_loss_clip": 0.01124508, + "auxiliary_loss_mlp": 0.01007904, + "balance_loss_clip": 1.02233171, + "balance_loss_mlp": 1.00636578, + "epoch": 0.4204893885649011, + "flos": 67842942935040.0, + "grad_norm": 0.7194227311339065, + "language_loss": 0.61665058, + "learning_rate": 2.6027528307460544e-06, + "loss": 0.63797474, + "num_input_tokens_seen": 75390540, + "step": 3497, + "time_per_iteration": 3.352672815322876 + }, + { + "auxiliary_loss_clip": 0.01190052, + "auxiliary_loss_mlp": 0.01027946, + "balance_loss_clip": 1.05610538, + "balance_loss_mlp": 1.02002454, + "epoch": 0.4206096314555402, + "flos": 21908777385600.0, + "grad_norm": 1.8305220898255, + "language_loss": 0.86701876, + "learning_rate": 2.602010031493217e-06, + "loss": 0.88919878, + "num_input_tokens_seen": 75408770, + "step": 3498, + "time_per_iteration": 2.6639575958251953 + }, + { + "auxiliary_loss_clip": 0.01207162, + "auxiliary_loss_mlp": 0.0102275, + "balance_loss_clip": 1.04749489, + "balance_loss_mlp": 1.01466751, + "epoch": 0.42072987434617926, + "flos": 29278185269760.0, + "grad_norm": 4.368826897972221, + "language_loss": 0.86865509, + "learning_rate": 2.6012671409159367e-06, + "loss": 0.8909542, + "num_input_tokens_seen": 75430105, + "step": 3499, + "time_per_iteration": 2.7664647102355957 + }, + { + "auxiliary_loss_clip": 0.01205232, + "auxiliary_loss_mlp": 0.01037716, + "balance_loss_clip": 1.05518222, + "balance_loss_mlp": 1.02893007, + "epoch": 0.42085011723681837, + "flos": 27600726170880.0, + "grad_norm": 2.627210180128501, + "language_loss": 0.81447041, + "learning_rate": 2.6005241591269097e-06, + "loss": 0.83689988, + "num_input_tokens_seen": 75449475, + "step": 3500, + "time_per_iteration": 2.7396342754364014 + }, + { + "auxiliary_loss_clip": 0.01213755, + "auxiliary_loss_mlp": 0.01023639, + "balance_loss_clip": 1.05498767, + "balance_loss_mlp": 1.01595616, + "epoch": 0.4209703601274575, + "flos": 27818632028160.0, + "grad_norm": 2.02678279924409, + "language_loss": 0.79510498, + "learning_rate": 2.5997810862388454e-06, + "loss": 0.8174789, + "num_input_tokens_seen": 75469315, + "step": 3501, + "time_per_iteration": 2.738645076751709 + }, + { + "auxiliary_loss_clip": 0.01205963, + "auxiliary_loss_mlp": 0.01029269, + "balance_loss_clip": 1.051121, + "balance_loss_mlp": 1.02082634, + "epoch": 0.42109060301809653, + "flos": 27525529048320.0, + "grad_norm": 2.1841386533929406, + "language_loss": 0.75523722, + "learning_rate": 2.599037922364467e-06, + "loss": 0.77758956, + "num_input_tokens_seen": 75488215, + "step": 3502, + "time_per_iteration": 3.744307041168213 + }, + { + "auxiliary_loss_clip": 0.01209602, + "auxiliary_loss_mlp": 0.01030395, + "balance_loss_clip": 1.05248713, + "balance_loss_mlp": 1.02208591, + "epoch": 0.42121084590873564, + "flos": 29314275459840.0, + "grad_norm": 2.1488024106432073, + "language_loss": 0.75348884, + "learning_rate": 2.5982946676165112e-06, + "loss": 0.77588886, + "num_input_tokens_seen": 75507985, + "step": 3503, + "time_per_iteration": 2.7900805473327637 + }, + { + "auxiliary_loss_clip": 0.01131276, + "auxiliary_loss_mlp": 0.0100088, + "balance_loss_clip": 1.0348742, + "balance_loss_mlp": 0.99977171, + "epoch": 0.42133108879937475, + "flos": 67398835178880.0, + "grad_norm": 0.7266416636390547, + "language_loss": 0.57687193, + "learning_rate": 2.5975513221077313e-06, + "loss": 0.59819353, + "num_input_tokens_seen": 75571955, + "step": 3504, + "time_per_iteration": 3.3578054904937744 + }, + { + "auxiliary_loss_clip": 0.01195815, + "auxiliary_loss_mlp": 0.01029814, + "balance_loss_clip": 1.05061758, + "balance_loss_mlp": 1.02182698, + "epoch": 0.4214513316900138, + "flos": 23106038538240.0, + "grad_norm": 2.6182016287867422, + "language_loss": 0.88377786, + "learning_rate": 2.5968078859508897e-06, + "loss": 0.90603411, + "num_input_tokens_seen": 75589155, + "step": 3505, + "time_per_iteration": 2.719320297241211 + }, + { + "auxiliary_loss_clip": 0.01199399, + "auxiliary_loss_mlp": 0.01031902, + "balance_loss_clip": 1.05500984, + "balance_loss_mlp": 1.02377772, + "epoch": 0.4215715745806529, + "flos": 15336190857600.0, + "grad_norm": 1.9505983765680337, + "language_loss": 0.79645962, + "learning_rate": 2.5960643592587673e-06, + "loss": 0.81877261, + "num_input_tokens_seen": 75606565, + "step": 3506, + "time_per_iteration": 2.632835626602173 + }, + { + "auxiliary_loss_clip": 0.01215928, + "auxiliary_loss_mlp": 0.01027869, + "balance_loss_clip": 1.05126381, + "balance_loss_mlp": 1.01964927, + "epoch": 0.42169181747129203, + "flos": 22127257860480.0, + "grad_norm": 1.819972382188843, + "language_loss": 0.80806363, + "learning_rate": 2.5953207421441553e-06, + "loss": 0.83050162, + "num_input_tokens_seen": 75625165, + "step": 3507, + "time_per_iteration": 2.718494176864624 + }, + { + "auxiliary_loss_clip": 0.01217218, + "auxiliary_loss_mlp": 0.01027033, + "balance_loss_clip": 1.05223978, + "balance_loss_mlp": 1.01936793, + "epoch": 0.4218120603619311, + "flos": 22630724841600.0, + "grad_norm": 2.1630611528109687, + "language_loss": 0.74967003, + "learning_rate": 2.5945770347198603e-06, + "loss": 0.77211249, + "num_input_tokens_seen": 75643320, + "step": 3508, + "time_per_iteration": 2.7466206550598145 + }, + { + "auxiliary_loss_clip": 0.01202296, + "auxiliary_loss_mlp": 0.01028326, + "balance_loss_clip": 1.05157089, + "balance_loss_mlp": 1.02054191, + "epoch": 0.4219323032525702, + "flos": 19682818629120.0, + "grad_norm": 1.7939661572847434, + "language_loss": 0.82244062, + "learning_rate": 2.593833237098701e-06, + "loss": 0.84474683, + "num_input_tokens_seen": 75660920, + "step": 3509, + "time_per_iteration": 2.6998708248138428 + }, + { + "auxiliary_loss_clip": 0.01194257, + "auxiliary_loss_mlp": 0.0103546, + "balance_loss_clip": 1.05055988, + "balance_loss_mlp": 1.02664495, + "epoch": 0.4220525461432093, + "flos": 30190747224960.0, + "grad_norm": 1.978322178432609, + "language_loss": 0.62711024, + "learning_rate": 2.593089349393512e-06, + "loss": 0.64940739, + "num_input_tokens_seen": 75681410, + "step": 3510, + "time_per_iteration": 2.721264600753784 + }, + { + "auxiliary_loss_clip": 0.01199305, + "auxiliary_loss_mlp": 0.01029857, + "balance_loss_clip": 1.05696404, + "balance_loss_mlp": 1.0213393, + "epoch": 0.42217278903384836, + "flos": 24315941278080.0, + "grad_norm": 1.9636588710034126, + "language_loss": 0.83273005, + "learning_rate": 2.592345371717141e-06, + "loss": 0.85502172, + "num_input_tokens_seen": 75700940, + "step": 3511, + "time_per_iteration": 2.7092642784118652 + }, + { + "auxiliary_loss_clip": 0.01204725, + "auxiliary_loss_mlp": 0.01032177, + "balance_loss_clip": 1.0625881, + "balance_loss_mlp": 1.02402878, + "epoch": 0.42229303192448747, + "flos": 17092474352640.0, + "grad_norm": 1.9055397784564367, + "language_loss": 0.71608269, + "learning_rate": 2.591601304182448e-06, + "loss": 0.73845172, + "num_input_tokens_seen": 75718910, + "step": 3512, + "time_per_iteration": 2.6489064693450928 + }, + { + "auxiliary_loss_clip": 0.01211996, + "auxiliary_loss_mlp": 0.01027803, + "balance_loss_clip": 1.05827236, + "balance_loss_mlp": 1.02033782, + "epoch": 0.4224132748151266, + "flos": 22784530878720.0, + "grad_norm": 1.64132866129588, + "language_loss": 0.79500747, + "learning_rate": 2.5908571469023067e-06, + "loss": 0.81740546, + "num_input_tokens_seen": 75738395, + "step": 3513, + "time_per_iteration": 2.695725440979004 + }, + { + "auxiliary_loss_clip": 0.01190424, + "auxiliary_loss_mlp": 0.01025748, + "balance_loss_clip": 1.05590963, + "balance_loss_mlp": 1.01771355, + "epoch": 0.42253351770576564, + "flos": 17819090576640.0, + "grad_norm": 2.194852130376686, + "language_loss": 0.75722146, + "learning_rate": 2.5901128999896067e-06, + "loss": 0.77938318, + "num_input_tokens_seen": 75753825, + "step": 3514, + "time_per_iteration": 2.6311874389648438 + }, + { + "auxiliary_loss_clip": 0.01194407, + "auxiliary_loss_mlp": 0.01030603, + "balance_loss_clip": 1.05374575, + "balance_loss_mlp": 1.02228212, + "epoch": 0.42265376059640475, + "flos": 28512390286080.0, + "grad_norm": 1.590033887400631, + "language_loss": 0.68279278, + "learning_rate": 2.5893685635572487e-06, + "loss": 0.7050429, + "num_input_tokens_seen": 75774675, + "step": 3515, + "time_per_iteration": 2.6597557067871094 + }, + { + "auxiliary_loss_clip": 0.01201453, + "auxiliary_loss_mlp": 0.0102525, + "balance_loss_clip": 1.05094516, + "balance_loss_mlp": 1.01713777, + "epoch": 0.4227740034870438, + "flos": 16253349753600.0, + "grad_norm": 1.9672393249646956, + "language_loss": 0.69402564, + "learning_rate": 2.5886241377181483e-06, + "loss": 0.71629262, + "num_input_tokens_seen": 75793545, + "step": 3516, + "time_per_iteration": 2.7290892601013184 + }, + { + "auxiliary_loss_clip": 0.01198527, + "auxiliary_loss_mlp": 0.01033729, + "balance_loss_clip": 1.05403495, + "balance_loss_mlp": 1.02461553, + "epoch": 0.4228942463776829, + "flos": 25295691623040.0, + "grad_norm": 1.8418860786204259, + "language_loss": 0.81307054, + "learning_rate": 2.587879622585234e-06, + "loss": 0.83539307, + "num_input_tokens_seen": 75812145, + "step": 3517, + "time_per_iteration": 2.753483295440674 + }, + { + "auxiliary_loss_clip": 0.01199076, + "auxiliary_loss_mlp": 0.01027321, + "balance_loss_clip": 1.05637574, + "balance_loss_mlp": 1.01928639, + "epoch": 0.423014489268322, + "flos": 26395779507840.0, + "grad_norm": 2.0556673737767994, + "language_loss": 0.76287746, + "learning_rate": 2.5871350182714486e-06, + "loss": 0.78514147, + "num_input_tokens_seen": 75833025, + "step": 3518, + "time_per_iteration": 2.677734136581421 + }, + { + "auxiliary_loss_clip": 0.01189238, + "auxiliary_loss_mlp": 0.01028253, + "balance_loss_clip": 1.05624044, + "balance_loss_mlp": 1.02027178, + "epoch": 0.4231347321589611, + "flos": 17274002711040.0, + "grad_norm": 2.8014016411875016, + "language_loss": 0.80527329, + "learning_rate": 2.586390324889748e-06, + "loss": 0.82744819, + "num_input_tokens_seen": 75848925, + "step": 3519, + "time_per_iteration": 2.560899257659912 + }, + { + "auxiliary_loss_clip": 0.01194642, + "auxiliary_loss_mlp": 0.01033599, + "balance_loss_clip": 1.05424333, + "balance_loss_mlp": 1.02476013, + "epoch": 0.4232549750496002, + "flos": 22999635475200.0, + "grad_norm": 5.037568643870271, + "language_loss": 0.6758939, + "learning_rate": 2.5856455425531003e-06, + "loss": 0.69817626, + "num_input_tokens_seen": 75870400, + "step": 3520, + "time_per_iteration": 2.6943652629852295 + }, + { + "auxiliary_loss_clip": 0.01197327, + "auxiliary_loss_mlp": 0.0102591, + "balance_loss_clip": 1.05593455, + "balance_loss_mlp": 1.01795244, + "epoch": 0.4233752179402393, + "flos": 21248343970560.0, + "grad_norm": 1.820253008895416, + "language_loss": 0.80697674, + "learning_rate": 2.5849006713744902e-06, + "loss": 0.82920909, + "num_input_tokens_seen": 75889195, + "step": 3521, + "time_per_iteration": 4.516894102096558 + }, + { + "auxiliary_loss_clip": 0.01205477, + "auxiliary_loss_mlp": 0.010306, + "balance_loss_clip": 1.05483532, + "balance_loss_mlp": 1.02276206, + "epoch": 0.42349546083087836, + "flos": 20704297599360.0, + "grad_norm": 2.0896612238235166, + "language_loss": 0.73088843, + "learning_rate": 2.5841557114669135e-06, + "loss": 0.75324917, + "num_input_tokens_seen": 75906055, + "step": 3522, + "time_per_iteration": 2.655472755432129 + }, + { + "auxiliary_loss_clip": 0.0119446, + "auxiliary_loss_mlp": 0.01028682, + "balance_loss_clip": 1.05702269, + "balance_loss_mlp": 1.01899672, + "epoch": 0.42361570372151747, + "flos": 18585065128320.0, + "grad_norm": 2.8022403065809676, + "language_loss": 0.67132562, + "learning_rate": 2.58341066294338e-06, + "loss": 0.69355702, + "num_input_tokens_seen": 75922720, + "step": 3523, + "time_per_iteration": 2.5752596855163574 + }, + { + "auxiliary_loss_clip": 0.01228422, + "auxiliary_loss_mlp": 0.01379123, + "balance_loss_clip": 1.05042732, + "balance_loss_mlp": 1.00019598, + "epoch": 0.4237359466121566, + "flos": 20959478795520.0, + "grad_norm": 2.1882763403412597, + "language_loss": 0.85340512, + "learning_rate": 2.5826655259169124e-06, + "loss": 0.8794806, + "num_input_tokens_seen": 75941375, + "step": 3524, + "time_per_iteration": 2.7317585945129395 + }, + { + "auxiliary_loss_clip": 0.01192389, + "auxiliary_loss_mlp": 0.01028807, + "balance_loss_clip": 1.05854416, + "balance_loss_mlp": 1.02104056, + "epoch": 0.42385618950279563, + "flos": 18038181582720.0, + "grad_norm": 2.0182312978467665, + "language_loss": 0.91101789, + "learning_rate": 2.5819203005005475e-06, + "loss": 0.93322986, + "num_input_tokens_seen": 75958710, + "step": 3525, + "time_per_iteration": 2.6305556297302246 + }, + { + "auxiliary_loss_clip": 0.01200767, + "auxiliary_loss_mlp": 0.01026455, + "balance_loss_clip": 1.05336261, + "balance_loss_mlp": 1.01819968, + "epoch": 0.42397643239343474, + "flos": 23769129559680.0, + "grad_norm": 1.6680033907060032, + "language_loss": 0.78477222, + "learning_rate": 2.581174986807336e-06, + "loss": 0.80704445, + "num_input_tokens_seen": 75978945, + "step": 3526, + "time_per_iteration": 2.6784186363220215 + }, + { + "auxiliary_loss_clip": 0.01187797, + "auxiliary_loss_mlp": 0.01379134, + "balance_loss_clip": 1.05337417, + "balance_loss_mlp": 1.00022793, + "epoch": 0.42409667528407385, + "flos": 16545088016640.0, + "grad_norm": 2.2555973537235547, + "language_loss": 0.9115569, + "learning_rate": 2.580429584950341e-06, + "loss": 0.93722624, + "num_input_tokens_seen": 75994695, + "step": 3527, + "time_per_iteration": 2.689711093902588 + }, + { + "auxiliary_loss_clip": 0.0122061, + "auxiliary_loss_mlp": 0.01028221, + "balance_loss_clip": 1.05178332, + "balance_loss_mlp": 1.01960206, + "epoch": 0.4242169181747129, + "flos": 16034186920320.0, + "grad_norm": 2.1259019156639423, + "language_loss": 0.66385806, + "learning_rate": 2.5796840950426397e-06, + "loss": 0.68634641, + "num_input_tokens_seen": 76011780, + "step": 3528, + "time_per_iteration": 3.6170473098754883 + }, + { + "auxiliary_loss_clip": 0.01188723, + "auxiliary_loss_mlp": 0.01034136, + "balance_loss_clip": 1.05385661, + "balance_loss_mlp": 1.02576196, + "epoch": 0.424337161065352, + "flos": 20084012611200.0, + "grad_norm": 1.8729347373273708, + "language_loss": 0.65850508, + "learning_rate": 2.578938517197322e-06, + "loss": 0.68073368, + "num_input_tokens_seen": 76029875, + "step": 3529, + "time_per_iteration": 2.6958961486816406 + }, + { + "auxiliary_loss_clip": 0.01197116, + "auxiliary_loss_mlp": 0.01029412, + "balance_loss_clip": 1.05321622, + "balance_loss_mlp": 1.02064371, + "epoch": 0.4244574039559911, + "flos": 23878369797120.0, + "grad_norm": 2.3382702443895447, + "language_loss": 0.62520301, + "learning_rate": 2.5781928515274916e-06, + "loss": 0.64746833, + "num_input_tokens_seen": 76048595, + "step": 3530, + "time_per_iteration": 2.72739577293396 + }, + { + "auxiliary_loss_clip": 0.01203819, + "auxiliary_loss_mlp": 0.01026386, + "balance_loss_clip": 1.05844927, + "balance_loss_mlp": 1.01893592, + "epoch": 0.4245776468466302, + "flos": 17565920542080.0, + "grad_norm": 2.15588233497027, + "language_loss": 0.67631912, + "learning_rate": 2.577447098146265e-06, + "loss": 0.69862115, + "num_input_tokens_seen": 76065770, + "step": 3531, + "time_per_iteration": 2.594045877456665 + }, + { + "auxiliary_loss_clip": 0.01215031, + "auxiliary_loss_mlp": 0.01025448, + "balance_loss_clip": 1.05138636, + "balance_loss_mlp": 1.01715136, + "epoch": 0.4246978897372693, + "flos": 27776256958080.0, + "grad_norm": 1.7560171540565854, + "language_loss": 0.78835768, + "learning_rate": 2.5767012571667724e-06, + "loss": 0.81076247, + "num_input_tokens_seen": 76085250, + "step": 3532, + "time_per_iteration": 2.8121721744537354 + }, + { + "auxiliary_loss_clip": 0.01199175, + "auxiliary_loss_mlp": 0.01040514, + "balance_loss_clip": 1.05377555, + "balance_loss_mlp": 1.03073931, + "epoch": 0.42481813262790835, + "flos": 15596615439360.0, + "grad_norm": 1.791820126952655, + "language_loss": 0.68822831, + "learning_rate": 2.5759553287021587e-06, + "loss": 0.71062517, + "num_input_tokens_seen": 76103580, + "step": 3533, + "time_per_iteration": 2.6634225845336914 + }, + { + "auxiliary_loss_clip": 0.01206806, + "auxiliary_loss_mlp": 0.01035015, + "balance_loss_clip": 1.05489373, + "balance_loss_mlp": 1.02634823, + "epoch": 0.42493837551854746, + "flos": 23951088881280.0, + "grad_norm": 2.6990736283871253, + "language_loss": 0.77707303, + "learning_rate": 2.5752093128655786e-06, + "loss": 0.79949123, + "num_input_tokens_seen": 76121825, + "step": 3534, + "time_per_iteration": 2.725147008895874 + }, + { + "auxiliary_loss_clip": 0.01200623, + "auxiliary_loss_mlp": 0.01033138, + "balance_loss_clip": 1.05304623, + "balance_loss_mlp": 1.0244596, + "epoch": 0.4250586184091866, + "flos": 20813466009600.0, + "grad_norm": 2.0911203249393773, + "language_loss": 0.73759532, + "learning_rate": 2.574463209770204e-06, + "loss": 0.75993299, + "num_input_tokens_seen": 76141140, + "step": 3535, + "time_per_iteration": 2.6707887649536133 + }, + { + "auxiliary_loss_clip": 0.01209175, + "auxiliary_loss_mlp": 0.0102655, + "balance_loss_clip": 1.04823661, + "balance_loss_mlp": 1.01805627, + "epoch": 0.42517886129982563, + "flos": 30371018607360.0, + "grad_norm": 1.6690288941845306, + "language_loss": 0.79261601, + "learning_rate": 2.5737170195292165e-06, + "loss": 0.81497324, + "num_input_tokens_seen": 76164475, + "step": 3536, + "time_per_iteration": 2.8049166202545166 + }, + { + "auxiliary_loss_clip": 0.01213001, + "auxiliary_loss_mlp": 0.01031116, + "balance_loss_clip": 1.04934585, + "balance_loss_mlp": 1.02249146, + "epoch": 0.42529910419046474, + "flos": 20080636732800.0, + "grad_norm": 2.4078215322081467, + "language_loss": 0.78222096, + "learning_rate": 2.572970742255814e-06, + "loss": 0.80466211, + "num_input_tokens_seen": 76182965, + "step": 3537, + "time_per_iteration": 2.733185052871704 + }, + { + "auxiliary_loss_clip": 0.01197331, + "auxiliary_loss_mlp": 0.01034139, + "balance_loss_clip": 1.05633402, + "balance_loss_mlp": 1.02668786, + "epoch": 0.42541934708110385, + "flos": 22632448694400.0, + "grad_norm": 1.680810940031312, + "language_loss": 0.81843042, + "learning_rate": 2.5722243780632046e-06, + "loss": 0.84074509, + "num_input_tokens_seen": 76201230, + "step": 3538, + "time_per_iteration": 2.6710755825042725 + }, + { + "auxiliary_loss_clip": 0.01140304, + "auxiliary_loss_mlp": 0.01012025, + "balance_loss_clip": 1.02573991, + "balance_loss_mlp": 1.01050508, + "epoch": 0.4255395899717429, + "flos": 66200676186240.0, + "grad_norm": 0.7553118697248312, + "language_loss": 0.60493124, + "learning_rate": 2.5714779270646125e-06, + "loss": 0.62645447, + "num_input_tokens_seen": 76262000, + "step": 3539, + "time_per_iteration": 3.280014753341675 + }, + { + "auxiliary_loss_clip": 0.0120778, + "auxiliary_loss_mlp": 0.01379417, + "balance_loss_clip": 1.05291736, + "balance_loss_mlp": 1.00010717, + "epoch": 0.425659832862382, + "flos": 17931814433280.0, + "grad_norm": 3.3576905689455225, + "language_loss": 0.77780831, + "learning_rate": 2.5707313893732735e-06, + "loss": 0.80368024, + "num_input_tokens_seen": 76280540, + "step": 3540, + "time_per_iteration": 2.6548871994018555 + }, + { + "auxiliary_loss_clip": 0.01228384, + "auxiliary_loss_mlp": 0.01028066, + "balance_loss_clip": 1.04113531, + "balance_loss_mlp": 1.01952457, + "epoch": 0.4257800757530211, + "flos": 24022550989440.0, + "grad_norm": 1.8092489518025043, + "language_loss": 0.77122009, + "learning_rate": 2.5699847651024364e-06, + "loss": 0.79378456, + "num_input_tokens_seen": 76301180, + "step": 3541, + "time_per_iteration": 2.9443392753601074 + }, + { + "auxiliary_loss_clip": 0.0119262, + "auxiliary_loss_mlp": 0.01030295, + "balance_loss_clip": 1.05378294, + "balance_loss_mlp": 1.02174163, + "epoch": 0.4259003186436602, + "flos": 23696015425920.0, + "grad_norm": 2.371474140832626, + "language_loss": 0.77078998, + "learning_rate": 2.5692380543653627e-06, + "loss": 0.79301918, + "num_input_tokens_seen": 76319335, + "step": 3542, + "time_per_iteration": 2.920558452606201 + }, + { + "auxiliary_loss_clip": 0.01200594, + "auxiliary_loss_mlp": 0.01379196, + "balance_loss_clip": 1.05585718, + "balance_loss_mlp": 1.00021851, + "epoch": 0.4260205615342993, + "flos": 15259772672640.0, + "grad_norm": 2.1915354552942152, + "language_loss": 0.69455743, + "learning_rate": 2.5684912572753293e-06, + "loss": 0.72035539, + "num_input_tokens_seen": 76335010, + "step": 3543, + "time_per_iteration": 2.62893009185791 + }, + { + "auxiliary_loss_clip": 0.01186179, + "auxiliary_loss_mlp": 0.01025406, + "balance_loss_clip": 1.05571628, + "balance_loss_mlp": 1.01820874, + "epoch": 0.4261408044249384, + "flos": 30665306736000.0, + "grad_norm": 1.8620278920612754, + "language_loss": 0.83671427, + "learning_rate": 2.5677443739456245e-06, + "loss": 0.85883003, + "num_input_tokens_seen": 76356670, + "step": 3544, + "time_per_iteration": 2.7705769538879395 + }, + { + "auxiliary_loss_clip": 0.01206631, + "auxiliary_loss_mlp": 0.01029542, + "balance_loss_clip": 1.05360544, + "balance_loss_mlp": 1.02125645, + "epoch": 0.42626104731557746, + "flos": 23257905240960.0, + "grad_norm": 2.2100183259667836, + "language_loss": 0.79297841, + "learning_rate": 2.5669974044895495e-06, + "loss": 0.81534016, + "num_input_tokens_seen": 76373065, + "step": 3545, + "time_per_iteration": 2.652148723602295 + }, + { + "auxiliary_loss_clip": 0.01222167, + "auxiliary_loss_mlp": 0.01031537, + "balance_loss_clip": 1.05146241, + "balance_loss_mlp": 1.02421737, + "epoch": 0.42638129020621657, + "flos": 25884770670720.0, + "grad_norm": 2.2854940884595725, + "language_loss": 0.79530501, + "learning_rate": 2.5662503490204187e-06, + "loss": 0.81784201, + "num_input_tokens_seen": 76393230, + "step": 3546, + "time_per_iteration": 2.788116455078125 + }, + { + "auxiliary_loss_clip": 0.01202489, + "auxiliary_loss_mlp": 0.01029566, + "balance_loss_clip": 1.05148077, + "balance_loss_mlp": 1.02161455, + "epoch": 0.4265015330968556, + "flos": 26502362138880.0, + "grad_norm": 1.8023456917829443, + "language_loss": 0.76146632, + "learning_rate": 2.5655032076515603e-06, + "loss": 0.78378689, + "num_input_tokens_seen": 76412555, + "step": 3547, + "time_per_iteration": 3.6706430912017822 + }, + { + "auxiliary_loss_clip": 0.01207832, + "auxiliary_loss_mlp": 0.01028525, + "balance_loss_clip": 1.05459595, + "balance_loss_mlp": 1.02082431, + "epoch": 0.42662177598749473, + "flos": 24389522288640.0, + "grad_norm": 1.968526883142543, + "language_loss": 0.8170855, + "learning_rate": 2.5647559804963155e-06, + "loss": 0.83944905, + "num_input_tokens_seen": 76432485, + "step": 3548, + "time_per_iteration": 2.693053960800171 + }, + { + "auxiliary_loss_clip": 0.01227909, + "auxiliary_loss_mlp": 0.01028562, + "balance_loss_clip": 1.05037999, + "balance_loss_mlp": 1.02074814, + "epoch": 0.42674201887813384, + "flos": 23148629089920.0, + "grad_norm": 1.8802572876915964, + "language_loss": 0.7856338, + "learning_rate": 2.5640086676680364e-06, + "loss": 0.80819857, + "num_input_tokens_seen": 76453980, + "step": 3549, + "time_per_iteration": 2.7725980281829834 + }, + { + "auxiliary_loss_clip": 0.0119792, + "auxiliary_loss_mlp": 0.01026267, + "balance_loss_clip": 1.05569053, + "balance_loss_mlp": 1.01779127, + "epoch": 0.4268622617687729, + "flos": 21689614552320.0, + "grad_norm": 2.2469020557803927, + "language_loss": 0.80611658, + "learning_rate": 2.5632612692800923e-06, + "loss": 0.82835841, + "num_input_tokens_seen": 76473045, + "step": 3550, + "time_per_iteration": 2.6866934299468994 + }, + { + "auxiliary_loss_clip": 0.01216398, + "auxiliary_loss_mlp": 0.01031667, + "balance_loss_clip": 1.05254376, + "balance_loss_mlp": 1.02260709, + "epoch": 0.426982504659412, + "flos": 23440151871360.0, + "grad_norm": 2.6558762739750743, + "language_loss": 0.75190413, + "learning_rate": 2.5625137854458603e-06, + "loss": 0.77438474, + "num_input_tokens_seen": 76492060, + "step": 3551, + "time_per_iteration": 2.7072315216064453 + }, + { + "auxiliary_loss_clip": 0.01206605, + "auxiliary_loss_mlp": 0.01029033, + "balance_loss_clip": 1.05130982, + "balance_loss_mlp": 1.0211947, + "epoch": 0.4271027475500511, + "flos": 18916556768640.0, + "grad_norm": 1.9266959442310996, + "language_loss": 0.80123103, + "learning_rate": 2.561766216278735e-06, + "loss": 0.82358736, + "num_input_tokens_seen": 76509655, + "step": 3552, + "time_per_iteration": 2.7418956756591797 + }, + { + "auxiliary_loss_clip": 0.01222276, + "auxiliary_loss_mlp": 0.01026608, + "balance_loss_clip": 1.05042696, + "balance_loss_mlp": 1.01800084, + "epoch": 0.4272229904406902, + "flos": 26870554500480.0, + "grad_norm": 1.7619986072459994, + "language_loss": 0.81725371, + "learning_rate": 2.561018561892121e-06, + "loss": 0.8397426, + "num_input_tokens_seen": 76528795, + "step": 3553, + "time_per_iteration": 2.8338499069213867 + }, + { + "auxiliary_loss_clip": 0.01202825, + "auxiliary_loss_mlp": 0.01030221, + "balance_loss_clip": 1.0509876, + "balance_loss_mlp": 1.02230525, + "epoch": 0.4273432333313293, + "flos": 23951376190080.0, + "grad_norm": 1.5008890249347062, + "language_loss": 0.76805723, + "learning_rate": 2.5602708223994363e-06, + "loss": 0.79038763, + "num_input_tokens_seen": 76550660, + "step": 3554, + "time_per_iteration": 3.5719454288482666 + }, + { + "auxiliary_loss_clip": 0.01213832, + "auxiliary_loss_mlp": 0.01031146, + "balance_loss_clip": 1.0480845, + "balance_loss_mlp": 1.02299809, + "epoch": 0.4274634762219684, + "flos": 29570354496000.0, + "grad_norm": 10.212342282466006, + "language_loss": 0.67886084, + "learning_rate": 2.559522997914115e-06, + "loss": 0.70131063, + "num_input_tokens_seen": 76570240, + "step": 3555, + "time_per_iteration": 2.885772705078125 + }, + { + "auxiliary_loss_clip": 0.0118642, + "auxiliary_loss_mlp": 0.01028708, + "balance_loss_clip": 1.0566864, + "balance_loss_mlp": 1.02124548, + "epoch": 0.42758371911260745, + "flos": 21434146047360.0, + "grad_norm": 2.025271084694579, + "language_loss": 0.84448314, + "learning_rate": 2.558775088549599e-06, + "loss": 0.86663449, + "num_input_tokens_seen": 76589820, + "step": 3556, + "time_per_iteration": 2.661388874053955 + }, + { + "auxiliary_loss_clip": 0.012043, + "auxiliary_loss_mlp": 0.01036434, + "balance_loss_clip": 1.05583692, + "balance_loss_mlp": 1.02818203, + "epoch": 0.42770396200324656, + "flos": 14752822072320.0, + "grad_norm": 2.8693976197525695, + "language_loss": 0.66006631, + "learning_rate": 2.5580270944193467e-06, + "loss": 0.68247366, + "num_input_tokens_seen": 76606640, + "step": 3557, + "time_per_iteration": 2.622823476791382 + }, + { + "auxiliary_loss_clip": 0.01086633, + "auxiliary_loss_mlp": 0.01005101, + "balance_loss_clip": 1.02311397, + "balance_loss_mlp": 1.0036881, + "epoch": 0.4278242048938857, + "flos": 70654712601600.0, + "grad_norm": 0.747517281332642, + "language_loss": 0.5548932, + "learning_rate": 2.557279015636827e-06, + "loss": 0.57581061, + "num_input_tokens_seen": 76667050, + "step": 3558, + "time_per_iteration": 3.216348886489868 + }, + { + "auxiliary_loss_clip": 0.0109502, + "auxiliary_loss_mlp": 0.01004719, + "balance_loss_clip": 1.02202153, + "balance_loss_mlp": 1.00330091, + "epoch": 0.42794444778452473, + "flos": 69366165033600.0, + "grad_norm": 0.7634530799201418, + "language_loss": 0.61221117, + "learning_rate": 2.5565308523155245e-06, + "loss": 0.63320857, + "num_input_tokens_seen": 76726650, + "step": 3559, + "time_per_iteration": 3.1383237838745117 + }, + { + "auxiliary_loss_clip": 0.01215032, + "auxiliary_loss_mlp": 0.01029383, + "balance_loss_clip": 1.05003059, + "balance_loss_mlp": 1.02181876, + "epoch": 0.42806469067516384, + "flos": 18215328481920.0, + "grad_norm": 2.3164516367929506, + "language_loss": 0.82225817, + "learning_rate": 2.5557826045689336e-06, + "loss": 0.84470236, + "num_input_tokens_seen": 76742890, + "step": 3560, + "time_per_iteration": 2.74298095703125 + }, + { + "auxiliary_loss_clip": 0.0112095, + "auxiliary_loss_mlp": 0.01000845, + "balance_loss_clip": 1.02465606, + "balance_loss_mlp": 0.99950409, + "epoch": 0.4281849335658029, + "flos": 54535814432640.0, + "grad_norm": 0.8270234251970139, + "language_loss": 0.58800709, + "learning_rate": 2.5550342725105643e-06, + "loss": 0.60922503, + "num_input_tokens_seen": 76801055, + "step": 3561, + "time_per_iteration": 3.235391139984131 + }, + { + "auxiliary_loss_clip": 0.01198515, + "auxiliary_loss_mlp": 0.0102742, + "balance_loss_clip": 1.057163, + "balance_loss_mlp": 1.01948047, + "epoch": 0.428305176456442, + "flos": 17274828723840.0, + "grad_norm": 1.6688933301620912, + "language_loss": 0.8066892, + "learning_rate": 2.554285856253937e-06, + "loss": 0.8289485, + "num_input_tokens_seen": 76819890, + "step": 3562, + "time_per_iteration": 2.657564640045166 + }, + { + "auxiliary_loss_clip": 0.01203411, + "auxiliary_loss_mlp": 0.01026238, + "balance_loss_clip": 1.05378604, + "balance_loss_mlp": 1.01796496, + "epoch": 0.4284254193470811, + "flos": 26359509749760.0, + "grad_norm": 1.928961297291992, + "language_loss": 0.7757535, + "learning_rate": 2.5535373559125855e-06, + "loss": 0.79804999, + "num_input_tokens_seen": 76840255, + "step": 3563, + "time_per_iteration": 2.7946360111236572 + }, + { + "auxiliary_loss_clip": 0.01241357, + "auxiliary_loss_mlp": 0.01031731, + "balance_loss_clip": 1.0469228, + "balance_loss_mlp": 1.0234524, + "epoch": 0.42854566223772017, + "flos": 29714248379520.0, + "grad_norm": 1.5251500218146867, + "language_loss": 0.81934583, + "learning_rate": 2.552788771600057e-06, + "loss": 0.84207672, + "num_input_tokens_seen": 76860565, + "step": 3564, + "time_per_iteration": 2.8705673217773438 + }, + { + "auxiliary_loss_clip": 0.01218691, + "auxiliary_loss_mlp": 0.01027401, + "balance_loss_clip": 1.05397737, + "balance_loss_mlp": 1.01947927, + "epoch": 0.4286659051283593, + "flos": 22018161277440.0, + "grad_norm": 3.34519401884347, + "language_loss": 0.82094926, + "learning_rate": 2.5520401034299118e-06, + "loss": 0.84341013, + "num_input_tokens_seen": 76878325, + "step": 3565, + "time_per_iteration": 2.7449278831481934 + }, + { + "auxiliary_loss_clip": 0.01200452, + "auxiliary_loss_mlp": 0.01029296, + "balance_loss_clip": 1.05455685, + "balance_loss_mlp": 1.02014065, + "epoch": 0.4287861480189984, + "flos": 13334422838400.0, + "grad_norm": 2.215165396950984, + "language_loss": 0.87848055, + "learning_rate": 2.551291351515722e-06, + "loss": 0.90077806, + "num_input_tokens_seen": 76895340, + "step": 3566, + "time_per_iteration": 2.7028934955596924 + }, + { + "auxiliary_loss_clip": 0.01208021, + "auxiliary_loss_mlp": 0.01378983, + "balance_loss_clip": 1.04711008, + "balance_loss_mlp": 1.0000608, + "epoch": 0.42890639090963745, + "flos": 26651535321600.0, + "grad_norm": 1.6609002676746043, + "language_loss": 0.85689139, + "learning_rate": 2.5505425159710726e-06, + "loss": 0.88276136, + "num_input_tokens_seen": 76915150, + "step": 3567, + "time_per_iteration": 2.8373303413391113 + }, + { + "auxiliary_loss_clip": 0.01213467, + "auxiliary_loss_mlp": 0.01379096, + "balance_loss_clip": 1.0526948, + "balance_loss_mlp": 1.00021577, + "epoch": 0.42902663380027656, + "flos": 24055768091520.0, + "grad_norm": 1.8159584637437316, + "language_loss": 0.82825434, + "learning_rate": 2.549793596909561e-06, + "loss": 0.85417992, + "num_input_tokens_seen": 76933770, + "step": 3568, + "time_per_iteration": 2.7631707191467285 + }, + { + "auxiliary_loss_clip": 0.01202057, + "auxiliary_loss_mlp": 0.01024404, + "balance_loss_clip": 1.05232108, + "balance_loss_mlp": 1.01613712, + "epoch": 0.42914687669091567, + "flos": 15632561975040.0, + "grad_norm": 2.5487784947116654, + "language_loss": 0.66327596, + "learning_rate": 2.5490445944447976e-06, + "loss": 0.68554056, + "num_input_tokens_seen": 76952265, + "step": 3569, + "time_per_iteration": 2.6921989917755127 + }, + { + "auxiliary_loss_clip": 0.01198466, + "auxiliary_loss_mlp": 0.01029096, + "balance_loss_clip": 1.05505943, + "balance_loss_mlp": 1.02112722, + "epoch": 0.4292671195815547, + "flos": 31467802440960.0, + "grad_norm": 2.1290460964381235, + "language_loss": 0.6520226, + "learning_rate": 2.548295508690406e-06, + "loss": 0.67429829, + "num_input_tokens_seen": 76973560, + "step": 3570, + "time_per_iteration": 2.7360920906066895 + }, + { + "auxiliary_loss_clip": 0.01197842, + "auxiliary_loss_mlp": 0.01034842, + "balance_loss_clip": 1.05237103, + "balance_loss_mlp": 1.02745974, + "epoch": 0.42938736247219383, + "flos": 30257756046720.0, + "grad_norm": 1.8567562091783478, + "language_loss": 0.7646277, + "learning_rate": 2.5475463397600217e-06, + "loss": 0.78695452, + "num_input_tokens_seen": 76993640, + "step": 3571, + "time_per_iteration": 2.717895746231079 + }, + { + "auxiliary_loss_clip": 0.01195441, + "auxiliary_loss_mlp": 0.01029942, + "balance_loss_clip": 1.05949736, + "balance_loss_mlp": 1.02165651, + "epoch": 0.42950760536283294, + "flos": 29349683291520.0, + "grad_norm": 2.8673432951653264, + "language_loss": 0.7700569, + "learning_rate": 2.546797087767293e-06, + "loss": 0.79231071, + "num_input_tokens_seen": 77013765, + "step": 3572, + "time_per_iteration": 2.649695634841919 + }, + { + "auxiliary_loss_clip": 0.01220475, + "auxiliary_loss_mlp": 0.01024958, + "balance_loss_clip": 1.05050611, + "balance_loss_mlp": 1.01688194, + "epoch": 0.429627848253472, + "flos": 26869943969280.0, + "grad_norm": 1.6687502925208417, + "language_loss": 0.87358552, + "learning_rate": 2.546047752825881e-06, + "loss": 0.89603984, + "num_input_tokens_seen": 77034370, + "step": 3573, + "time_per_iteration": 3.7552976608276367 + }, + { + "auxiliary_loss_clip": 0.01226771, + "auxiliary_loss_mlp": 0.01029074, + "balance_loss_clip": 1.0478543, + "balance_loss_mlp": 1.02132583, + "epoch": 0.4297480911441111, + "flos": 13881270470400.0, + "grad_norm": 2.041749825840658, + "language_loss": 0.93492067, + "learning_rate": 2.5452983350494595e-06, + "loss": 0.95747912, + "num_input_tokens_seen": 77049925, + "step": 3574, + "time_per_iteration": 3.612565755844116 + }, + { + "auxiliary_loss_clip": 0.01200448, + "auxiliary_loss_mlp": 0.01379206, + "balance_loss_clip": 1.0563606, + "balance_loss_mlp": 1.00022292, + "epoch": 0.4298683340347502, + "flos": 20741141975040.0, + "grad_norm": 2.054890396399492, + "language_loss": 0.64966851, + "learning_rate": 2.544548834551713e-06, + "loss": 0.67546505, + "num_input_tokens_seen": 77068930, + "step": 3575, + "time_per_iteration": 2.6746695041656494 + }, + { + "auxiliary_loss_clip": 0.01206044, + "auxiliary_loss_mlp": 0.01378841, + "balance_loss_clip": 1.04788363, + "balance_loss_mlp": 1.00003886, + "epoch": 0.4299885769253893, + "flos": 20882126856960.0, + "grad_norm": 3.264957744346618, + "language_loss": 0.94486934, + "learning_rate": 2.5437992514463424e-06, + "loss": 0.9707182, + "num_input_tokens_seen": 77082255, + "step": 3576, + "time_per_iteration": 2.713831663131714 + }, + { + "auxiliary_loss_clip": 0.01198414, + "auxiliary_loss_mlp": 0.01029158, + "balance_loss_clip": 1.05582714, + "balance_loss_mlp": 1.02105784, + "epoch": 0.4301088198160284, + "flos": 25484618183040.0, + "grad_norm": 1.9205962412319855, + "language_loss": 0.87968677, + "learning_rate": 2.5430495858470565e-06, + "loss": 0.90196252, + "num_input_tokens_seen": 77101725, + "step": 3577, + "time_per_iteration": 2.7741806507110596 + }, + { + "auxiliary_loss_clip": 0.01195342, + "auxiliary_loss_mlp": 0.01030502, + "balance_loss_clip": 1.05549216, + "balance_loss_mlp": 1.02200818, + "epoch": 0.43022906270666744, + "flos": 18259427404800.0, + "grad_norm": 2.177115358470198, + "language_loss": 0.77275258, + "learning_rate": 2.54229983786758e-06, + "loss": 0.79501104, + "num_input_tokens_seen": 77119670, + "step": 3578, + "time_per_iteration": 2.65705943107605 + }, + { + "auxiliary_loss_clip": 0.01205296, + "auxiliary_loss_mlp": 0.01030633, + "balance_loss_clip": 1.05112123, + "balance_loss_mlp": 1.02252698, + "epoch": 0.43034930559730655, + "flos": 23399536567680.0, + "grad_norm": 2.0934935636039245, + "language_loss": 0.85069704, + "learning_rate": 2.541550007621651e-06, + "loss": 0.87305629, + "num_input_tokens_seen": 77138160, + "step": 3579, + "time_per_iteration": 2.7344863414764404 + }, + { + "auxiliary_loss_clip": 0.01197247, + "auxiliary_loss_mlp": 0.0102618, + "balance_loss_clip": 1.05597198, + "balance_loss_mlp": 1.01828837, + "epoch": 0.43046954848794566, + "flos": 28184382264960.0, + "grad_norm": 3.851478138872625, + "language_loss": 0.80070138, + "learning_rate": 2.5408000952230156e-06, + "loss": 0.82293564, + "num_input_tokens_seen": 77156950, + "step": 3580, + "time_per_iteration": 3.60856294631958 + }, + { + "auxiliary_loss_clip": 0.01226055, + "auxiliary_loss_mlp": 0.01032117, + "balance_loss_clip": 1.05157375, + "balance_loss_mlp": 1.02311659, + "epoch": 0.4305897913785847, + "flos": 28580476515840.0, + "grad_norm": 2.0172017270873015, + "language_loss": 0.90494937, + "learning_rate": 2.5400501007854357e-06, + "loss": 0.92753112, + "num_input_tokens_seen": 77176395, + "step": 3581, + "time_per_iteration": 2.7793185710906982 + }, + { + "auxiliary_loss_clip": 0.01220477, + "auxiliary_loss_mlp": 0.01023771, + "balance_loss_clip": 1.04747796, + "balance_loss_mlp": 1.01540828, + "epoch": 0.43071003426922383, + "flos": 20448721353600.0, + "grad_norm": 1.986604092613336, + "language_loss": 0.75438094, + "learning_rate": 2.539300024422685e-06, + "loss": 0.7768234, + "num_input_tokens_seen": 77194340, + "step": 3582, + "time_per_iteration": 2.6813430786132812 + }, + { + "auxiliary_loss_clip": 0.01117745, + "auxiliary_loss_mlp": 0.01009353, + "balance_loss_clip": 1.01774883, + "balance_loss_mlp": 1.00804186, + "epoch": 0.43083027715986294, + "flos": 51997969883520.0, + "grad_norm": 0.7923648976428543, + "language_loss": 0.60908127, + "learning_rate": 2.538549866248549e-06, + "loss": 0.63035214, + "num_input_tokens_seen": 77249320, + "step": 3583, + "time_per_iteration": 3.1149098873138428 + }, + { + "auxiliary_loss_clip": 0.01198225, + "auxiliary_loss_mlp": 0.01026399, + "balance_loss_clip": 1.05428159, + "balance_loss_mlp": 1.01809549, + "epoch": 0.430950520050502, + "flos": 16690885320960.0, + "grad_norm": 2.31636855653466, + "language_loss": 0.8150211, + "learning_rate": 2.5377996263768274e-06, + "loss": 0.83726734, + "num_input_tokens_seen": 77267400, + "step": 3584, + "time_per_iteration": 2.6429827213287354 + }, + { + "auxiliary_loss_clip": 0.01191185, + "auxiliary_loss_mlp": 0.01027639, + "balance_loss_clip": 1.05253386, + "balance_loss_mlp": 1.01954508, + "epoch": 0.4310707629411411, + "flos": 24608433726720.0, + "grad_norm": 1.7620985615691651, + "language_loss": 0.68677175, + "learning_rate": 2.5370493049213293e-06, + "loss": 0.70896, + "num_input_tokens_seen": 77287045, + "step": 3585, + "time_per_iteration": 2.6643104553222656 + }, + { + "auxiliary_loss_clip": 0.012439, + "auxiliary_loss_mlp": 0.01026617, + "balance_loss_clip": 1.04600477, + "balance_loss_mlp": 1.01816463, + "epoch": 0.4311910058317802, + "flos": 26432983019520.0, + "grad_norm": 2.289031859757682, + "language_loss": 0.79791176, + "learning_rate": 2.536298901995878e-06, + "loss": 0.82061696, + "num_input_tokens_seen": 77306255, + "step": 3586, + "time_per_iteration": 3.011251211166382 + }, + { + "auxiliary_loss_clip": 0.012128, + "auxiliary_loss_mlp": 0.01024549, + "balance_loss_clip": 1.05642045, + "balance_loss_mlp": 1.01621068, + "epoch": 0.43131124872241927, + "flos": 25155891889920.0, + "grad_norm": 1.7377671147665918, + "language_loss": 0.8009761, + "learning_rate": 2.535548417714311e-06, + "loss": 0.8233496, + "num_input_tokens_seen": 77325555, + "step": 3587, + "time_per_iteration": 2.9761576652526855 + }, + { + "auxiliary_loss_clip": 0.01201974, + "auxiliary_loss_mlp": 0.01031208, + "balance_loss_clip": 1.05327487, + "balance_loss_mlp": 1.02304864, + "epoch": 0.4314314916130584, + "flos": 21614812479360.0, + "grad_norm": 1.5221543895428131, + "language_loss": 0.87171245, + "learning_rate": 2.534797852190474e-06, + "loss": 0.89404428, + "num_input_tokens_seen": 77345735, + "step": 3588, + "time_per_iteration": 2.694197654724121 + }, + { + "auxiliary_loss_clip": 0.01192883, + "auxiliary_loss_mlp": 0.01026978, + "balance_loss_clip": 1.05223632, + "balance_loss_mlp": 1.01888335, + "epoch": 0.4315517345036975, + "flos": 19275016544640.0, + "grad_norm": 2.154932052291633, + "language_loss": 0.81918848, + "learning_rate": 2.5340472055382283e-06, + "loss": 0.84138703, + "num_input_tokens_seen": 77361765, + "step": 3589, + "time_per_iteration": 2.647594690322876 + }, + { + "auxiliary_loss_clip": 0.01211231, + "auxiliary_loss_mlp": 0.01023858, + "balance_loss_clip": 1.04697275, + "balance_loss_mlp": 1.0159483, + "epoch": 0.43167197739433655, + "flos": 24273853516800.0, + "grad_norm": 1.9830537171877427, + "language_loss": 0.80979186, + "learning_rate": 2.5332964778714468e-06, + "loss": 0.83214277, + "num_input_tokens_seen": 77378950, + "step": 3590, + "time_per_iteration": 2.7538247108459473 + }, + { + "auxiliary_loss_clip": 0.01217858, + "auxiliary_loss_mlp": 0.01028661, + "balance_loss_clip": 1.05578732, + "balance_loss_mlp": 1.02013755, + "epoch": 0.43179222028497566, + "flos": 16867816738560.0, + "grad_norm": 1.5464821024896154, + "language_loss": 0.66013968, + "learning_rate": 2.5325456693040123e-06, + "loss": 0.68260491, + "num_input_tokens_seen": 77396145, + "step": 3591, + "time_per_iteration": 2.6930081844329834 + }, + { + "auxiliary_loss_clip": 0.01203042, + "auxiliary_loss_mlp": 0.01023456, + "balance_loss_clip": 1.05271184, + "balance_loss_mlp": 1.01480198, + "epoch": 0.43191246317561477, + "flos": 17639214243840.0, + "grad_norm": 2.5076104873875673, + "language_loss": 0.74859589, + "learning_rate": 2.531794779949824e-06, + "loss": 0.77086091, + "num_input_tokens_seen": 77414045, + "step": 3592, + "time_per_iteration": 3.6384167671203613 + }, + { + "auxiliary_loss_clip": 0.01203871, + "auxiliary_loss_mlp": 0.01025398, + "balance_loss_clip": 1.04903877, + "balance_loss_mlp": 1.01753056, + "epoch": 0.4320327060662538, + "flos": 23878800760320.0, + "grad_norm": 1.7200798015265175, + "language_loss": 0.87949055, + "learning_rate": 2.5310438099227903e-06, + "loss": 0.90178323, + "num_input_tokens_seen": 77431310, + "step": 3593, + "time_per_iteration": 2.819674015045166 + }, + { + "auxiliary_loss_clip": 0.01093435, + "auxiliary_loss_mlp": 0.01002193, + "balance_loss_clip": 1.01778698, + "balance_loss_mlp": 1.00073278, + "epoch": 0.43215294895689293, + "flos": 66394917959040.0, + "grad_norm": 0.8027667326188312, + "language_loss": 0.53437471, + "learning_rate": 2.530292759336833e-06, + "loss": 0.55533099, + "num_input_tokens_seen": 77492045, + "step": 3594, + "time_per_iteration": 3.265228509902954 + }, + { + "auxiliary_loss_clip": 0.01206108, + "auxiliary_loss_mlp": 0.01027198, + "balance_loss_clip": 1.05413485, + "balance_loss_mlp": 1.01882315, + "epoch": 0.432273191847532, + "flos": 20594267262720.0, + "grad_norm": 2.6176210429331888, + "language_loss": 0.69691396, + "learning_rate": 2.5295416283058855e-06, + "loss": 0.71924698, + "num_input_tokens_seen": 77510910, + "step": 3595, + "time_per_iteration": 2.702402114868164 + }, + { + "auxiliary_loss_clip": 0.01199055, + "auxiliary_loss_mlp": 0.01378939, + "balance_loss_clip": 1.0499512, + "balance_loss_mlp": 1.0, + "epoch": 0.4323934347381711, + "flos": 19282127437440.0, + "grad_norm": 1.7079186910341269, + "language_loss": 0.66085768, + "learning_rate": 2.5287904169438943e-06, + "loss": 0.68663764, + "num_input_tokens_seen": 77530115, + "step": 3596, + "time_per_iteration": 2.6910202503204346 + }, + { + "auxiliary_loss_clip": 0.01250254, + "auxiliary_loss_mlp": 0.01035912, + "balance_loss_clip": 1.04994559, + "balance_loss_mlp": 1.02701902, + "epoch": 0.4325136776288102, + "flos": 21726315273600.0, + "grad_norm": 4.09073897966191, + "language_loss": 0.64243668, + "learning_rate": 2.528039125364817e-06, + "loss": 0.66529834, + "num_input_tokens_seen": 77548920, + "step": 3597, + "time_per_iteration": 2.9000985622406006 + }, + { + "auxiliary_loss_clip": 0.01219626, + "auxiliary_loss_mlp": 0.01028885, + "balance_loss_clip": 1.05308747, + "balance_loss_mlp": 1.02012944, + "epoch": 0.43263392051944927, + "flos": 22340746344960.0, + "grad_norm": 2.437368799358761, + "language_loss": 0.76034212, + "learning_rate": 2.5272877536826246e-06, + "loss": 0.78282726, + "num_input_tokens_seen": 77567715, + "step": 3598, + "time_per_iteration": 2.7418041229248047 + }, + { + "auxiliary_loss_clip": 0.01220831, + "auxiliary_loss_mlp": 0.0102688, + "balance_loss_clip": 1.04454017, + "balance_loss_mlp": 1.01882732, + "epoch": 0.4327541634100884, + "flos": 29168406328320.0, + "grad_norm": 2.580792464198517, + "language_loss": 0.70328295, + "learning_rate": 2.5265363020112986e-06, + "loss": 0.72575998, + "num_input_tokens_seen": 77588035, + "step": 3599, + "time_per_iteration": 4.739888906478882 + }, + { + "auxiliary_loss_clip": 0.01198055, + "auxiliary_loss_mlp": 0.01027407, + "balance_loss_clip": 1.05652189, + "balance_loss_mlp": 1.01880646, + "epoch": 0.4328744063007275, + "flos": 26067448264320.0, + "grad_norm": 1.9923870972490967, + "language_loss": 0.84307837, + "learning_rate": 2.5257847704648344e-06, + "loss": 0.86533296, + "num_input_tokens_seen": 77609265, + "step": 3600, + "time_per_iteration": 3.8272695541381836 + }, + { + "auxiliary_loss_clip": 0.01189666, + "auxiliary_loss_mlp": 0.01027144, + "balance_loss_clip": 1.05648923, + "balance_loss_mlp": 1.01924086, + "epoch": 0.43299464919136654, + "flos": 16581357774720.0, + "grad_norm": 1.7089033785298955, + "language_loss": 0.75226361, + "learning_rate": 2.525033159157239e-06, + "loss": 0.77443177, + "num_input_tokens_seen": 77625580, + "step": 3601, + "time_per_iteration": 2.757535696029663 + }, + { + "auxiliary_loss_clip": 0.01192554, + "auxiliary_loss_mlp": 0.01029322, + "balance_loss_clip": 1.05255854, + "balance_loss_mlp": 1.02008963, + "epoch": 0.43311489208200565, + "flos": 16107265140480.0, + "grad_norm": 1.7574774467729641, + "language_loss": 0.77382052, + "learning_rate": 2.52428146820253e-06, + "loss": 0.79603928, + "num_input_tokens_seen": 77643835, + "step": 3602, + "time_per_iteration": 2.560293197631836 + }, + { + "auxiliary_loss_clip": 0.01218402, + "auxiliary_loss_mlp": 0.01032104, + "balance_loss_clip": 1.05301952, + "balance_loss_mlp": 1.02327025, + "epoch": 0.43323513497264476, + "flos": 22930220442240.0, + "grad_norm": 1.9410886269254, + "language_loss": 0.81800151, + "learning_rate": 2.52352969771474e-06, + "loss": 0.84050661, + "num_input_tokens_seen": 77663060, + "step": 3603, + "time_per_iteration": 2.711829900741577 + }, + { + "auxiliary_loss_clip": 0.01205116, + "auxiliary_loss_mlp": 0.01026551, + "balance_loss_clip": 1.05119109, + "balance_loss_mlp": 1.01828432, + "epoch": 0.4333553778632838, + "flos": 25299031587840.0, + "grad_norm": 2.1656681234912925, + "language_loss": 0.88472712, + "learning_rate": 2.5227778478079106e-06, + "loss": 0.90704381, + "num_input_tokens_seen": 77682470, + "step": 3604, + "time_per_iteration": 2.7112224102020264 + }, + { + "auxiliary_loss_clip": 0.01190708, + "auxiliary_loss_mlp": 0.01033813, + "balance_loss_clip": 1.05144644, + "balance_loss_mlp": 1.02575469, + "epoch": 0.43347562075392293, + "flos": 19387165783680.0, + "grad_norm": 1.5657419862134492, + "language_loss": 0.76768911, + "learning_rate": 2.522025918596098e-06, + "loss": 0.78993434, + "num_input_tokens_seen": 77700770, + "step": 3605, + "time_per_iteration": 2.6188690662384033 + }, + { + "auxiliary_loss_clip": 0.01195805, + "auxiliary_loss_mlp": 0.01027114, + "balance_loss_clip": 1.05195951, + "balance_loss_mlp": 1.01986623, + "epoch": 0.43359586364456204, + "flos": 26325969425280.0, + "grad_norm": 2.3992434607726856, + "language_loss": 0.65655434, + "learning_rate": 2.521273910193368e-06, + "loss": 0.67878354, + "num_input_tokens_seen": 77723950, + "step": 3606, + "time_per_iteration": 3.718946933746338 + }, + { + "auxiliary_loss_clip": 0.0120459, + "auxiliary_loss_mlp": 0.01027599, + "balance_loss_clip": 1.05516279, + "balance_loss_mlp": 1.01946926, + "epoch": 0.4337161065352011, + "flos": 15989261984640.0, + "grad_norm": 2.160042653447159, + "language_loss": 0.87410343, + "learning_rate": 2.5205218227138006e-06, + "loss": 0.89642537, + "num_input_tokens_seen": 77736905, + "step": 3607, + "time_per_iteration": 2.6271777153015137 + }, + { + "auxiliary_loss_clip": 0.01190794, + "auxiliary_loss_mlp": 0.01032976, + "balance_loss_clip": 1.05694497, + "balance_loss_mlp": 1.02493513, + "epoch": 0.4338363494258402, + "flos": 20224710184320.0, + "grad_norm": 1.897483342880546, + "language_loss": 0.7920233, + "learning_rate": 2.519769656271486e-06, + "loss": 0.81426096, + "num_input_tokens_seen": 77754325, + "step": 3608, + "time_per_iteration": 2.6998531818389893 + }, + { + "auxiliary_loss_clip": 0.01207249, + "auxiliary_loss_mlp": 0.01034851, + "balance_loss_clip": 1.04484224, + "balance_loss_mlp": 1.0264405, + "epoch": 0.43395659231647926, + "flos": 20083904870400.0, + "grad_norm": 2.064782230577604, + "language_loss": 0.67244709, + "learning_rate": 2.5190174109805285e-06, + "loss": 0.69486809, + "num_input_tokens_seen": 77774150, + "step": 3609, + "time_per_iteration": 2.7911384105682373 + }, + { + "auxiliary_loss_clip": 0.01192121, + "auxiliary_loss_mlp": 0.01029492, + "balance_loss_clip": 1.04805934, + "balance_loss_mlp": 1.02127218, + "epoch": 0.43407683520711837, + "flos": 19901801894400.0, + "grad_norm": 1.8546324886485135, + "language_loss": 0.63904309, + "learning_rate": 2.518265086955042e-06, + "loss": 0.66125929, + "num_input_tokens_seen": 77791870, + "step": 3610, + "time_per_iteration": 2.63204026222229 + }, + { + "auxiliary_loss_clip": 0.01186659, + "auxiliary_loss_mlp": 0.01028922, + "balance_loss_clip": 1.05375481, + "balance_loss_mlp": 1.02075028, + "epoch": 0.4341970780977575, + "flos": 23108732058240.0, + "grad_norm": 1.847288782165669, + "language_loss": 0.83564359, + "learning_rate": 2.5175126843091534e-06, + "loss": 0.85779941, + "num_input_tokens_seen": 77811240, + "step": 3611, + "time_per_iteration": 2.5815610885620117 + }, + { + "auxiliary_loss_clip": 0.0120866, + "auxiliary_loss_mlp": 0.01032479, + "balance_loss_clip": 1.05334425, + "balance_loss_mlp": 1.0247128, + "epoch": 0.43431732098839654, + "flos": 37408288406400.0, + "grad_norm": 2.3743094309367994, + "language_loss": 0.75287199, + "learning_rate": 2.5167602031570034e-06, + "loss": 0.77528334, + "num_input_tokens_seen": 77831425, + "step": 3612, + "time_per_iteration": 2.794874429702759 + }, + { + "auxiliary_loss_clip": 0.01187528, + "auxiliary_loss_mlp": 0.01031886, + "balance_loss_clip": 1.05452871, + "balance_loss_mlp": 1.02336574, + "epoch": 0.43443756387903565, + "flos": 31868206323840.0, + "grad_norm": 1.7406351801717914, + "language_loss": 0.73937315, + "learning_rate": 2.51600764361274e-06, + "loss": 0.76156723, + "num_input_tokens_seen": 77852950, + "step": 3613, + "time_per_iteration": 2.703700542449951 + }, + { + "auxiliary_loss_clip": 0.01188482, + "auxiliary_loss_mlp": 0.01024968, + "balance_loss_clip": 1.05463827, + "balance_loss_mlp": 1.01643312, + "epoch": 0.43455780676967476, + "flos": 23477139901440.0, + "grad_norm": 4.806191469063813, + "language_loss": 0.78647673, + "learning_rate": 2.5152550057905283e-06, + "loss": 0.80861127, + "num_input_tokens_seen": 77872840, + "step": 3614, + "time_per_iteration": 2.6043941974639893 + }, + { + "auxiliary_loss_clip": 0.01196403, + "auxiliary_loss_mlp": 0.01379537, + "balance_loss_clip": 1.05261469, + "balance_loss_mlp": 1.0000906, + "epoch": 0.4346780496603138, + "flos": 24207060176640.0, + "grad_norm": 2.260836406405554, + "language_loss": 0.7679109, + "learning_rate": 2.5145022898045415e-06, + "loss": 0.7936703, + "num_input_tokens_seen": 77892025, + "step": 3615, + "time_per_iteration": 2.71989369392395 + }, + { + "auxiliary_loss_clip": 0.01206229, + "auxiliary_loss_mlp": 0.01031164, + "balance_loss_clip": 1.05084944, + "balance_loss_mlp": 1.02211583, + "epoch": 0.4347982925509529, + "flos": 17092366611840.0, + "grad_norm": 2.695798228805677, + "language_loss": 0.89893937, + "learning_rate": 2.5137494957689664e-06, + "loss": 0.92131329, + "num_input_tokens_seen": 77907635, + "step": 3616, + "time_per_iteration": 2.6164746284484863 + }, + { + "auxiliary_loss_clip": 0.01110077, + "auxiliary_loss_mlp": 0.01000565, + "balance_loss_clip": 1.02231455, + "balance_loss_mlp": 0.99924797, + "epoch": 0.43491853544159204, + "flos": 60945544696320.0, + "grad_norm": 0.7654981100890261, + "language_loss": 0.57331467, + "learning_rate": 2.5129966237980016e-06, + "loss": 0.59442115, + "num_input_tokens_seen": 77970630, + "step": 3617, + "time_per_iteration": 3.270381212234497 + }, + { + "auxiliary_loss_clip": 0.01217537, + "auxiliary_loss_mlp": 0.01025283, + "balance_loss_clip": 1.05217624, + "balance_loss_mlp": 1.01690292, + "epoch": 0.4350387783322311, + "flos": 21944652094080.0, + "grad_norm": 2.0703294937692758, + "language_loss": 0.78121006, + "learning_rate": 2.512243674005857e-06, + "loss": 0.80363834, + "num_input_tokens_seen": 77989995, + "step": 3618, + "time_per_iteration": 2.684422016143799 + }, + { + "auxiliary_loss_clip": 0.01229504, + "auxiliary_loss_mlp": 0.01030833, + "balance_loss_clip": 1.04890335, + "balance_loss_mlp": 1.02242839, + "epoch": 0.4351590212228702, + "flos": 25082705928960.0, + "grad_norm": 1.744149823393357, + "language_loss": 0.86254644, + "learning_rate": 2.5114906465067537e-06, + "loss": 0.88514972, + "num_input_tokens_seen": 78010980, + "step": 3619, + "time_per_iteration": 2.8504114151000977 + }, + { + "auxiliary_loss_clip": 0.01194866, + "auxiliary_loss_mlp": 0.01026774, + "balance_loss_clip": 1.0506568, + "balance_loss_mlp": 1.01890004, + "epoch": 0.4352792641135093, + "flos": 21506541909120.0, + "grad_norm": 2.5891783797284393, + "language_loss": 0.74885595, + "learning_rate": 2.5107375414149264e-06, + "loss": 0.77107227, + "num_input_tokens_seen": 78030225, + "step": 3620, + "time_per_iteration": 2.733981132507324 + }, + { + "auxiliary_loss_clip": 0.01211864, + "auxiliary_loss_mlp": 0.01031624, + "balance_loss_clip": 1.04514587, + "balance_loss_mlp": 1.02283812, + "epoch": 0.43539950700414837, + "flos": 16253457494400.0, + "grad_norm": 2.3115821331313167, + "language_loss": 0.71799755, + "learning_rate": 2.5099843588446197e-06, + "loss": 0.74043238, + "num_input_tokens_seen": 78048545, + "step": 3621, + "time_per_iteration": 2.8059377670288086 + }, + { + "auxiliary_loss_clip": 0.01226997, + "auxiliary_loss_mlp": 0.01028959, + "balance_loss_clip": 1.05018842, + "balance_loss_mlp": 1.02124655, + "epoch": 0.4355197498947875, + "flos": 16691819074560.0, + "grad_norm": 2.7425448995929527, + "language_loss": 0.61342657, + "learning_rate": 2.509231098910091e-06, + "loss": 0.63598621, + "num_input_tokens_seen": 78068415, + "step": 3622, + "time_per_iteration": 2.7888357639312744 + }, + { + "auxiliary_loss_clip": 0.01199665, + "auxiliary_loss_mlp": 0.01028003, + "balance_loss_clip": 1.05243587, + "balance_loss_mlp": 1.01984859, + "epoch": 0.4356399927854266, + "flos": 16362733645440.0, + "grad_norm": 2.265021615240272, + "language_loss": 0.75025213, + "learning_rate": 2.508477761725611e-06, + "loss": 0.77252883, + "num_input_tokens_seen": 78086690, + "step": 3623, + "time_per_iteration": 2.7176413536071777 + }, + { + "auxiliary_loss_clip": 0.01203029, + "auxiliary_loss_mlp": 0.01032967, + "balance_loss_clip": 1.05527544, + "balance_loss_mlp": 1.0246526, + "epoch": 0.43576023567606564, + "flos": 17202037812480.0, + "grad_norm": 2.2427034382213518, + "language_loss": 0.80679637, + "learning_rate": 2.507724347405458e-06, + "loss": 0.8291564, + "num_input_tokens_seen": 78104640, + "step": 3624, + "time_per_iteration": 2.69073224067688 + }, + { + "auxiliary_loss_clip": 0.0121194, + "auxiliary_loss_mlp": 0.01029452, + "balance_loss_clip": 1.04301143, + "balance_loss_mlp": 1.02188182, + "epoch": 0.43588047856670475, + "flos": 15917656222080.0, + "grad_norm": 2.2098334682212735, + "language_loss": 0.81652921, + "learning_rate": 2.5069708560639243e-06, + "loss": 0.83894312, + "num_input_tokens_seen": 78122550, + "step": 3625, + "time_per_iteration": 3.640756130218506 + }, + { + "auxiliary_loss_clip": 0.01218365, + "auxiliary_loss_mlp": 0.01028716, + "balance_loss_clip": 1.05228972, + "balance_loss_mlp": 1.01963842, + "epoch": 0.4360007214573438, + "flos": 23659566099840.0, + "grad_norm": 2.1664482757293246, + "language_loss": 0.61016941, + "learning_rate": 2.5062172878153158e-06, + "loss": 0.63264024, + "num_input_tokens_seen": 78141825, + "step": 3626, + "time_per_iteration": 3.652808666229248 + }, + { + "auxiliary_loss_clip": 0.012391, + "auxiliary_loss_mlp": 0.0102389, + "balance_loss_clip": 1.0477519, + "balance_loss_mlp": 1.01562333, + "epoch": 0.4361209643479829, + "flos": 21978767036160.0, + "grad_norm": 1.7491968541347107, + "language_loss": 0.87258559, + "learning_rate": 2.505463642773947e-06, + "loss": 0.89521551, + "num_input_tokens_seen": 78161790, + "step": 3627, + "time_per_iteration": 2.8082633018493652 + }, + { + "auxiliary_loss_clip": 0.01211033, + "auxiliary_loss_mlp": 0.01379252, + "balance_loss_clip": 1.04816818, + "balance_loss_mlp": 0.9999938, + "epoch": 0.43624120723862203, + "flos": 17420159151360.0, + "grad_norm": 2.6879823183210476, + "language_loss": 0.7520467, + "learning_rate": 2.504709921054146e-06, + "loss": 0.77794951, + "num_input_tokens_seen": 78178605, + "step": 3628, + "time_per_iteration": 2.719780683517456 + }, + { + "auxiliary_loss_clip": 0.01204773, + "auxiliary_loss_mlp": 0.01033378, + "balance_loss_clip": 1.04597557, + "balance_loss_mlp": 1.02394283, + "epoch": 0.4363614501292611, + "flos": 17895293280000.0, + "grad_norm": 2.650678548705882, + "language_loss": 0.83825505, + "learning_rate": 2.50395612277025e-06, + "loss": 0.86063659, + "num_input_tokens_seen": 78194460, + "step": 3629, + "time_per_iteration": 2.759937047958374 + }, + { + "auxiliary_loss_clip": 0.0121121, + "auxiliary_loss_mlp": 0.01036371, + "balance_loss_clip": 1.05224133, + "balance_loss_mlp": 1.02809787, + "epoch": 0.4364816930199002, + "flos": 20302888135680.0, + "grad_norm": 2.382377957467713, + "language_loss": 0.73119235, + "learning_rate": 2.503202248036612e-06, + "loss": 0.75366813, + "num_input_tokens_seen": 78213315, + "step": 3630, + "time_per_iteration": 2.7164394855499268 + }, + { + "auxiliary_loss_clip": 0.01189075, + "auxiliary_loss_mlp": 0.01028738, + "balance_loss_clip": 1.05650306, + "balance_loss_mlp": 1.01982737, + "epoch": 0.4366019359105393, + "flos": 24061334699520.0, + "grad_norm": 1.814655250377807, + "language_loss": 0.73498923, + "learning_rate": 2.5024482969675927e-06, + "loss": 0.75716734, + "num_input_tokens_seen": 78233270, + "step": 3631, + "time_per_iteration": 3.6205122470855713 + }, + { + "auxiliary_loss_clip": 0.0122665, + "auxiliary_loss_mlp": 0.01027996, + "balance_loss_clip": 1.04820824, + "balance_loss_mlp": 1.02008033, + "epoch": 0.43672217880117836, + "flos": 21754109422080.0, + "grad_norm": 2.180437280620313, + "language_loss": 0.84128165, + "learning_rate": 2.501694269677566e-06, + "loss": 0.86382806, + "num_input_tokens_seen": 78251040, + "step": 3632, + "time_per_iteration": 2.7247684001922607 + }, + { + "auxiliary_loss_clip": 0.01201437, + "auxiliary_loss_mlp": 0.01030879, + "balance_loss_clip": 1.05382776, + "balance_loss_mlp": 1.02223659, + "epoch": 0.4368424216918175, + "flos": 18035200753920.0, + "grad_norm": 2.7291984849642548, + "language_loss": 0.80365038, + "learning_rate": 2.500940166280918e-06, + "loss": 0.82597351, + "num_input_tokens_seen": 78269470, + "step": 3633, + "time_per_iteration": 2.617938995361328 + }, + { + "auxiliary_loss_clip": 0.01192512, + "auxiliary_loss_mlp": 0.01027514, + "balance_loss_clip": 1.05205655, + "balance_loss_mlp": 1.0190382, + "epoch": 0.4369626645824566, + "flos": 25447127362560.0, + "grad_norm": 1.9788686130910311, + "language_loss": 0.7973935, + "learning_rate": 2.500185986892045e-06, + "loss": 0.81959379, + "num_input_tokens_seen": 78288955, + "step": 3634, + "time_per_iteration": 2.7738595008850098 + }, + { + "auxiliary_loss_clip": 0.01193978, + "auxiliary_loss_mlp": 0.01028659, + "balance_loss_clip": 1.05324936, + "balance_loss_mlp": 1.01971793, + "epoch": 0.43708290747309564, + "flos": 25302694775040.0, + "grad_norm": 2.105291814603591, + "language_loss": 0.77380949, + "learning_rate": 2.499431731625355e-06, + "loss": 0.79603589, + "num_input_tokens_seen": 78307980, + "step": 3635, + "time_per_iteration": 2.6547489166259766 + }, + { + "auxiliary_loss_clip": 0.01190913, + "auxiliary_loss_mlp": 0.01030123, + "balance_loss_clip": 1.05608094, + "balance_loss_mlp": 1.02169502, + "epoch": 0.43720315036373475, + "flos": 31575103344000.0, + "grad_norm": 2.0500711545073225, + "language_loss": 0.7923677, + "learning_rate": 2.4986774005952686e-06, + "loss": 0.81457806, + "num_input_tokens_seen": 78330355, + "step": 3636, + "time_per_iteration": 2.7786779403686523 + }, + { + "auxiliary_loss_clip": 0.01196268, + "auxiliary_loss_mlp": 0.01029815, + "balance_loss_clip": 1.05499029, + "balance_loss_mlp": 1.02176213, + "epoch": 0.43732339325437386, + "flos": 23112000195840.0, + "grad_norm": 2.1227522359920745, + "language_loss": 0.84607053, + "learning_rate": 2.4979229939162166e-06, + "loss": 0.86833137, + "num_input_tokens_seen": 78349135, + "step": 3637, + "time_per_iteration": 2.657740592956543 + }, + { + "auxiliary_loss_clip": 0.01191698, + "auxiliary_loss_mlp": 0.01028943, + "balance_loss_clip": 1.05213857, + "balance_loss_mlp": 1.02106023, + "epoch": 0.4374436361450129, + "flos": 27746272080000.0, + "grad_norm": 1.5891371037434743, + "language_loss": 0.80674624, + "learning_rate": 2.4971685117026433e-06, + "loss": 0.82895267, + "num_input_tokens_seen": 78368900, + "step": 3638, + "time_per_iteration": 2.747014284133911 + }, + { + "auxiliary_loss_clip": 0.01202081, + "auxiliary_loss_mlp": 0.01027571, + "balance_loss_clip": 1.05567157, + "balance_loss_mlp": 1.01963162, + "epoch": 0.437563879035652, + "flos": 24172370616960.0, + "grad_norm": 1.5981616171966635, + "language_loss": 0.76784205, + "learning_rate": 2.4964139540690018e-06, + "loss": 0.7901386, + "num_input_tokens_seen": 78392235, + "step": 3639, + "time_per_iteration": 2.674344778060913 + }, + { + "auxiliary_loss_clip": 0.01216292, + "auxiliary_loss_mlp": 0.01031082, + "balance_loss_clip": 1.05192482, + "balance_loss_mlp": 1.02280951, + "epoch": 0.4376841219262911, + "flos": 23477211728640.0, + "grad_norm": 2.1043080062164226, + "language_loss": 0.72889042, + "learning_rate": 2.495659321129758e-06, + "loss": 0.75136417, + "num_input_tokens_seen": 78409980, + "step": 3640, + "time_per_iteration": 2.7585091590881348 + }, + { + "auxiliary_loss_clip": 0.01193609, + "auxiliary_loss_mlp": 0.01029414, + "balance_loss_clip": 1.05342424, + "balance_loss_mlp": 1.02130187, + "epoch": 0.4378043648169302, + "flos": 25447809720960.0, + "grad_norm": 1.822642169945931, + "language_loss": 0.75623339, + "learning_rate": 2.494904612999389e-06, + "loss": 0.7784636, + "num_input_tokens_seen": 78428690, + "step": 3641, + "time_per_iteration": 2.6547443866729736 + }, + { + "auxiliary_loss_clip": 0.01095011, + "auxiliary_loss_mlp": 0.01002698, + "balance_loss_clip": 1.02315354, + "balance_loss_mlp": 1.0013032, + "epoch": 0.4379246077075693, + "flos": 53914056986880.0, + "grad_norm": 0.7490365129493706, + "language_loss": 0.56547254, + "learning_rate": 2.4941498297923843e-06, + "loss": 0.58644962, + "num_input_tokens_seen": 78489260, + "step": 3642, + "time_per_iteration": 3.2083818912506104 + }, + { + "auxiliary_loss_clip": 0.0119521, + "auxiliary_loss_mlp": 0.01026828, + "balance_loss_clip": 1.05517113, + "balance_loss_mlp": 1.01890099, + "epoch": 0.43804485059820836, + "flos": 20588305605120.0, + "grad_norm": 2.012173828496252, + "language_loss": 0.69686431, + "learning_rate": 2.4933949716232424e-06, + "loss": 0.71908474, + "num_input_tokens_seen": 78506785, + "step": 3643, + "time_per_iteration": 2.703558921813965 + }, + { + "auxiliary_loss_clip": 0.01214013, + "auxiliary_loss_mlp": 0.01025768, + "balance_loss_clip": 1.05189633, + "balance_loss_mlp": 1.0178045, + "epoch": 0.43816509348884747, + "flos": 23876214981120.0, + "grad_norm": 2.130227807616928, + "language_loss": 0.74085844, + "learning_rate": 2.492640038606476e-06, + "loss": 0.76325631, + "num_input_tokens_seen": 78525150, + "step": 3644, + "time_per_iteration": 2.716697931289673 + }, + { + "auxiliary_loss_clip": 0.01196831, + "auxiliary_loss_mlp": 0.01026006, + "balance_loss_clip": 1.05167198, + "balance_loss_mlp": 1.01695836, + "epoch": 0.4382853363794866, + "flos": 14684448533760.0, + "grad_norm": 2.214802485243486, + "language_loss": 0.78299594, + "learning_rate": 2.491885030856608e-06, + "loss": 0.8052243, + "num_input_tokens_seen": 78543245, + "step": 3645, + "time_per_iteration": 2.714293956756592 + }, + { + "auxiliary_loss_clip": 0.01211353, + "auxiliary_loss_mlp": 0.01033122, + "balance_loss_clip": 1.05638504, + "balance_loss_mlp": 1.02415144, + "epoch": 0.43840557927012563, + "flos": 17165301177600.0, + "grad_norm": 2.4523435927139587, + "language_loss": 0.82862616, + "learning_rate": 2.4911299484881713e-06, + "loss": 0.85107094, + "num_input_tokens_seen": 78560775, + "step": 3646, + "time_per_iteration": 2.628676176071167 + }, + { + "auxiliary_loss_clip": 0.01199311, + "auxiliary_loss_mlp": 0.01028968, + "balance_loss_clip": 1.04996121, + "balance_loss_mlp": 1.02118397, + "epoch": 0.43852582216076474, + "flos": 19390685316480.0, + "grad_norm": 1.6756369912474152, + "language_loss": 0.81187809, + "learning_rate": 2.490374791615712e-06, + "loss": 0.83416092, + "num_input_tokens_seen": 78580800, + "step": 3647, + "time_per_iteration": 2.8086512088775635 + }, + { + "auxiliary_loss_clip": 0.01191945, + "auxiliary_loss_mlp": 0.01379492, + "balance_loss_clip": 1.05629897, + "balance_loss_mlp": 1.00007987, + "epoch": 0.43864606505140386, + "flos": 18075133699200.0, + "grad_norm": 2.380196550690307, + "language_loss": 0.78160805, + "learning_rate": 2.4896195603537867e-06, + "loss": 0.80732238, + "num_input_tokens_seen": 78595410, + "step": 3648, + "time_per_iteration": 2.6159920692443848 + }, + { + "auxiliary_loss_clip": 0.01216829, + "auxiliary_loss_mlp": 0.01025941, + "balance_loss_clip": 1.05188823, + "balance_loss_mlp": 1.01736963, + "epoch": 0.4387663079420429, + "flos": 19644896845440.0, + "grad_norm": 1.8195430214182902, + "language_loss": 0.73639649, + "learning_rate": 2.488864254816964e-06, + "loss": 0.75882423, + "num_input_tokens_seen": 78614100, + "step": 3649, + "time_per_iteration": 2.767664670944214 + }, + { + "auxiliary_loss_clip": 0.01198354, + "auxiliary_loss_mlp": 0.01033011, + "balance_loss_clip": 1.05424702, + "balance_loss_mlp": 1.02355766, + "epoch": 0.438886550832682, + "flos": 19719339782400.0, + "grad_norm": 3.0817345644255667, + "language_loss": 0.68291348, + "learning_rate": 2.4881088751198218e-06, + "loss": 0.70522714, + "num_input_tokens_seen": 78632260, + "step": 3650, + "time_per_iteration": 2.7058041095733643 + }, + { + "auxiliary_loss_clip": 0.01205583, + "auxiliary_loss_mlp": 0.01020955, + "balance_loss_clip": 1.050174, + "balance_loss_mlp": 1.01278913, + "epoch": 0.43900679372332113, + "flos": 14536675981440.0, + "grad_norm": 2.7768751299689853, + "language_loss": 0.63906419, + "learning_rate": 2.4873534213769517e-06, + "loss": 0.66132963, + "num_input_tokens_seen": 78647490, + "step": 3651, + "time_per_iteration": 3.6441245079040527 + }, + { + "auxiliary_loss_clip": 0.01214046, + "auxiliary_loss_mlp": 0.01026649, + "balance_loss_clip": 1.05512607, + "balance_loss_mlp": 1.01805997, + "epoch": 0.4391270366139602, + "flos": 24056234968320.0, + "grad_norm": 1.6025442337188793, + "language_loss": 0.71914351, + "learning_rate": 2.4865978937029547e-06, + "loss": 0.74155051, + "num_input_tokens_seen": 78666470, + "step": 3652, + "time_per_iteration": 4.589456558227539 + }, + { + "auxiliary_loss_clip": 0.01210901, + "auxiliary_loss_mlp": 0.01030895, + "balance_loss_clip": 1.0488205, + "balance_loss_mlp": 1.02215743, + "epoch": 0.4392472795045993, + "flos": 31538510363520.0, + "grad_norm": 1.5586367837588389, + "language_loss": 0.66604298, + "learning_rate": 2.485842292212445e-06, + "loss": 0.68846095, + "num_input_tokens_seen": 78687685, + "step": 3653, + "time_per_iteration": 2.862787961959839 + }, + { + "auxiliary_loss_clip": 0.01191554, + "auxiliary_loss_mlp": 0.01031243, + "balance_loss_clip": 1.0575068, + "balance_loss_mlp": 1.02296138, + "epoch": 0.4393675223952384, + "flos": 14866300114560.0, + "grad_norm": 2.4539909939609204, + "language_loss": 0.80134726, + "learning_rate": 2.485086617020045e-06, + "loss": 0.8235752, + "num_input_tokens_seen": 78706180, + "step": 3654, + "time_per_iteration": 2.647303819656372 + }, + { + "auxiliary_loss_clip": 0.01196462, + "auxiliary_loss_mlp": 0.01026201, + "balance_loss_clip": 1.05100131, + "balance_loss_mlp": 1.01723027, + "epoch": 0.43948776528587746, + "flos": 14825900292480.0, + "grad_norm": 2.0817638989512135, + "language_loss": 0.82087183, + "learning_rate": 2.4843308682403903e-06, + "loss": 0.84309846, + "num_input_tokens_seen": 78723095, + "step": 3655, + "time_per_iteration": 2.846405506134033 + }, + { + "auxiliary_loss_clip": 0.01188066, + "auxiliary_loss_mlp": 0.01030053, + "balance_loss_clip": 1.05525041, + "balance_loss_mlp": 1.02157116, + "epoch": 0.4396080081765166, + "flos": 13914523486080.0, + "grad_norm": 1.6819258980565523, + "language_loss": 0.82351792, + "learning_rate": 2.4835750459881294e-06, + "loss": 0.84569907, + "num_input_tokens_seen": 78739720, + "step": 3656, + "time_per_iteration": 2.573734998703003 + }, + { + "auxiliary_loss_clip": 0.01196154, + "auxiliary_loss_mlp": 0.01028763, + "balance_loss_clip": 1.05019736, + "balance_loss_mlp": 1.02014971, + "epoch": 0.43972825106715563, + "flos": 18222978078720.0, + "grad_norm": 1.8376565448969984, + "language_loss": 0.82116389, + "learning_rate": 2.4828191503779177e-06, + "loss": 0.843413, + "num_input_tokens_seen": 78757820, + "step": 3657, + "time_per_iteration": 3.573002815246582 + }, + { + "auxiliary_loss_clip": 0.01212212, + "auxiliary_loss_mlp": 0.01025027, + "balance_loss_clip": 1.04980421, + "balance_loss_mlp": 1.01693273, + "epoch": 0.43984849395779474, + "flos": 16873239692160.0, + "grad_norm": 2.262083573162609, + "language_loss": 0.89928102, + "learning_rate": 2.482063181524425e-06, + "loss": 0.92165339, + "num_input_tokens_seen": 78773720, + "step": 3658, + "time_per_iteration": 2.7116546630859375 + }, + { + "auxiliary_loss_clip": 0.01191638, + "auxiliary_loss_mlp": 0.01027697, + "balance_loss_clip": 1.05727315, + "balance_loss_mlp": 1.01859498, + "epoch": 0.43996873684843385, + "flos": 18691504104960.0, + "grad_norm": 3.072933714241406, + "language_loss": 0.81514072, + "learning_rate": 2.4813071395423307e-06, + "loss": 0.83733416, + "num_input_tokens_seen": 78791285, + "step": 3659, + "time_per_iteration": 2.683872938156128 + }, + { + "auxiliary_loss_clip": 0.01199134, + "auxiliary_loss_mlp": 0.01033839, + "balance_loss_clip": 1.05361462, + "balance_loss_mlp": 1.02507114, + "epoch": 0.4400889797390729, + "flos": 23653460787840.0, + "grad_norm": 2.1998647845675157, + "language_loss": 0.64279848, + "learning_rate": 2.4805510245463263e-06, + "loss": 0.66512823, + "num_input_tokens_seen": 78811440, + "step": 3660, + "time_per_iteration": 2.7874951362609863 + }, + { + "auxiliary_loss_clip": 0.0119783, + "auxiliary_loss_mlp": 0.01028767, + "balance_loss_clip": 1.05430293, + "balance_loss_mlp": 1.01956367, + "epoch": 0.440209222629712, + "flos": 23149203707520.0, + "grad_norm": 1.9328202700742132, + "language_loss": 0.60068977, + "learning_rate": 2.4797948366511137e-06, + "loss": 0.62295574, + "num_input_tokens_seen": 78831150, + "step": 3661, + "time_per_iteration": 2.6881039142608643 + }, + { + "auxiliary_loss_clip": 0.01209941, + "auxiliary_loss_mlp": 0.010287, + "balance_loss_clip": 1.04837894, + "balance_loss_mlp": 1.02041531, + "epoch": 0.4403294655203511, + "flos": 24823394668800.0, + "grad_norm": 2.037057592497111, + "language_loss": 0.76194012, + "learning_rate": 2.4790385759714055e-06, + "loss": 0.78432655, + "num_input_tokens_seen": 78850215, + "step": 3662, + "time_per_iteration": 2.7816505432128906 + }, + { + "auxiliary_loss_clip": 0.01196498, + "auxiliary_loss_mlp": 0.01028773, + "balance_loss_clip": 1.05566764, + "balance_loss_mlp": 1.02054811, + "epoch": 0.4404497084109902, + "flos": 22565080736640.0, + "grad_norm": 1.8226516992290251, + "language_loss": 0.70965987, + "learning_rate": 2.478282242621926e-06, + "loss": 0.73191261, + "num_input_tokens_seen": 78870675, + "step": 3663, + "time_per_iteration": 2.648496389389038 + }, + { + "auxiliary_loss_clip": 0.01120412, + "auxiliary_loss_mlp": 0.01000289, + "balance_loss_clip": 1.02286732, + "balance_loss_mlp": 0.99900132, + "epoch": 0.4405699513016293, + "flos": 64967073448320.0, + "grad_norm": 0.8570630398891631, + "language_loss": 0.59631371, + "learning_rate": 2.477525836717411e-06, + "loss": 0.61752069, + "num_input_tokens_seen": 78938440, + "step": 3664, + "time_per_iteration": 3.4813878536224365 + }, + { + "auxiliary_loss_clip": 0.01199784, + "auxiliary_loss_mlp": 0.01029515, + "balance_loss_clip": 1.05475712, + "balance_loss_mlp": 1.0206635, + "epoch": 0.4406901941922684, + "flos": 35661952978560.0, + "grad_norm": 2.2525932793204917, + "language_loss": 0.79409742, + "learning_rate": 2.476769358372606e-06, + "loss": 0.8163904, + "num_input_tokens_seen": 78960090, + "step": 3665, + "time_per_iteration": 2.811152219772339 + }, + { + "auxiliary_loss_clip": 0.01208456, + "auxiliary_loss_mlp": 0.01029178, + "balance_loss_clip": 1.0509088, + "balance_loss_mlp": 1.02168274, + "epoch": 0.44081043708290746, + "flos": 18040767361920.0, + "grad_norm": 2.5520306062461975, + "language_loss": 0.74989283, + "learning_rate": 2.4760128077022683e-06, + "loss": 0.77226919, + "num_input_tokens_seen": 78978225, + "step": 3666, + "time_per_iteration": 2.7623894214630127 + }, + { + "auxiliary_loss_clip": 0.0121004, + "auxiliary_loss_mlp": 0.01026056, + "balance_loss_clip": 1.04778945, + "balance_loss_mlp": 1.01797938, + "epoch": 0.44093067997354657, + "flos": 30153507799680.0, + "grad_norm": 1.577994973055739, + "language_loss": 0.68351102, + "learning_rate": 2.4752561848211672e-06, + "loss": 0.70587194, + "num_input_tokens_seen": 79000625, + "step": 3667, + "time_per_iteration": 2.8311164379119873 + }, + { + "auxiliary_loss_clip": 0.01193869, + "auxiliary_loss_mlp": 0.01032002, + "balance_loss_clip": 1.05547452, + "balance_loss_mlp": 1.02449787, + "epoch": 0.4410509228641857, + "flos": 23255068066560.0, + "grad_norm": 1.7945303540358335, + "language_loss": 0.70984185, + "learning_rate": 2.4744994898440797e-06, + "loss": 0.73210061, + "num_input_tokens_seen": 79019415, + "step": 3668, + "time_per_iteration": 2.6450588703155518 + }, + { + "auxiliary_loss_clip": 0.01218616, + "auxiliary_loss_mlp": 0.01029271, + "balance_loss_clip": 1.04994977, + "balance_loss_mlp": 1.02109337, + "epoch": 0.44117116575482473, + "flos": 19500571998720.0, + "grad_norm": 3.7391123609341785, + "language_loss": 0.83712709, + "learning_rate": 2.473742722885797e-06, + "loss": 0.85960603, + "num_input_tokens_seen": 79038435, + "step": 3669, + "time_per_iteration": 2.740476369857788 + }, + { + "auxiliary_loss_clip": 0.01199735, + "auxiliary_loss_mlp": 0.01379662, + "balance_loss_clip": 1.05663216, + "balance_loss_mlp": 1.00016868, + "epoch": 0.44129140864546385, + "flos": 27053124353280.0, + "grad_norm": 2.559023726565751, + "language_loss": 0.6512534, + "learning_rate": 2.4729858840611197e-06, + "loss": 0.67704737, + "num_input_tokens_seen": 79057345, + "step": 3670, + "time_per_iteration": 2.679344415664673 + }, + { + "auxiliary_loss_clip": 0.01188036, + "auxiliary_loss_mlp": 0.01031563, + "balance_loss_clip": 1.05586004, + "balance_loss_mlp": 1.02368307, + "epoch": 0.4414116515361029, + "flos": 26102101910400.0, + "grad_norm": 2.112635342380397, + "language_loss": 0.73046261, + "learning_rate": 2.4722289734848605e-06, + "loss": 0.75265861, + "num_input_tokens_seen": 79077810, + "step": 3671, + "time_per_iteration": 2.6760785579681396 + }, + { + "auxiliary_loss_clip": 0.01216908, + "auxiliary_loss_mlp": 0.01024521, + "balance_loss_clip": 1.0549469, + "balance_loss_mlp": 1.01639724, + "epoch": 0.441531894426742, + "flos": 21906083865600.0, + "grad_norm": 2.6129939665247814, + "language_loss": 0.78230894, + "learning_rate": 2.471471991271841e-06, + "loss": 0.80472326, + "num_input_tokens_seen": 79094935, + "step": 3672, + "time_per_iteration": 2.746500253677368 + }, + { + "auxiliary_loss_clip": 0.01188671, + "auxiliary_loss_mlp": 0.01031655, + "balance_loss_clip": 1.05343521, + "balance_loss_mlp": 1.02264261, + "epoch": 0.4416521373173811, + "flos": 23437099215360.0, + "grad_norm": 1.9728557166772114, + "language_loss": 0.79772735, + "learning_rate": 2.470714937536896e-06, + "loss": 0.81993055, + "num_input_tokens_seen": 79113660, + "step": 3673, + "time_per_iteration": 2.6974937915802 + }, + { + "auxiliary_loss_clip": 0.01220626, + "auxiliary_loss_mlp": 0.01029433, + "balance_loss_clip": 1.04866958, + "balance_loss_mlp": 1.02141023, + "epoch": 0.4417723802080202, + "flos": 20334345471360.0, + "grad_norm": 2.0217752393523973, + "language_loss": 0.70719343, + "learning_rate": 2.469957812394868e-06, + "loss": 0.72969401, + "num_input_tokens_seen": 79132470, + "step": 3674, + "time_per_iteration": 2.734570026397705 + }, + { + "auxiliary_loss_clip": 0.01192398, + "auxiliary_loss_mlp": 0.01024394, + "balance_loss_clip": 1.06035924, + "balance_loss_mlp": 1.01606464, + "epoch": 0.4418926230986593, + "flos": 18880682060160.0, + "grad_norm": 2.1974032719404004, + "language_loss": 0.75864315, + "learning_rate": 2.4692006159606148e-06, + "loss": 0.78081113, + "num_input_tokens_seen": 79150000, + "step": 3675, + "time_per_iteration": 2.6684296131134033 + }, + { + "auxiliary_loss_clip": 0.01190312, + "auxiliary_loss_mlp": 0.01028129, + "balance_loss_clip": 1.05629718, + "balance_loss_mlp": 1.01996863, + "epoch": 0.4420128659892984, + "flos": 19464409981440.0, + "grad_norm": 1.8963482612804075, + "language_loss": 0.78328735, + "learning_rate": 2.468443348349e-06, + "loss": 0.80547178, + "num_input_tokens_seen": 79167875, + "step": 3676, + "time_per_iteration": 2.620682716369629 + }, + { + "auxiliary_loss_clip": 0.01217589, + "auxiliary_loss_mlp": 0.01032289, + "balance_loss_clip": 1.0461247, + "balance_loss_mlp": 1.02331829, + "epoch": 0.44213310887993745, + "flos": 17894359526400.0, + "grad_norm": 2.957901092326129, + "language_loss": 0.82633513, + "learning_rate": 2.467686009674902e-06, + "loss": 0.84883392, + "num_input_tokens_seen": 79182325, + "step": 3677, + "time_per_iteration": 3.696080446243286 + }, + { + "auxiliary_loss_clip": 0.01192269, + "auxiliary_loss_mlp": 0.01034795, + "balance_loss_clip": 1.05271983, + "balance_loss_mlp": 1.02527595, + "epoch": 0.44225335177057656, + "flos": 19204667758080.0, + "grad_norm": 2.419152919450061, + "language_loss": 0.8506549, + "learning_rate": 2.466928600053209e-06, + "loss": 0.87292552, + "num_input_tokens_seen": 79197630, + "step": 3678, + "time_per_iteration": 3.538076877593994 + }, + { + "auxiliary_loss_clip": 0.01203854, + "auxiliary_loss_mlp": 0.0102555, + "balance_loss_clip": 1.05119395, + "balance_loss_mlp": 1.01678824, + "epoch": 0.4423735946612157, + "flos": 23471321898240.0, + "grad_norm": 1.9671185164158955, + "language_loss": 0.71619433, + "learning_rate": 2.466171119598818e-06, + "loss": 0.73848832, + "num_input_tokens_seen": 79217600, + "step": 3679, + "time_per_iteration": 2.726132392883301 + }, + { + "auxiliary_loss_clip": 0.01203637, + "auxiliary_loss_mlp": 0.01035034, + "balance_loss_clip": 1.05360329, + "balance_loss_mlp": 1.0254854, + "epoch": 0.44249383755185473, + "flos": 26685398868480.0, + "grad_norm": 1.7634111394187748, + "language_loss": 0.77345884, + "learning_rate": 2.465413568426639e-06, + "loss": 0.79584551, + "num_input_tokens_seen": 79238550, + "step": 3680, + "time_per_iteration": 2.6657142639160156 + }, + { + "auxiliary_loss_clip": 0.011953, + "auxiliary_loss_mlp": 0.01026769, + "balance_loss_clip": 1.05518138, + "balance_loss_mlp": 1.01871014, + "epoch": 0.44261408044249384, + "flos": 23147659422720.0, + "grad_norm": 1.98946212753432, + "language_loss": 0.81061727, + "learning_rate": 2.464655946651591e-06, + "loss": 0.832838, + "num_input_tokens_seen": 79257555, + "step": 3681, + "time_per_iteration": 2.653651714324951 + }, + { + "auxiliary_loss_clip": 0.01198955, + "auxiliary_loss_mlp": 0.01032661, + "balance_loss_clip": 1.05435276, + "balance_loss_mlp": 1.023983, + "epoch": 0.44273432333313295, + "flos": 24462564595200.0, + "grad_norm": 1.9677222226353346, + "language_loss": 0.80806649, + "learning_rate": 2.4638982543886065e-06, + "loss": 0.83038265, + "num_input_tokens_seen": 79277595, + "step": 3682, + "time_per_iteration": 2.690866470336914 + }, + { + "auxiliary_loss_clip": 0.01204643, + "auxiliary_loss_mlp": 0.01033679, + "balance_loss_clip": 1.05833268, + "balance_loss_mlp": 1.0243094, + "epoch": 0.442854566223772, + "flos": 17528932512000.0, + "grad_norm": 2.2807101653801336, + "language_loss": 0.87053114, + "learning_rate": 2.4631404917526254e-06, + "loss": 0.8929143, + "num_input_tokens_seen": 79294550, + "step": 3683, + "time_per_iteration": 3.639434576034546 + }, + { + "auxiliary_loss_clip": 0.01188016, + "auxiliary_loss_mlp": 0.01025554, + "balance_loss_clip": 1.05115378, + "balance_loss_mlp": 1.01753116, + "epoch": 0.4429748091144111, + "flos": 24896293320960.0, + "grad_norm": 1.5512658826724446, + "language_loss": 0.79332894, + "learning_rate": 2.4623826588586e-06, + "loss": 0.81546474, + "num_input_tokens_seen": 79314820, + "step": 3684, + "time_per_iteration": 2.671163320541382 + }, + { + "auxiliary_loss_clip": 0.01198703, + "auxiliary_loss_mlp": 0.01028018, + "balance_loss_clip": 1.04992843, + "balance_loss_mlp": 1.01908362, + "epoch": 0.4430950520050502, + "flos": 21614704738560.0, + "grad_norm": 1.4554385056790606, + "language_loss": 0.82498038, + "learning_rate": 2.461624755821492e-06, + "loss": 0.8472476, + "num_input_tokens_seen": 79334300, + "step": 3685, + "time_per_iteration": 2.6854796409606934 + }, + { + "auxiliary_loss_clip": 0.01218927, + "auxiliary_loss_mlp": 0.01027839, + "balance_loss_clip": 1.05400944, + "balance_loss_mlp": 1.01947093, + "epoch": 0.4432152948956893, + "flos": 24572271709440.0, + "grad_norm": 1.5495748588029061, + "language_loss": 0.76490855, + "learning_rate": 2.4608667827562763e-06, + "loss": 0.78737628, + "num_input_tokens_seen": 79353630, + "step": 3686, + "time_per_iteration": 2.7353384494781494 + }, + { + "auxiliary_loss_clip": 0.01205736, + "auxiliary_loss_mlp": 0.01029219, + "balance_loss_clip": 1.05753851, + "balance_loss_mlp": 1.02073717, + "epoch": 0.4433355377863284, + "flos": 21762261809280.0, + "grad_norm": 2.9112736576901077, + "language_loss": 0.90244126, + "learning_rate": 2.460108739777936e-06, + "loss": 0.92479086, + "num_input_tokens_seen": 79372765, + "step": 3687, + "time_per_iteration": 2.6623408794403076 + }, + { + "auxiliary_loss_clip": 0.01205781, + "auxiliary_loss_mlp": 0.01029718, + "balance_loss_clip": 1.05315542, + "balance_loss_mlp": 1.02141488, + "epoch": 0.44345578067696745, + "flos": 20084479488000.0, + "grad_norm": 1.7289835459919083, + "language_loss": 0.76311892, + "learning_rate": 2.4593506270014656e-06, + "loss": 0.78547394, + "num_input_tokens_seen": 79391735, + "step": 3688, + "time_per_iteration": 2.6701653003692627 + }, + { + "auxiliary_loss_clip": 0.01207227, + "auxiliary_loss_mlp": 0.01025435, + "balance_loss_clip": 1.05111146, + "balance_loss_mlp": 1.01725101, + "epoch": 0.44357602356760656, + "flos": 24169497528960.0, + "grad_norm": 1.5781137516157528, + "language_loss": 0.81838489, + "learning_rate": 2.45859244454187e-06, + "loss": 0.84071153, + "num_input_tokens_seen": 79411525, + "step": 3689, + "time_per_iteration": 2.7484123706817627 + }, + { + "auxiliary_loss_clip": 0.01192982, + "auxiliary_loss_mlp": 0.01025479, + "balance_loss_clip": 1.05443943, + "balance_loss_mlp": 1.01746869, + "epoch": 0.44369626645824567, + "flos": 22707717644160.0, + "grad_norm": 1.8487270128508886, + "language_loss": 0.66266537, + "learning_rate": 2.4578341925141655e-06, + "loss": 0.68485004, + "num_input_tokens_seen": 79430740, + "step": 3690, + "time_per_iteration": 2.647047519683838 + }, + { + "auxiliary_loss_clip": 0.01202498, + "auxiliary_loss_mlp": 0.01031874, + "balance_loss_clip": 1.05401111, + "balance_loss_mlp": 1.02310634, + "epoch": 0.4438165093488847, + "flos": 38030225420160.0, + "grad_norm": 2.564980587405266, + "language_loss": 0.72047037, + "learning_rate": 2.457075871033378e-06, + "loss": 0.74281406, + "num_input_tokens_seen": 79452615, + "step": 3691, + "time_per_iteration": 2.787203073501587 + }, + { + "auxiliary_loss_clip": 0.01214305, + "auxiliary_loss_mlp": 0.01028472, + "balance_loss_clip": 1.0535965, + "balance_loss_mlp": 1.02010357, + "epoch": 0.44393675223952384, + "flos": 15523213996800.0, + "grad_norm": 1.9382079746543188, + "language_loss": 0.88369977, + "learning_rate": 2.4563174802145445e-06, + "loss": 0.90612757, + "num_input_tokens_seen": 79469865, + "step": 3692, + "time_per_iteration": 2.6969985961914062 + }, + { + "auxiliary_loss_clip": 0.01108431, + "auxiliary_loss_mlp": 0.01005539, + "balance_loss_clip": 1.02323294, + "balance_loss_mlp": 1.00408483, + "epoch": 0.44405699513016295, + "flos": 64574893779840.0, + "grad_norm": 0.6372330479888297, + "language_loss": 0.48610711, + "learning_rate": 2.455559020172712e-06, + "loss": 0.50724685, + "num_input_tokens_seen": 79537220, + "step": 3693, + "time_per_iteration": 3.3146986961364746 + }, + { + "auxiliary_loss_clip": 0.01232871, + "auxiliary_loss_mlp": 0.0103789, + "balance_loss_clip": 1.05821609, + "balance_loss_mlp": 1.02907419, + "epoch": 0.444177238020802, + "flos": 23987394552960.0, + "grad_norm": 2.3960069687590986, + "language_loss": 0.89399487, + "learning_rate": 2.4548004910229385e-06, + "loss": 0.91670251, + "num_input_tokens_seen": 79554795, + "step": 3694, + "time_per_iteration": 2.817686080932617 + }, + { + "auxiliary_loss_clip": 0.01199325, + "auxiliary_loss_mlp": 0.01379392, + "balance_loss_clip": 1.05390382, + "balance_loss_mlp": 1.00022507, + "epoch": 0.4442974809114411, + "flos": 22563069575040.0, + "grad_norm": 1.9200129402340305, + "language_loss": 0.87047887, + "learning_rate": 2.4540418928802913e-06, + "loss": 0.8962661, + "num_input_tokens_seen": 79573530, + "step": 3695, + "time_per_iteration": 2.6836631298065186 + }, + { + "auxiliary_loss_clip": 0.01205004, + "auxiliary_loss_mlp": 0.01033052, + "balance_loss_clip": 1.05191898, + "balance_loss_mlp": 1.02410579, + "epoch": 0.4444177238020802, + "flos": 17675699483520.0, + "grad_norm": 2.1116014897990483, + "language_loss": 0.65883797, + "learning_rate": 2.4532832258598506e-06, + "loss": 0.68121856, + "num_input_tokens_seen": 79591360, + "step": 3696, + "time_per_iteration": 2.6822662353515625 + }, + { + "auxiliary_loss_clip": 0.01186155, + "auxiliary_loss_mlp": 0.01029943, + "balance_loss_clip": 1.05470896, + "balance_loss_mlp": 1.02150869, + "epoch": 0.4445379666927193, + "flos": 28621594609920.0, + "grad_norm": 1.736066058196946, + "language_loss": 0.81089389, + "learning_rate": 2.4525244900767047e-06, + "loss": 0.8330549, + "num_input_tokens_seen": 79612175, + "step": 3697, + "time_per_iteration": 2.6124510765075684 + }, + { + "auxiliary_loss_clip": 0.01109827, + "auxiliary_loss_mlp": 0.01000451, + "balance_loss_clip": 1.04049706, + "balance_loss_mlp": 0.99921721, + "epoch": 0.4446582095833584, + "flos": 70487370115200.0, + "grad_norm": 0.7784131361368414, + "language_loss": 0.60469878, + "learning_rate": 2.4517656856459536e-06, + "loss": 0.62580156, + "num_input_tokens_seen": 79678020, + "step": 3698, + "time_per_iteration": 3.326637029647827 + }, + { + "auxiliary_loss_clip": 0.0119804, + "auxiliary_loss_mlp": 0.01027464, + "balance_loss_clip": 1.0537219, + "balance_loss_mlp": 1.01885438, + "epoch": 0.4447784524739975, + "flos": 26505199313280.0, + "grad_norm": 1.5808156796791488, + "language_loss": 0.67965746, + "learning_rate": 2.4510068126827073e-06, + "loss": 0.70191246, + "num_input_tokens_seen": 79699020, + "step": 3699, + "time_per_iteration": 2.7146644592285156 + }, + { + "auxiliary_loss_clip": 0.01201196, + "auxiliary_loss_mlp": 0.01029132, + "balance_loss_clip": 1.05081201, + "balance_loss_mlp": 1.02093041, + "epoch": 0.44489869536463655, + "flos": 11656209553920.0, + "grad_norm": 2.380030049310432, + "language_loss": 0.81913733, + "learning_rate": 2.450247871302086e-06, + "loss": 0.84144068, + "num_input_tokens_seen": 79716795, + "step": 3700, + "time_per_iteration": 2.6823365688323975 + }, + { + "auxiliary_loss_clip": 0.01199836, + "auxiliary_loss_mlp": 0.01030328, + "balance_loss_clip": 1.05370188, + "balance_loss_mlp": 1.02189994, + "epoch": 0.44501893825527566, + "flos": 20448469958400.0, + "grad_norm": 2.33481180909058, + "language_loss": 0.83410037, + "learning_rate": 2.44948886161922e-06, + "loss": 0.85640204, + "num_input_tokens_seen": 79735810, + "step": 3701, + "time_per_iteration": 2.646738052368164 + }, + { + "auxiliary_loss_clip": 0.01198804, + "auxiliary_loss_mlp": 0.01032826, + "balance_loss_clip": 1.05407286, + "balance_loss_mlp": 1.02424943, + "epoch": 0.4451391811459148, + "flos": 18261079430400.0, + "grad_norm": 1.7174848439895112, + "language_loss": 0.84853679, + "learning_rate": 2.4487297837492524e-06, + "loss": 0.87085313, + "num_input_tokens_seen": 79754975, + "step": 3702, + "time_per_iteration": 3.564713478088379 + }, + { + "auxiliary_loss_clip": 0.01212186, + "auxiliary_loss_mlp": 0.01030182, + "balance_loss_clip": 1.05006707, + "balance_loss_mlp": 1.021456, + "epoch": 0.44525942403655383, + "flos": 16910155895040.0, + "grad_norm": 1.852561256797135, + "language_loss": 0.62313306, + "learning_rate": 2.4479706378073323e-06, + "loss": 0.64555681, + "num_input_tokens_seen": 79773515, + "step": 3703, + "time_per_iteration": 4.504130601882935 + }, + { + "auxiliary_loss_clip": 0.01196553, + "auxiliary_loss_mlp": 0.01027474, + "balance_loss_clip": 1.04587984, + "balance_loss_mlp": 1.01932621, + "epoch": 0.44537966692719294, + "flos": 23258838994560.0, + "grad_norm": 1.637692857160993, + "language_loss": 0.84038812, + "learning_rate": 2.447211423908623e-06, + "loss": 0.86262846, + "num_input_tokens_seen": 79793560, + "step": 3704, + "time_per_iteration": 2.70210862159729 + }, + { + "auxiliary_loss_clip": 0.0119719, + "auxiliary_loss_mlp": 0.010303, + "balance_loss_clip": 1.05218208, + "balance_loss_mlp": 1.02256906, + "epoch": 0.445499909817832, + "flos": 21724160457600.0, + "grad_norm": 2.5436119346848036, + "language_loss": 0.74957949, + "learning_rate": 2.4464521421682966e-06, + "loss": 0.7718544, + "num_input_tokens_seen": 79811150, + "step": 3705, + "time_per_iteration": 2.619410753250122 + }, + { + "auxiliary_loss_clip": 0.0118802, + "auxiliary_loss_mlp": 0.01028239, + "balance_loss_clip": 1.05126297, + "balance_loss_mlp": 1.02058625, + "epoch": 0.4456201527084711, + "flos": 23987969170560.0, + "grad_norm": 1.3276122478052141, + "language_loss": 0.87671626, + "learning_rate": 2.4456927927015345e-06, + "loss": 0.89887881, + "num_input_tokens_seen": 79832190, + "step": 3706, + "time_per_iteration": 2.684204339981079 + }, + { + "auxiliary_loss_clip": 0.01212706, + "auxiliary_loss_mlp": 0.01028458, + "balance_loss_clip": 1.05601931, + "balance_loss_mlp": 1.01930833, + "epoch": 0.4457403955991102, + "flos": 18807065136000.0, + "grad_norm": 2.4716242819115135, + "language_loss": 0.76172233, + "learning_rate": 2.4449333756235307e-06, + "loss": 0.78413403, + "num_input_tokens_seen": 79848905, + "step": 3707, + "time_per_iteration": 2.6597952842712402 + }, + { + "auxiliary_loss_clip": 0.01202332, + "auxiliary_loss_mlp": 0.01030424, + "balance_loss_clip": 1.05576313, + "balance_loss_mlp": 1.02232695, + "epoch": 0.4458606384897493, + "flos": 19207756327680.0, + "grad_norm": 2.675219495835348, + "language_loss": 0.79242158, + "learning_rate": 2.4441738910494876e-06, + "loss": 0.81474912, + "num_input_tokens_seen": 79863640, + "step": 3708, + "time_per_iteration": 2.6917824745178223 + }, + { + "auxiliary_loss_clip": 0.01209651, + "auxiliary_loss_mlp": 0.01032553, + "balance_loss_clip": 1.0506624, + "balance_loss_mlp": 1.02298653, + "epoch": 0.4459808813803884, + "flos": 21361283308800.0, + "grad_norm": 6.949435009485993, + "language_loss": 0.82160974, + "learning_rate": 2.4434143390946176e-06, + "loss": 0.84403181, + "num_input_tokens_seen": 79882450, + "step": 3709, + "time_per_iteration": 3.657869338989258 + }, + { + "auxiliary_loss_clip": 0.01210197, + "auxiliary_loss_mlp": 0.01031018, + "balance_loss_clip": 1.05010545, + "balance_loss_mlp": 1.02291203, + "epoch": 0.4461011242710275, + "flos": 23288967527040.0, + "grad_norm": 2.485359101741412, + "language_loss": 0.8551752, + "learning_rate": 2.4426547198741457e-06, + "loss": 0.87758732, + "num_input_tokens_seen": 79900655, + "step": 3710, + "time_per_iteration": 2.7274398803710938 + }, + { + "auxiliary_loss_clip": 0.01220415, + "auxiliary_loss_mlp": 0.0102844, + "balance_loss_clip": 1.05131185, + "balance_loss_mlp": 1.01989877, + "epoch": 0.44622136716166655, + "flos": 20193001453440.0, + "grad_norm": 2.0018131292753414, + "language_loss": 0.74378526, + "learning_rate": 2.441895033503305e-06, + "loss": 0.76627386, + "num_input_tokens_seen": 79918575, + "step": 3711, + "time_per_iteration": 2.879054307937622 + }, + { + "auxiliary_loss_clip": 0.0119457, + "auxiliary_loss_mlp": 0.01032291, + "balance_loss_clip": 1.051319, + "balance_loss_mlp": 1.02310026, + "epoch": 0.44634161005230566, + "flos": 21283033530240.0, + "grad_norm": 1.620987725187861, + "language_loss": 0.81971025, + "learning_rate": 2.4411352800973375e-06, + "loss": 0.84197891, + "num_input_tokens_seen": 79937010, + "step": 3712, + "time_per_iteration": 2.6446096897125244 + }, + { + "auxiliary_loss_clip": 0.01207162, + "auxiliary_loss_mlp": 0.01029824, + "balance_loss_clip": 1.04786897, + "balance_loss_mlp": 1.02141953, + "epoch": 0.44646185294294477, + "flos": 22929358515840.0, + "grad_norm": 2.566996676805878, + "language_loss": 0.75353777, + "learning_rate": 2.4403754597715005e-06, + "loss": 0.77590764, + "num_input_tokens_seen": 79956455, + "step": 3713, + "time_per_iteration": 2.72251296043396 + }, + { + "auxiliary_loss_clip": 0.01203493, + "auxiliary_loss_mlp": 0.01029843, + "balance_loss_clip": 1.04815745, + "balance_loss_mlp": 1.02108693, + "epoch": 0.4465820958335838, + "flos": 22637692080000.0, + "grad_norm": 5.057211583878384, + "language_loss": 0.93169451, + "learning_rate": 2.4396155726410553e-06, + "loss": 0.95402789, + "num_input_tokens_seen": 79975065, + "step": 3714, + "time_per_iteration": 2.688377857208252 + }, + { + "auxiliary_loss_clip": 0.01200684, + "auxiliary_loss_mlp": 0.01030638, + "balance_loss_clip": 1.05336785, + "balance_loss_mlp": 1.02178669, + "epoch": 0.44670233872422294, + "flos": 22672525294080.0, + "grad_norm": 4.273930767696802, + "language_loss": 0.90705526, + "learning_rate": 2.438855618821278e-06, + "loss": 0.9293685, + "num_input_tokens_seen": 79990865, + "step": 3715, + "time_per_iteration": 2.620181083679199 + }, + { + "auxiliary_loss_clip": 0.01185425, + "auxiliary_loss_mlp": 0.01028917, + "balance_loss_clip": 1.04868817, + "balance_loss_mlp": 1.0208106, + "epoch": 0.44682258161486205, + "flos": 23582178247680.0, + "grad_norm": 1.6251154255080225, + "language_loss": 0.67636764, + "learning_rate": 2.4380955984274517e-06, + "loss": 0.698511, + "num_input_tokens_seen": 80009520, + "step": 3716, + "time_per_iteration": 2.726823568344116 + }, + { + "auxiliary_loss_clip": 0.01195338, + "auxiliary_loss_mlp": 0.01027281, + "balance_loss_clip": 1.05376852, + "balance_loss_mlp": 1.01872754, + "epoch": 0.4469428245055011, + "flos": 26501356558080.0, + "grad_norm": 1.8700097920062475, + "language_loss": 0.76685345, + "learning_rate": 2.4373355115748716e-06, + "loss": 0.78907967, + "num_input_tokens_seen": 80030350, + "step": 3717, + "time_per_iteration": 2.6929550170898438 + }, + { + "auxiliary_loss_clip": 0.01195493, + "auxiliary_loss_mlp": 0.01029917, + "balance_loss_clip": 1.05009091, + "balance_loss_mlp": 1.02135146, + "epoch": 0.4470630673961402, + "flos": 21504925797120.0, + "grad_norm": 3.156012840907761, + "language_loss": 0.72299147, + "learning_rate": 2.436575358378842e-06, + "loss": 0.74524552, + "num_input_tokens_seen": 80049840, + "step": 3718, + "time_per_iteration": 2.69863224029541 + }, + { + "auxiliary_loss_clip": 0.01210926, + "auxiliary_loss_mlp": 0.01030683, + "balance_loss_clip": 1.05282629, + "balance_loss_mlp": 1.0221417, + "epoch": 0.44718331028677927, + "flos": 16173986653440.0, + "grad_norm": 2.8617990418262353, + "language_loss": 0.83427095, + "learning_rate": 2.4358151389546782e-06, + "loss": 0.85668701, + "num_input_tokens_seen": 80066525, + "step": 3719, + "time_per_iteration": 2.631152391433716 + }, + { + "auxiliary_loss_clip": 0.01186727, + "auxiliary_loss_mlp": 0.01031248, + "balance_loss_clip": 1.05399239, + "balance_loss_mlp": 1.02305818, + "epoch": 0.4473035531774184, + "flos": 19681238430720.0, + "grad_norm": 2.776527888205176, + "language_loss": 0.76025409, + "learning_rate": 2.4350548534177035e-06, + "loss": 0.78243387, + "num_input_tokens_seen": 80083355, + "step": 3720, + "time_per_iteration": 2.7050766944885254 + }, + { + "auxiliary_loss_clip": 0.01210225, + "auxiliary_loss_mlp": 0.0102687, + "balance_loss_clip": 1.04845619, + "balance_loss_mlp": 1.01845419, + "epoch": 0.4474237960680575, + "flos": 41427590515200.0, + "grad_norm": 1.561574646110216, + "language_loss": 0.66840279, + "learning_rate": 2.434294501883254e-06, + "loss": 0.69077373, + "num_input_tokens_seen": 80106450, + "step": 3721, + "time_per_iteration": 2.87019681930542 + }, + { + "auxiliary_loss_clip": 0.0119106, + "auxiliary_loss_mlp": 0.01030126, + "balance_loss_clip": 1.04727316, + "balance_loss_mlp": 1.02163219, + "epoch": 0.44754403895869654, + "flos": 22891328991360.0, + "grad_norm": 1.7208822032821733, + "language_loss": 0.65780884, + "learning_rate": 2.433534084466674e-06, + "loss": 0.68002069, + "num_input_tokens_seen": 80125670, + "step": 3722, + "time_per_iteration": 2.7296416759490967 + }, + { + "auxiliary_loss_clip": 0.0118287, + "auxiliary_loss_mlp": 0.01030497, + "balance_loss_clip": 1.05212271, + "balance_loss_mlp": 1.02242672, + "epoch": 0.44766428184933565, + "flos": 25630271832960.0, + "grad_norm": 1.4585599940711413, + "language_loss": 0.70910919, + "learning_rate": 2.4327736012833178e-06, + "loss": 0.7312429, + "num_input_tokens_seen": 80147390, + "step": 3723, + "time_per_iteration": 2.6065430641174316 + }, + { + "auxiliary_loss_clip": 0.011944, + "auxiliary_loss_mlp": 0.01025201, + "balance_loss_clip": 1.0531075, + "balance_loss_mlp": 1.01633167, + "epoch": 0.44778452473997477, + "flos": 20448972748800.0, + "grad_norm": 1.9667459951807704, + "language_loss": 0.76299095, + "learning_rate": 2.4320130524485506e-06, + "loss": 0.78518695, + "num_input_tokens_seen": 80166185, + "step": 3724, + "time_per_iteration": 2.675131320953369 + }, + { + "auxiliary_loss_clip": 0.01197726, + "auxiliary_loss_mlp": 0.01028516, + "balance_loss_clip": 1.05369043, + "balance_loss_mlp": 1.02074409, + "epoch": 0.4479047676306138, + "flos": 21975462984960.0, + "grad_norm": 1.423949724702474, + "language_loss": 0.79702741, + "learning_rate": 2.431252438077746e-06, + "loss": 0.8192898, + "num_input_tokens_seen": 80185685, + "step": 3725, + "time_per_iteration": 2.666611671447754 + }, + { + "auxiliary_loss_clip": 0.01198368, + "auxiliary_loss_mlp": 0.01379473, + "balance_loss_clip": 1.05275834, + "balance_loss_mlp": 1.00028729, + "epoch": 0.44802501052125293, + "flos": 21467219495040.0, + "grad_norm": 2.643932304124272, + "language_loss": 0.77572936, + "learning_rate": 2.4304917582862906e-06, + "loss": 0.80150777, + "num_input_tokens_seen": 80204865, + "step": 3726, + "time_per_iteration": 2.6545753479003906 + }, + { + "auxiliary_loss_clip": 0.01189408, + "auxiliary_loss_mlp": 0.01027684, + "balance_loss_clip": 1.0559051, + "balance_loss_mlp": 1.0193038, + "epoch": 0.44814525341189204, + "flos": 22126970551680.0, + "grad_norm": 1.960594925481167, + "language_loss": 0.87862539, + "learning_rate": 2.4297310131895774e-06, + "loss": 0.90079629, + "num_input_tokens_seen": 80223410, + "step": 3727, + "time_per_iteration": 2.5844247341156006 + }, + { + "auxiliary_loss_clip": 0.01197484, + "auxiliary_loss_mlp": 0.01029458, + "balance_loss_clip": 1.05485725, + "balance_loss_mlp": 1.02095866, + "epoch": 0.4482654963025311, + "flos": 16653933204480.0, + "grad_norm": 2.298805181652212, + "language_loss": 0.7447378, + "learning_rate": 2.4289702029030113e-06, + "loss": 0.76700723, + "num_input_tokens_seen": 80240880, + "step": 3728, + "time_per_iteration": 3.538877248764038 + }, + { + "auxiliary_loss_clip": 0.0119424, + "auxiliary_loss_mlp": 0.01030989, + "balance_loss_clip": 1.05389714, + "balance_loss_mlp": 1.0221262, + "epoch": 0.4483857391931702, + "flos": 18841251905280.0, + "grad_norm": 1.8199582530779987, + "language_loss": 0.82776636, + "learning_rate": 2.4282093275420057e-06, + "loss": 0.85001868, + "num_input_tokens_seen": 80259910, + "step": 3729, + "time_per_iteration": 3.4563236236572266 + }, + { + "auxiliary_loss_clip": 0.01199252, + "auxiliary_loss_mlp": 0.01027041, + "balance_loss_clip": 1.05309927, + "balance_loss_mlp": 1.01919127, + "epoch": 0.4485059820838093, + "flos": 20372590477440.0, + "grad_norm": 2.068723254269659, + "language_loss": 0.70394385, + "learning_rate": 2.4274483872219863e-06, + "loss": 0.72620678, + "num_input_tokens_seen": 80277270, + "step": 3730, + "time_per_iteration": 2.498856782913208 + }, + { + "auxiliary_loss_clip": 0.01192356, + "auxiliary_loss_mlp": 0.0102596, + "balance_loss_clip": 1.05269504, + "balance_loss_mlp": 1.01793694, + "epoch": 0.4486262249744484, + "flos": 20047742853120.0, + "grad_norm": 2.1027318999046765, + "language_loss": 0.93652761, + "learning_rate": 2.426687382058386e-06, + "loss": 0.95871085, + "num_input_tokens_seen": 80295550, + "step": 3731, + "time_per_iteration": 2.5569732189178467 + }, + { + "auxiliary_loss_clip": 0.01103707, + "auxiliary_loss_mlp": 0.01001415, + "balance_loss_clip": 1.03534842, + "balance_loss_mlp": 1.00009823, + "epoch": 0.4487464678650875, + "flos": 64595684776320.0, + "grad_norm": 0.8584252673701235, + "language_loss": 0.59774756, + "learning_rate": 2.425926312166649e-06, + "loss": 0.61879873, + "num_input_tokens_seen": 80348425, + "step": 3732, + "time_per_iteration": 3.0471301078796387 + }, + { + "auxiliary_loss_clip": 0.01206819, + "auxiliary_loss_mlp": 0.01031616, + "balance_loss_clip": 1.0502249, + "balance_loss_mlp": 1.02339041, + "epoch": 0.4488667107557266, + "flos": 20769798049920.0, + "grad_norm": 2.8795115458890805, + "language_loss": 0.73589993, + "learning_rate": 2.42516517766223e-06, + "loss": 0.75828433, + "num_input_tokens_seen": 80366505, + "step": 3733, + "time_per_iteration": 2.7534396648406982 + }, + { + "auxiliary_loss_clip": 0.01188987, + "auxiliary_loss_mlp": 0.01028151, + "balance_loss_clip": 1.05709755, + "balance_loss_mlp": 1.02027738, + "epoch": 0.44898695364636565, + "flos": 23951735326080.0, + "grad_norm": 1.7088837418384584, + "language_loss": 0.68204755, + "learning_rate": 2.4244039786605907e-06, + "loss": 0.70421892, + "num_input_tokens_seen": 80387510, + "step": 3734, + "time_per_iteration": 2.6865015029907227 + }, + { + "auxiliary_loss_clip": 0.01223662, + "auxiliary_loss_mlp": 0.01036042, + "balance_loss_clip": 1.047575, + "balance_loss_mlp": 1.02733994, + "epoch": 0.44910719653700476, + "flos": 18624351628800.0, + "grad_norm": 2.281741187737899, + "language_loss": 0.82391882, + "learning_rate": 2.4236427152772055e-06, + "loss": 0.84651583, + "num_input_tokens_seen": 80405915, + "step": 3735, + "time_per_iteration": 3.650080919265747 + }, + { + "auxiliary_loss_clip": 0.01106377, + "auxiliary_loss_mlp": 0.01002793, + "balance_loss_clip": 1.02175784, + "balance_loss_mlp": 1.00150585, + "epoch": 0.4492274394276438, + "flos": 57033435749760.0, + "grad_norm": 0.8208288719140442, + "language_loss": 0.57380342, + "learning_rate": 2.422881387627557e-06, + "loss": 0.59489512, + "num_input_tokens_seen": 80458365, + "step": 3736, + "time_per_iteration": 3.0186078548431396 + }, + { + "auxiliary_loss_clip": 0.01204698, + "auxiliary_loss_mlp": 0.01028127, + "balance_loss_clip": 1.05137062, + "balance_loss_mlp": 1.01974678, + "epoch": 0.4493476823182829, + "flos": 23254888498560.0, + "grad_norm": 1.5992484938385407, + "language_loss": 0.77254182, + "learning_rate": 2.422119995827139e-06, + "loss": 0.79487002, + "num_input_tokens_seen": 80478490, + "step": 3737, + "time_per_iteration": 2.6563456058502197 + }, + { + "auxiliary_loss_clip": 0.01201813, + "auxiliary_loss_mlp": 0.01025558, + "balance_loss_clip": 1.05706406, + "balance_loss_mlp": 1.01708198, + "epoch": 0.44946792520892204, + "flos": 15815131827840.0, + "grad_norm": 2.8015282269243778, + "language_loss": 0.73922169, + "learning_rate": 2.4213585399914528e-06, + "loss": 0.76149541, + "num_input_tokens_seen": 80495695, + "step": 3738, + "time_per_iteration": 2.5848000049591064 + }, + { + "auxiliary_loss_clip": 0.0119584, + "auxiliary_loss_mlp": 0.01028381, + "balance_loss_clip": 1.05402434, + "balance_loss_mlp": 1.02027512, + "epoch": 0.4495881680995611, + "flos": 19610063631360.0, + "grad_norm": 1.717336217688089, + "language_loss": 0.8539995, + "learning_rate": 2.4205970202360113e-06, + "loss": 0.87624168, + "num_input_tokens_seen": 80515260, + "step": 3739, + "time_per_iteration": 2.6338984966278076 + }, + { + "auxiliary_loss_clip": 0.01214138, + "auxiliary_loss_mlp": 0.01031283, + "balance_loss_clip": 1.04910827, + "balance_loss_mlp": 1.02273607, + "epoch": 0.4497084109902002, + "flos": 26031465815040.0, + "grad_norm": 4.029173269838788, + "language_loss": 0.78385162, + "learning_rate": 2.4198354366763354e-06, + "loss": 0.80630583, + "num_input_tokens_seen": 80533900, + "step": 3740, + "time_per_iteration": 2.8050689697265625 + }, + { + "auxiliary_loss_clip": 0.01207121, + "auxiliary_loss_mlp": 0.01029074, + "balance_loss_clip": 1.0529722, + "balance_loss_mlp": 1.02086651, + "epoch": 0.4498286538808393, + "flos": 14793688771200.0, + "grad_norm": 2.3237568040435335, + "language_loss": 0.78639197, + "learning_rate": 2.4190737894279587e-06, + "loss": 0.80875397, + "num_input_tokens_seen": 80551270, + "step": 3741, + "time_per_iteration": 2.6748814582824707 + }, + { + "auxiliary_loss_clip": 0.01196426, + "auxiliary_loss_mlp": 0.01032787, + "balance_loss_clip": 1.04364467, + "balance_loss_mlp": 1.02441823, + "epoch": 0.44994889677147837, + "flos": 15450171690240.0, + "grad_norm": 3.180530195605207, + "language_loss": 0.8013494, + "learning_rate": 2.4183120786064203e-06, + "loss": 0.82364154, + "num_input_tokens_seen": 80568145, + "step": 3742, + "time_per_iteration": 2.856226921081543 + }, + { + "auxiliary_loss_clip": 0.01197792, + "auxiliary_loss_mlp": 0.01379463, + "balance_loss_clip": 1.05493093, + "balance_loss_mlp": 1.00029564, + "epoch": 0.4500691396621175, + "flos": 21798316085760.0, + "grad_norm": 2.1525400514819455, + "language_loss": 0.85224766, + "learning_rate": 2.417550304327273e-06, + "loss": 0.87802017, + "num_input_tokens_seen": 80586185, + "step": 3743, + "time_per_iteration": 2.681993007659912 + }, + { + "auxiliary_loss_clip": 0.01190508, + "auxiliary_loss_mlp": 0.01028487, + "balance_loss_clip": 1.05707681, + "balance_loss_mlp": 1.01904559, + "epoch": 0.4501893825527566, + "flos": 32382016421760.0, + "grad_norm": 1.5915577621228274, + "language_loss": 0.75780237, + "learning_rate": 2.4167884667060763e-06, + "loss": 0.77999234, + "num_input_tokens_seen": 80608895, + "step": 3744, + "time_per_iteration": 2.699380397796631 + }, + { + "auxiliary_loss_clip": 0.01202868, + "auxiliary_loss_mlp": 0.01031261, + "balance_loss_clip": 1.05043495, + "balance_loss_mlp": 1.02179623, + "epoch": 0.45030962544339564, + "flos": 16544944362240.0, + "grad_norm": 2.2125333818583397, + "language_loss": 0.8761003, + "learning_rate": 2.4160265658584e-06, + "loss": 0.89844155, + "num_input_tokens_seen": 80623785, + "step": 3745, + "time_per_iteration": 2.6892099380493164 + }, + { + "auxiliary_loss_clip": 0.01200009, + "auxiliary_loss_mlp": 0.0102803, + "balance_loss_clip": 1.05476797, + "balance_loss_mlp": 1.01979268, + "epoch": 0.45042986833403476, + "flos": 19573039687680.0, + "grad_norm": 1.9573862219553644, + "language_loss": 0.68257868, + "learning_rate": 2.4152646018998253e-06, + "loss": 0.70485902, + "num_input_tokens_seen": 80642735, + "step": 3746, + "time_per_iteration": 2.772172451019287 + }, + { + "auxiliary_loss_clip": 0.01191838, + "auxiliary_loss_mlp": 0.01026499, + "balance_loss_clip": 1.05318499, + "balance_loss_mlp": 1.01830947, + "epoch": 0.45055011122467387, + "flos": 23112467072640.0, + "grad_norm": 1.799229184318272, + "language_loss": 0.72052729, + "learning_rate": 2.4145025749459403e-06, + "loss": 0.74271071, + "num_input_tokens_seen": 80663760, + "step": 3747, + "time_per_iteration": 2.746121883392334 + }, + { + "auxiliary_loss_clip": 0.01238589, + "auxiliary_loss_mlp": 0.01025298, + "balance_loss_clip": 1.04639041, + "balance_loss_mlp": 1.01658392, + "epoch": 0.4506703541153129, + "flos": 19934623946880.0, + "grad_norm": 1.872013875691128, + "language_loss": 0.70097822, + "learning_rate": 2.413740485112344e-06, + "loss": 0.72361708, + "num_input_tokens_seen": 80682100, + "step": 3748, + "time_per_iteration": 2.811337471008301 + }, + { + "auxiliary_loss_clip": 0.01202792, + "auxiliary_loss_mlp": 0.01026092, + "balance_loss_clip": 1.05505872, + "balance_loss_mlp": 1.01803327, + "epoch": 0.45079059700595203, + "flos": 19499530504320.0, + "grad_norm": 1.5802254310453625, + "language_loss": 0.82384229, + "learning_rate": 2.412978332514646e-06, + "loss": 0.84613115, + "num_input_tokens_seen": 80700880, + "step": 3749, + "time_per_iteration": 2.6944291591644287 + }, + { + "auxiliary_loss_clip": 0.01208064, + "auxiliary_loss_mlp": 0.01024465, + "balance_loss_clip": 1.05416512, + "balance_loss_mlp": 1.01599503, + "epoch": 0.4509108398965911, + "flos": 27636313570560.0, + "grad_norm": 1.946752076134735, + "language_loss": 0.723046, + "learning_rate": 2.4122161172684623e-06, + "loss": 0.74537128, + "num_input_tokens_seen": 80721675, + "step": 3750, + "time_per_iteration": 2.707136869430542 + }, + { + "auxiliary_loss_clip": 0.01208564, + "auxiliary_loss_mlp": 0.01028936, + "balance_loss_clip": 1.05409932, + "balance_loss_mlp": 1.02009618, + "epoch": 0.4510310827872302, + "flos": 20995712640000.0, + "grad_norm": 2.2385236476853807, + "language_loss": 0.84550881, + "learning_rate": 2.4114538394894216e-06, + "loss": 0.8678838, + "num_input_tokens_seen": 80739315, + "step": 3751, + "time_per_iteration": 2.666091203689575 + }, + { + "auxiliary_loss_clip": 0.01197965, + "auxiliary_loss_mlp": 0.01030324, + "balance_loss_clip": 1.04672313, + "balance_loss_mlp": 1.02197337, + "epoch": 0.4511513256778693, + "flos": 16216684945920.0, + "grad_norm": 2.44654534281039, + "language_loss": 0.83087969, + "learning_rate": 2.410691499293161e-06, + "loss": 0.85316253, + "num_input_tokens_seen": 80757470, + "step": 3752, + "time_per_iteration": 2.681516647338867 + }, + { + "auxiliary_loss_clip": 0.01193764, + "auxiliary_loss_mlp": 0.0102631, + "balance_loss_clip": 1.05172205, + "balance_loss_mlp": 1.01851356, + "epoch": 0.45127156856850836, + "flos": 25186702780800.0, + "grad_norm": 1.586392752283445, + "language_loss": 0.74386185, + "learning_rate": 2.409929096795326e-06, + "loss": 0.76606262, + "num_input_tokens_seen": 80777840, + "step": 3753, + "time_per_iteration": 2.7063841819763184 + }, + { + "auxiliary_loss_clip": 0.01195518, + "auxiliary_loss_mlp": 0.01027114, + "balance_loss_clip": 1.05158138, + "balance_loss_mlp": 1.01831055, + "epoch": 0.4513918114591475, + "flos": 20412523422720.0, + "grad_norm": 1.9201063046867146, + "language_loss": 0.79101706, + "learning_rate": 2.409166632111573e-06, + "loss": 0.81324339, + "num_input_tokens_seen": 80795975, + "step": 3754, + "time_per_iteration": 2.7380311489105225 + }, + { + "auxiliary_loss_clip": 0.01204234, + "auxiliary_loss_mlp": 0.01028629, + "balance_loss_clip": 1.05392396, + "balance_loss_mlp": 1.01990294, + "epoch": 0.4515120543497866, + "flos": 26648482665600.0, + "grad_norm": 2.7402587974683112, + "language_loss": 0.8044951, + "learning_rate": 2.4084041053575674e-06, + "loss": 0.82682371, + "num_input_tokens_seen": 80815395, + "step": 3755, + "time_per_iteration": 4.595381736755371 + }, + { + "auxiliary_loss_clip": 0.01210543, + "auxiliary_loss_mlp": 0.01030443, + "balance_loss_clip": 1.0543642, + "balance_loss_mlp": 1.02223527, + "epoch": 0.45163229724042564, + "flos": 20595093275520.0, + "grad_norm": 2.4542752333162903, + "language_loss": 0.72089237, + "learning_rate": 2.4076415166489834e-06, + "loss": 0.74330223, + "num_input_tokens_seen": 80834805, + "step": 3756, + "time_per_iteration": 2.6670942306518555 + }, + { + "auxiliary_loss_clip": 0.01229242, + "auxiliary_loss_mlp": 0.01028211, + "balance_loss_clip": 1.04994071, + "balance_loss_mlp": 1.01992583, + "epoch": 0.45175254013106475, + "flos": 21689004021120.0, + "grad_norm": 1.5492965262862866, + "language_loss": 0.79070437, + "learning_rate": 2.406878866101506e-06, + "loss": 0.81327885, + "num_input_tokens_seen": 80853770, + "step": 3757, + "time_per_iteration": 2.7972614765167236 + }, + { + "auxiliary_loss_clip": 0.01187942, + "auxiliary_loss_mlp": 0.01031355, + "balance_loss_clip": 1.05585027, + "balance_loss_mlp": 1.02323663, + "epoch": 0.45187278302170386, + "flos": 18878850466560.0, + "grad_norm": 1.9212872169468453, + "language_loss": 0.78180444, + "learning_rate": 2.4061161538308273e-06, + "loss": 0.8039974, + "num_input_tokens_seen": 80870615, + "step": 3758, + "time_per_iteration": 2.623138189315796 + }, + { + "auxiliary_loss_clip": 0.01194719, + "auxiliary_loss_mlp": 0.01028668, + "balance_loss_clip": 1.05198503, + "balance_loss_mlp": 1.01979852, + "epoch": 0.4519930259123429, + "flos": 18582479349120.0, + "grad_norm": 1.9547224988294605, + "language_loss": 0.8895992, + "learning_rate": 2.4053533799526523e-06, + "loss": 0.91183305, + "num_input_tokens_seen": 80886335, + "step": 3759, + "time_per_iteration": 2.6533939838409424 + }, + { + "auxiliary_loss_clip": 0.01196435, + "auxiliary_loss_mlp": 0.01024878, + "balance_loss_clip": 1.04982877, + "balance_loss_mlp": 1.01681376, + "epoch": 0.452113268802982, + "flos": 25192377129600.0, + "grad_norm": 1.632012112591302, + "language_loss": 0.86336803, + "learning_rate": 2.404590544582691e-06, + "loss": 0.8855812, + "num_input_tokens_seen": 80904570, + "step": 3760, + "time_per_iteration": 2.7153921127319336 + }, + { + "auxiliary_loss_clip": 0.01219291, + "auxiliary_loss_mlp": 0.01034461, + "balance_loss_clip": 1.04490018, + "balance_loss_mlp": 1.02597332, + "epoch": 0.45223351169362114, + "flos": 39378922312320.0, + "grad_norm": 1.7791637298314593, + "language_loss": 0.80996591, + "learning_rate": 2.403827647836666e-06, + "loss": 0.83250344, + "num_input_tokens_seen": 80925125, + "step": 3761, + "time_per_iteration": 3.828139066696167 + }, + { + "auxiliary_loss_clip": 0.01187209, + "auxiliary_loss_mlp": 0.01028977, + "balance_loss_clip": 1.05374074, + "balance_loss_mlp": 1.02064466, + "epoch": 0.4523537545842602, + "flos": 21582169994880.0, + "grad_norm": 1.8774990827853522, + "language_loss": 0.69278812, + "learning_rate": 2.4030646898303075e-06, + "loss": 0.71494997, + "num_input_tokens_seen": 80946615, + "step": 3762, + "time_per_iteration": 2.6818454265594482 + }, + { + "auxiliary_loss_clip": 0.01205737, + "auxiliary_loss_mlp": 0.01025221, + "balance_loss_clip": 1.04927278, + "balance_loss_mlp": 1.0166018, + "epoch": 0.4524739974748993, + "flos": 28439527547520.0, + "grad_norm": 2.0810120464566215, + "language_loss": 0.82121134, + "learning_rate": 2.4023016706793566e-06, + "loss": 0.84352088, + "num_input_tokens_seen": 80966410, + "step": 3763, + "time_per_iteration": 2.740304470062256 + }, + { + "auxiliary_loss_clip": 0.01122265, + "auxiliary_loss_mlp": 0.01003054, + "balance_loss_clip": 1.02022386, + "balance_loss_mlp": 1.00177813, + "epoch": 0.4525942403655384, + "flos": 61556492148480.0, + "grad_norm": 0.7932476344604141, + "language_loss": 0.56881869, + "learning_rate": 2.401538590499561e-06, + "loss": 0.59007186, + "num_input_tokens_seen": 81026865, + "step": 3764, + "time_per_iteration": 3.3312273025512695 + }, + { + "auxiliary_loss_clip": 0.01197087, + "auxiliary_loss_mlp": 0.01379623, + "balance_loss_clip": 1.05260754, + "balance_loss_mlp": 1.00024652, + "epoch": 0.45271448325617747, + "flos": 27529838680320.0, + "grad_norm": 1.7853660596930285, + "language_loss": 0.71737158, + "learning_rate": 2.400775449406682e-06, + "loss": 0.74313867, + "num_input_tokens_seen": 81050060, + "step": 3765, + "time_per_iteration": 2.711268901824951 + }, + { + "auxiliary_loss_clip": 0.01191029, + "auxiliary_loss_mlp": 0.01028666, + "balance_loss_clip": 1.04840243, + "balance_loss_mlp": 1.02114964, + "epoch": 0.4528347261468166, + "flos": 22452608275200.0, + "grad_norm": 2.06895259188629, + "language_loss": 0.73052466, + "learning_rate": 2.400012247516485e-06, + "loss": 0.75272155, + "num_input_tokens_seen": 81070625, + "step": 3766, + "time_per_iteration": 2.701770067214966 + }, + { + "auxiliary_loss_clip": 0.01215945, + "auxiliary_loss_mlp": 0.01028183, + "balance_loss_clip": 1.04959476, + "balance_loss_mlp": 1.01980591, + "epoch": 0.45295496903745563, + "flos": 21103875469440.0, + "grad_norm": 1.8518736904325548, + "language_loss": 0.90418637, + "learning_rate": 2.3992489849447484e-06, + "loss": 0.92662764, + "num_input_tokens_seen": 81089080, + "step": 3767, + "time_per_iteration": 2.7284703254699707 + }, + { + "auxiliary_loss_clip": 0.01220693, + "auxiliary_loss_mlp": 0.01034389, + "balance_loss_clip": 1.05237937, + "balance_loss_mlp": 1.02467942, + "epoch": 0.45307521192809475, + "flos": 23221168606080.0, + "grad_norm": 1.6911707113642709, + "language_loss": 0.79006457, + "learning_rate": 2.3984856618072584e-06, + "loss": 0.81261539, + "num_input_tokens_seen": 81109115, + "step": 3768, + "time_per_iteration": 2.735497236251831 + }, + { + "auxiliary_loss_clip": 0.01213561, + "auxiliary_loss_mlp": 0.01029343, + "balance_loss_clip": 1.04780698, + "balance_loss_mlp": 1.02114177, + "epoch": 0.45319545481873386, + "flos": 15560094286080.0, + "grad_norm": 1.7585099684478342, + "language_loss": 0.73789132, + "learning_rate": 2.3977222782198098e-06, + "loss": 0.76032031, + "num_input_tokens_seen": 81127750, + "step": 3769, + "time_per_iteration": 2.6750869750976562 + }, + { + "auxiliary_loss_clip": 0.01201529, + "auxiliary_loss_mlp": 0.01025854, + "balance_loss_clip": 1.04896379, + "balance_loss_mlp": 1.01732481, + "epoch": 0.4533156977093729, + "flos": 21944759834880.0, + "grad_norm": 1.7779295486900708, + "language_loss": 0.7526859, + "learning_rate": 2.3969588342982077e-06, + "loss": 0.7749598, + "num_input_tokens_seen": 81147125, + "step": 3770, + "time_per_iteration": 2.730757474899292 + }, + { + "auxiliary_loss_clip": 0.01192637, + "auxiliary_loss_mlp": 0.01028525, + "balance_loss_clip": 1.05359113, + "balance_loss_mlp": 1.02050257, + "epoch": 0.453435940600012, + "flos": 24242180699520.0, + "grad_norm": 1.8344593997313412, + "language_loss": 0.72681844, + "learning_rate": 2.396195330158267e-06, + "loss": 0.74903011, + "num_input_tokens_seen": 81167015, + "step": 3771, + "time_per_iteration": 2.687596559524536 + }, + { + "auxiliary_loss_clip": 0.01189753, + "auxiliary_loss_mlp": 0.01029603, + "balance_loss_clip": 1.05678391, + "balance_loss_mlp": 1.02065659, + "epoch": 0.45355618349065113, + "flos": 23440367352960.0, + "grad_norm": 9.058413007715439, + "language_loss": 0.79505318, + "learning_rate": 2.3954317659158094e-06, + "loss": 0.81724679, + "num_input_tokens_seen": 81187350, + "step": 3772, + "time_per_iteration": 2.6342484951019287 + }, + { + "auxiliary_loss_clip": 0.01081842, + "auxiliary_loss_mlp": 0.01003548, + "balance_loss_clip": 1.02105439, + "balance_loss_mlp": 1.00211132, + "epoch": 0.4536764263812902, + "flos": 66903161448960.0, + "grad_norm": 0.9269300004362739, + "language_loss": 0.56983393, + "learning_rate": 2.394668141686667e-06, + "loss": 0.59068787, + "num_input_tokens_seen": 81249315, + "step": 3773, + "time_per_iteration": 3.1885299682617188 + }, + { + "auxiliary_loss_clip": 0.01190206, + "auxiliary_loss_mlp": 0.01032774, + "balance_loss_clip": 1.05186701, + "balance_loss_mlp": 1.02456653, + "epoch": 0.4537966692719293, + "flos": 42739766254080.0, + "grad_norm": 2.1451590185733096, + "language_loss": 0.70172703, + "learning_rate": 2.3939044575866813e-06, + "loss": 0.72395682, + "num_input_tokens_seen": 81272065, + "step": 3774, + "time_per_iteration": 2.800583600997925 + }, + { + "auxiliary_loss_clip": 0.01198732, + "auxiliary_loss_mlp": 0.0137937, + "balance_loss_clip": 1.04961109, + "balance_loss_mlp": 1.00018334, + "epoch": 0.4539169121625684, + "flos": 35549480517120.0, + "grad_norm": 5.291183414572345, + "language_loss": 0.75668424, + "learning_rate": 2.3931407137317024e-06, + "loss": 0.78246534, + "num_input_tokens_seen": 81292220, + "step": 3775, + "time_per_iteration": 2.8589096069335938 + }, + { + "auxiliary_loss_clip": 0.01205569, + "auxiliary_loss_mlp": 0.01026708, + "balance_loss_clip": 1.04724288, + "balance_loss_mlp": 1.01839328, + "epoch": 0.45403715505320746, + "flos": 18514716341760.0, + "grad_norm": 1.8095383495037414, + "language_loss": 0.84886765, + "learning_rate": 2.3923769102375907e-06, + "loss": 0.87119043, + "num_input_tokens_seen": 81311085, + "step": 3776, + "time_per_iteration": 2.69923996925354 + }, + { + "auxiliary_loss_clip": 0.01211224, + "auxiliary_loss_mlp": 0.01030128, + "balance_loss_clip": 1.04837847, + "balance_loss_mlp": 1.02200437, + "epoch": 0.4541573979438466, + "flos": 25045825639680.0, + "grad_norm": 4.993143344465535, + "language_loss": 0.78538203, + "learning_rate": 2.391613047220213e-06, + "loss": 0.80779552, + "num_input_tokens_seen": 81330985, + "step": 3777, + "time_per_iteration": 2.7335386276245117 + }, + { + "auxiliary_loss_clip": 0.01232328, + "auxiliary_loss_mlp": 0.01027748, + "balance_loss_clip": 1.05067825, + "balance_loss_mlp": 1.01899266, + "epoch": 0.4542776408344857, + "flos": 18332397884160.0, + "grad_norm": 1.780761601580375, + "language_loss": 0.78882933, + "learning_rate": 2.390849124795447e-06, + "loss": 0.8114301, + "num_input_tokens_seen": 81346985, + "step": 3778, + "time_per_iteration": 2.6789822578430176 + }, + { + "auxiliary_loss_clip": 0.01189944, + "auxiliary_loss_mlp": 0.01028331, + "balance_loss_clip": 1.05641329, + "balance_loss_mlp": 1.02011788, + "epoch": 0.45439788372512474, + "flos": 20701173116160.0, + "grad_norm": 1.8870703837302418, + "language_loss": 0.84746337, + "learning_rate": 2.3900851430791804e-06, + "loss": 0.86964607, + "num_input_tokens_seen": 81365005, + "step": 3779, + "time_per_iteration": 2.6341006755828857 + }, + { + "auxiliary_loss_clip": 0.01191221, + "auxiliary_loss_mlp": 0.01033176, + "balance_loss_clip": 1.05715942, + "balance_loss_mlp": 1.02435434, + "epoch": 0.45451812661576385, + "flos": 22309432663680.0, + "grad_norm": 2.1191762138099097, + "language_loss": 0.84559023, + "learning_rate": 2.389321102187307e-06, + "loss": 0.86783415, + "num_input_tokens_seen": 81383785, + "step": 3780, + "time_per_iteration": 3.5856573581695557 + }, + { + "auxiliary_loss_clip": 0.0120463, + "auxiliary_loss_mlp": 0.01379965, + "balance_loss_clip": 1.05255556, + "balance_loss_mlp": 1.00022578, + "epoch": 0.4546383695064029, + "flos": 21763303303680.0, + "grad_norm": 1.8676596294120922, + "language_loss": 0.81404293, + "learning_rate": 2.3885570022357326e-06, + "loss": 0.83988887, + "num_input_tokens_seen": 81402915, + "step": 3781, + "time_per_iteration": 3.6090729236602783 + }, + { + "auxiliary_loss_clip": 0.01120176, + "auxiliary_loss_mlp": 0.01000931, + "balance_loss_clip": 1.01839757, + "balance_loss_mlp": 0.99958962, + "epoch": 0.454758612397042, + "flos": 64242755694720.0, + "grad_norm": 0.8558324775381786, + "language_loss": 0.6087631, + "learning_rate": 2.38779284334037e-06, + "loss": 0.62997413, + "num_input_tokens_seen": 81467890, + "step": 3782, + "time_per_iteration": 4.330377817153931 + }, + { + "auxiliary_loss_clip": 0.01204824, + "auxiliary_loss_mlp": 0.01027759, + "balance_loss_clip": 1.04532385, + "balance_loss_mlp": 1.01936042, + "epoch": 0.4548788552876811, + "flos": 27304175485440.0, + "grad_norm": 1.8086463568393745, + "language_loss": 0.78907025, + "learning_rate": 2.387028625617141e-06, + "loss": 0.811396, + "num_input_tokens_seen": 81487105, + "step": 3783, + "time_per_iteration": 2.796022653579712 + }, + { + "auxiliary_loss_clip": 0.01188826, + "auxiliary_loss_mlp": 0.01030436, + "balance_loss_clip": 1.04786062, + "balance_loss_mlp": 1.02209115, + "epoch": 0.4549990981783202, + "flos": 22857142222080.0, + "grad_norm": 5.723010136386818, + "language_loss": 0.84787691, + "learning_rate": 2.3862643491819766e-06, + "loss": 0.87006956, + "num_input_tokens_seen": 81505670, + "step": 3784, + "time_per_iteration": 2.657317876815796 + }, + { + "auxiliary_loss_clip": 0.01193022, + "auxiliary_loss_mlp": 0.01027138, + "balance_loss_clip": 1.05317497, + "balance_loss_mlp": 1.01892471, + "epoch": 0.4551193410689593, + "flos": 23258587599360.0, + "grad_norm": 1.9012158812107895, + "language_loss": 0.84600794, + "learning_rate": 2.3855000141508186e-06, + "loss": 0.86820954, + "num_input_tokens_seen": 81525825, + "step": 3785, + "time_per_iteration": 2.6715967655181885 + }, + { + "auxiliary_loss_clip": 0.01208862, + "auxiliary_loss_mlp": 0.01031021, + "balance_loss_clip": 1.05567169, + "balance_loss_mlp": 1.02247977, + "epoch": 0.4552395839595984, + "flos": 20777519473920.0, + "grad_norm": 2.57939430381117, + "language_loss": 0.83752084, + "learning_rate": 2.3847356206396143e-06, + "loss": 0.85991967, + "num_input_tokens_seen": 81543135, + "step": 3786, + "time_per_iteration": 3.6224236488342285 + }, + { + "auxiliary_loss_clip": 0.01188218, + "auxiliary_loss_mlp": 0.01028419, + "balance_loss_clip": 1.05659974, + "balance_loss_mlp": 1.01969898, + "epoch": 0.45535982685023746, + "flos": 23257510191360.0, + "grad_norm": 1.505268203683487, + "language_loss": 0.7840727, + "learning_rate": 2.3839711687643227e-06, + "loss": 0.80623913, + "num_input_tokens_seen": 81564360, + "step": 3787, + "time_per_iteration": 2.6389896869659424 + }, + { + "auxiliary_loss_clip": 0.01201021, + "auxiliary_loss_mlp": 0.01029122, + "balance_loss_clip": 1.05827451, + "balance_loss_mlp": 1.01987159, + "epoch": 0.45548006974087657, + "flos": 19646117907840.0, + "grad_norm": 2.90414586087923, + "language_loss": 0.74261642, + "learning_rate": 2.38320665864091e-06, + "loss": 0.76491779, + "num_input_tokens_seen": 81583710, + "step": 3788, + "time_per_iteration": 2.704502582550049 + }, + { + "auxiliary_loss_clip": 0.01234149, + "auxiliary_loss_mlp": 0.01031603, + "balance_loss_clip": 1.04733157, + "balance_loss_mlp": 1.02294302, + "epoch": 0.4556003126315157, + "flos": 20047778766720.0, + "grad_norm": 1.7227119238950324, + "language_loss": 0.81901503, + "learning_rate": 2.3824420903853516e-06, + "loss": 0.84167254, + "num_input_tokens_seen": 81602175, + "step": 3789, + "time_per_iteration": 2.8030292987823486 + }, + { + "auxiliary_loss_clip": 0.01196377, + "auxiliary_loss_mlp": 0.0103153, + "balance_loss_clip": 1.0554657, + "balance_loss_mlp": 1.02332878, + "epoch": 0.45572055552215474, + "flos": 22959738443520.0, + "grad_norm": 2.3558075969870553, + "language_loss": 0.81909692, + "learning_rate": 2.3816774641136324e-06, + "loss": 0.84137607, + "num_input_tokens_seen": 81619430, + "step": 3790, + "time_per_iteration": 2.711970806121826 + }, + { + "auxiliary_loss_clip": 0.01195096, + "auxiliary_loss_mlp": 0.01379567, + "balance_loss_clip": 1.05434108, + "balance_loss_mlp": 1.00022209, + "epoch": 0.45584079841279385, + "flos": 33109925535360.0, + "grad_norm": 1.7234023635412417, + "language_loss": 0.71635652, + "learning_rate": 2.380912779941745e-06, + "loss": 0.7421031, + "num_input_tokens_seen": 81642550, + "step": 3791, + "time_per_iteration": 2.8364346027374268 + }, + { + "auxiliary_loss_clip": 0.01196407, + "auxiliary_loss_mlp": 0.01041017, + "balance_loss_clip": 1.05174911, + "balance_loss_mlp": 1.03153408, + "epoch": 0.45596104130343296, + "flos": 27272179445760.0, + "grad_norm": 2.568484358698114, + "language_loss": 0.83089691, + "learning_rate": 2.3801480379856918e-06, + "loss": 0.85327113, + "num_input_tokens_seen": 81664260, + "step": 3792, + "time_per_iteration": 2.7267165184020996 + }, + { + "auxiliary_loss_clip": 0.01207315, + "auxiliary_loss_mlp": 0.01028471, + "balance_loss_clip": 1.0540055, + "balance_loss_mlp": 1.01996493, + "epoch": 0.456081284194072, + "flos": 21579799697280.0, + "grad_norm": 1.6855968423453078, + "language_loss": 0.83701068, + "learning_rate": 2.379383238361484e-06, + "loss": 0.85936856, + "num_input_tokens_seen": 81683620, + "step": 3793, + "time_per_iteration": 2.6948025226593018 + }, + { + "auxiliary_loss_clip": 0.01191137, + "auxiliary_loss_mlp": 0.0103129, + "balance_loss_clip": 1.05201447, + "balance_loss_mlp": 1.02283835, + "epoch": 0.4562015270847111, + "flos": 35918822113920.0, + "grad_norm": 1.8352249607053603, + "language_loss": 0.79850698, + "learning_rate": 2.3786183811851407e-06, + "loss": 0.82073128, + "num_input_tokens_seen": 81704325, + "step": 3794, + "time_per_iteration": 2.7473297119140625 + }, + { + "auxiliary_loss_clip": 0.01191043, + "auxiliary_loss_mlp": 0.01034934, + "balance_loss_clip": 1.05908144, + "balance_loss_mlp": 1.0254513, + "epoch": 0.45632176997535023, + "flos": 13589783602560.0, + "grad_norm": 1.7728314788917396, + "language_loss": 0.80291438, + "learning_rate": 2.3778534665726892e-06, + "loss": 0.82517421, + "num_input_tokens_seen": 81721155, + "step": 3795, + "time_per_iteration": 2.5893068313598633 + }, + { + "auxiliary_loss_clip": 0.01185836, + "auxiliary_loss_mlp": 0.01025311, + "balance_loss_clip": 1.05439436, + "balance_loss_mlp": 1.01741886, + "epoch": 0.4564420128659893, + "flos": 32635401937920.0, + "grad_norm": 2.1062057547210737, + "language_loss": 0.72775924, + "learning_rate": 2.377088494640168e-06, + "loss": 0.74987066, + "num_input_tokens_seen": 81742905, + "step": 3796, + "time_per_iteration": 2.726823329925537 + }, + { + "auxiliary_loss_clip": 0.01193962, + "auxiliary_loss_mlp": 0.01025964, + "balance_loss_clip": 1.05541825, + "balance_loss_mlp": 1.01814365, + "epoch": 0.4565622557566284, + "flos": 20377690208640.0, + "grad_norm": 1.8546962117731065, + "language_loss": 0.78198814, + "learning_rate": 2.3763234655036216e-06, + "loss": 0.80418742, + "num_input_tokens_seen": 81762105, + "step": 3797, + "time_per_iteration": 2.6834259033203125 + }, + { + "auxiliary_loss_clip": 0.01207633, + "auxiliary_loss_mlp": 0.01029454, + "balance_loss_clip": 1.04590726, + "balance_loss_mlp": 1.0212642, + "epoch": 0.45668249864726745, + "flos": 25374372364800.0, + "grad_norm": 1.8373830368488537, + "language_loss": 0.86752951, + "learning_rate": 2.3755583792791046e-06, + "loss": 0.88990033, + "num_input_tokens_seen": 81781975, + "step": 3798, + "time_per_iteration": 2.7261831760406494 + }, + { + "auxiliary_loss_clip": 0.01199442, + "auxiliary_loss_mlp": 0.01027967, + "balance_loss_clip": 1.05558705, + "balance_loss_mlp": 1.01950908, + "epoch": 0.45680274153790656, + "flos": 15559806977280.0, + "grad_norm": 2.0030616870333007, + "language_loss": 0.74705708, + "learning_rate": 2.3747932360826803e-06, + "loss": 0.7693311, + "num_input_tokens_seen": 81798905, + "step": 3799, + "time_per_iteration": 2.681525230407715 + }, + { + "auxiliary_loss_clip": 0.01197186, + "auxiliary_loss_mlp": 0.01027002, + "balance_loss_clip": 1.05575645, + "balance_loss_mlp": 1.01827002, + "epoch": 0.4569229844285457, + "flos": 19792884879360.0, + "grad_norm": 1.9651913469688747, + "language_loss": 0.82127833, + "learning_rate": 2.3740280360304205e-06, + "loss": 0.84352022, + "num_input_tokens_seen": 81816630, + "step": 3800, + "time_per_iteration": 2.602782726287842 + }, + { + "auxiliary_loss_clip": 0.01211052, + "auxiliary_loss_mlp": 0.01029832, + "balance_loss_clip": 1.05199051, + "balance_loss_mlp": 1.02163005, + "epoch": 0.45704322731918473, + "flos": 24093941270400.0, + "grad_norm": 1.6275884560420801, + "language_loss": 0.67931175, + "learning_rate": 2.3732627792384038e-06, + "loss": 0.7017206, + "num_input_tokens_seen": 81837700, + "step": 3801, + "time_per_iteration": 2.7638015747070312 + }, + { + "auxiliary_loss_clip": 0.01188818, + "auxiliary_loss_mlp": 0.01028078, + "balance_loss_clip": 1.05627489, + "balance_loss_mlp": 1.01960242, + "epoch": 0.45716347020982384, + "flos": 31317803245440.0, + "grad_norm": 1.8991749805647313, + "language_loss": 0.75636917, + "learning_rate": 2.3724974658227207e-06, + "loss": 0.77853811, + "num_input_tokens_seen": 81858490, + "step": 3802, + "time_per_iteration": 2.6652779579162598 + }, + { + "auxiliary_loss_clip": 0.01208467, + "auxiliary_loss_mlp": 0.01379513, + "balance_loss_clip": 1.05655122, + "balance_loss_mlp": 1.00019264, + "epoch": 0.45728371310046295, + "flos": 26501392471680.0, + "grad_norm": 2.0936042131741357, + "language_loss": 0.71237528, + "learning_rate": 2.3717320958994687e-06, + "loss": 0.73825508, + "num_input_tokens_seen": 81876050, + "step": 3803, + "time_per_iteration": 2.7391650676727295 + }, + { + "auxiliary_loss_clip": 0.01207814, + "auxiliary_loss_mlp": 0.01025309, + "balance_loss_clip": 1.04528415, + "balance_loss_mlp": 1.01709569, + "epoch": 0.457403955991102, + "flos": 17929408222080.0, + "grad_norm": 1.8281019571713433, + "language_loss": 0.70430231, + "learning_rate": 2.3709666695847534e-06, + "loss": 0.72663343, + "num_input_tokens_seen": 81894230, + "step": 3804, + "time_per_iteration": 2.75313663482666 + }, + { + "auxiliary_loss_clip": 0.01212628, + "auxiliary_loss_mlp": 0.01028464, + "balance_loss_clip": 1.04703522, + "balance_loss_mlp": 1.02052426, + "epoch": 0.4575241988817411, + "flos": 42230660837760.0, + "grad_norm": 1.6669169843992588, + "language_loss": 0.70282829, + "learning_rate": 2.370201186994689e-06, + "loss": 0.72523916, + "num_input_tokens_seen": 81917915, + "step": 3805, + "time_per_iteration": 2.9236996173858643 + }, + { + "auxiliary_loss_clip": 0.01195654, + "auxiliary_loss_mlp": 0.01028267, + "balance_loss_clip": 1.0536499, + "balance_loss_mlp": 1.0199765, + "epoch": 0.45764444177238023, + "flos": 30117309868800.0, + "grad_norm": 1.721323437646134, + "language_loss": 0.70060062, + "learning_rate": 2.369435648245399e-06, + "loss": 0.72283989, + "num_input_tokens_seen": 81938130, + "step": 3806, + "time_per_iteration": 3.714301347732544 + }, + { + "auxiliary_loss_clip": 0.01203155, + "auxiliary_loss_mlp": 0.0103622, + "balance_loss_clip": 1.05179846, + "balance_loss_mlp": 1.02740502, + "epoch": 0.4577646846630193, + "flos": 24060293205120.0, + "grad_norm": 1.6247356576344953, + "language_loss": 0.85047442, + "learning_rate": 2.368670053453015e-06, + "loss": 0.87286818, + "num_input_tokens_seen": 81959820, + "step": 3807, + "time_per_iteration": 3.58158540725708 + }, + { + "auxiliary_loss_clip": 0.01205401, + "auxiliary_loss_mlp": 0.01026188, + "balance_loss_clip": 1.05883491, + "balance_loss_mlp": 1.01815319, + "epoch": 0.4578849275536584, + "flos": 17418578952960.0, + "grad_norm": 2.091536086805311, + "language_loss": 0.74262148, + "learning_rate": 2.3679044027336757e-06, + "loss": 0.7649374, + "num_input_tokens_seen": 81975710, + "step": 3808, + "time_per_iteration": 3.479231357574463 + }, + { + "auxiliary_loss_clip": 0.01189401, + "auxiliary_loss_mlp": 0.01027392, + "balance_loss_clip": 1.05581808, + "balance_loss_mlp": 1.01936936, + "epoch": 0.4580051704442975, + "flos": 13510169107200.0, + "grad_norm": 2.556490167406868, + "language_loss": 0.68924904, + "learning_rate": 2.3671386962035326e-06, + "loss": 0.71141696, + "num_input_tokens_seen": 81993180, + "step": 3809, + "time_per_iteration": 2.6062498092651367 + }, + { + "auxiliary_loss_clip": 0.01199205, + "auxiliary_loss_mlp": 0.01036673, + "balance_loss_clip": 1.05575728, + "balance_loss_mlp": 1.02734506, + "epoch": 0.45812541333493656, + "flos": 18037606965120.0, + "grad_norm": 1.8767514878591076, + "language_loss": 0.68614918, + "learning_rate": 2.3663729339787405e-06, + "loss": 0.70850796, + "num_input_tokens_seen": 82010115, + "step": 3810, + "time_per_iteration": 2.6042275428771973 + }, + { + "auxiliary_loss_clip": 0.01187949, + "auxiliary_loss_mlp": 0.01031972, + "balance_loss_clip": 1.05623245, + "balance_loss_mlp": 1.02304947, + "epoch": 0.45824565622557567, + "flos": 20222196232320.0, + "grad_norm": 2.144001803140168, + "language_loss": 0.73180544, + "learning_rate": 2.365607116175466e-06, + "loss": 0.75400466, + "num_input_tokens_seen": 82025540, + "step": 3811, + "time_per_iteration": 2.6566452980041504 + }, + { + "auxiliary_loss_clip": 0.01186238, + "auxiliary_loss_mlp": 0.01029493, + "balance_loss_clip": 1.05500495, + "balance_loss_mlp": 1.02159548, + "epoch": 0.4583658991162148, + "flos": 19864885691520.0, + "grad_norm": 2.2085988948330946, + "language_loss": 0.66857958, + "learning_rate": 2.3648412429098825e-06, + "loss": 0.69073689, + "num_input_tokens_seen": 82043890, + "step": 3812, + "time_per_iteration": 2.6122543811798096 + }, + { + "auxiliary_loss_clip": 0.01211905, + "auxiliary_loss_mlp": 0.01030123, + "balance_loss_clip": 1.0539515, + "balance_loss_mlp": 1.02106309, + "epoch": 0.45848614200685384, + "flos": 21029935322880.0, + "grad_norm": 1.7058913333378085, + "language_loss": 0.81677443, + "learning_rate": 2.364075314298172e-06, + "loss": 0.83919472, + "num_input_tokens_seen": 82061345, + "step": 3813, + "time_per_iteration": 3.6316702365875244 + }, + { + "auxiliary_loss_clip": 0.01200031, + "auxiliary_loss_mlp": 0.01379218, + "balance_loss_clip": 1.05586171, + "balance_loss_mlp": 1.00005257, + "epoch": 0.45860638489749295, + "flos": 21069293650560.0, + "grad_norm": 2.0112598495155494, + "language_loss": 0.70093846, + "learning_rate": 2.3633093304565267e-06, + "loss": 0.726731, + "num_input_tokens_seen": 82080400, + "step": 3814, + "time_per_iteration": 2.6663029193878174 + }, + { + "auxiliary_loss_clip": 0.01193893, + "auxiliary_loss_mlp": 0.0104078, + "balance_loss_clip": 1.05968106, + "balance_loss_mlp": 1.03233433, + "epoch": 0.458726627788132, + "flos": 26833889692800.0, + "grad_norm": 2.209131934214169, + "language_loss": 0.62349629, + "learning_rate": 2.3625432915011443e-06, + "loss": 0.64584297, + "num_input_tokens_seen": 82102310, + "step": 3815, + "time_per_iteration": 2.6730897426605225 + }, + { + "auxiliary_loss_clip": 0.01196352, + "auxiliary_loss_mlp": 0.0103303, + "balance_loss_clip": 1.0515461, + "balance_loss_mlp": 1.02373791, + "epoch": 0.4588468706787711, + "flos": 24097927680000.0, + "grad_norm": 1.634699643324753, + "language_loss": 0.65174454, + "learning_rate": 2.3617771975482334e-06, + "loss": 0.67403835, + "num_input_tokens_seen": 82121140, + "step": 3816, + "time_per_iteration": 2.72660493850708 + }, + { + "auxiliary_loss_clip": 0.01217639, + "auxiliary_loss_mlp": 0.01031316, + "balance_loss_clip": 1.04849589, + "balance_loss_mlp": 1.02304864, + "epoch": 0.4589671135694102, + "flos": 17889331622400.0, + "grad_norm": 1.5280173819384004, + "language_loss": 0.74362135, + "learning_rate": 2.3610110487140083e-06, + "loss": 0.7661109, + "num_input_tokens_seen": 82139575, + "step": 3817, + "time_per_iteration": 2.744433879852295 + }, + { + "auxiliary_loss_clip": 0.01202993, + "auxiliary_loss_mlp": 0.01029197, + "balance_loss_clip": 1.05195332, + "balance_loss_mlp": 1.02066207, + "epoch": 0.4590873564600493, + "flos": 25626967781760.0, + "grad_norm": 1.6307001822860516, + "language_loss": 0.81013179, + "learning_rate": 2.360244845114695e-06, + "loss": 0.83245373, + "num_input_tokens_seen": 82159195, + "step": 3818, + "time_per_iteration": 2.715616464614868 + }, + { + "auxiliary_loss_clip": 0.01200054, + "auxiliary_loss_mlp": 0.01030045, + "balance_loss_clip": 1.05395222, + "balance_loss_mlp": 1.02140498, + "epoch": 0.4592075993506884, + "flos": 18514788168960.0, + "grad_norm": 3.0481681217389647, + "language_loss": 0.6841284, + "learning_rate": 2.3594785868665245e-06, + "loss": 0.70642936, + "num_input_tokens_seen": 82175500, + "step": 3819, + "time_per_iteration": 2.667027235031128 + }, + { + "auxiliary_loss_clip": 0.01216243, + "auxiliary_loss_mlp": 0.01379609, + "balance_loss_clip": 1.05207574, + "balance_loss_mlp": 1.00012779, + "epoch": 0.4593278422413275, + "flos": 20631111638400.0, + "grad_norm": 2.4103799741228946, + "language_loss": 0.80595541, + "learning_rate": 2.3587122740857386e-06, + "loss": 0.83191395, + "num_input_tokens_seen": 82192600, + "step": 3820, + "time_per_iteration": 2.830845832824707 + }, + { + "auxiliary_loss_clip": 0.01195732, + "auxiliary_loss_mlp": 0.01029056, + "balance_loss_clip": 1.05388474, + "balance_loss_mlp": 1.02131319, + "epoch": 0.45944808513196655, + "flos": 21358517961600.0, + "grad_norm": 1.8259260984271954, + "language_loss": 0.78047675, + "learning_rate": 2.357945906888586e-06, + "loss": 0.80272466, + "num_input_tokens_seen": 82212040, + "step": 3821, + "time_per_iteration": 2.6365814208984375 + }, + { + "auxiliary_loss_clip": 0.01199176, + "auxiliary_loss_mlp": 0.01026526, + "balance_loss_clip": 1.0561564, + "balance_loss_mlp": 1.01775193, + "epoch": 0.45956832802260567, + "flos": 21427789340160.0, + "grad_norm": 2.4672067318162085, + "language_loss": 0.79736781, + "learning_rate": 2.357179485391324e-06, + "loss": 0.81962478, + "num_input_tokens_seen": 82229895, + "step": 3822, + "time_per_iteration": 2.676821231842041 + }, + { + "auxiliary_loss_clip": 0.011852, + "auxiliary_loss_mlp": 0.01032526, + "balance_loss_clip": 1.05594754, + "balance_loss_mlp": 1.02460504, + "epoch": 0.4596885709132448, + "flos": 22382654538240.0, + "grad_norm": 1.79813696164669, + "language_loss": 0.86382419, + "learning_rate": 2.3564130097102173e-06, + "loss": 0.88600147, + "num_input_tokens_seen": 82249550, + "step": 3823, + "time_per_iteration": 2.6026313304901123 + }, + { + "auxiliary_loss_clip": 0.01194842, + "auxiliary_loss_mlp": 0.01027969, + "balance_loss_clip": 1.05129719, + "balance_loss_mlp": 1.01993978, + "epoch": 0.45980881380388383, + "flos": 28981957806720.0, + "grad_norm": 2.066174819258397, + "language_loss": 0.7517637, + "learning_rate": 2.355646479961541e-06, + "loss": 0.77399182, + "num_input_tokens_seen": 82268860, + "step": 3824, + "time_per_iteration": 2.7754480838775635 + }, + { + "auxiliary_loss_clip": 0.01186004, + "auxiliary_loss_mlp": 0.01024292, + "balance_loss_clip": 1.05412459, + "balance_loss_mlp": 1.01601267, + "epoch": 0.45992905669452294, + "flos": 33396599980800.0, + "grad_norm": 1.7816835225742387, + "language_loss": 0.71202248, + "learning_rate": 2.354879896261576e-06, + "loss": 0.73412544, + "num_input_tokens_seen": 82289070, + "step": 3825, + "time_per_iteration": 2.67010498046875 + }, + { + "auxiliary_loss_clip": 0.01207037, + "auxiliary_loss_mlp": 0.01034791, + "balance_loss_clip": 1.05087066, + "balance_loss_mlp": 1.02652991, + "epoch": 0.46004929958516205, + "flos": 36318184502400.0, + "grad_norm": 1.8428919483118322, + "language_loss": 0.5706557, + "learning_rate": 2.3541132587266133e-06, + "loss": 0.59307396, + "num_input_tokens_seen": 82311790, + "step": 3826, + "time_per_iteration": 2.7985970973968506 + }, + { + "auxiliary_loss_clip": 0.01218745, + "auxiliary_loss_mlp": 0.01025496, + "balance_loss_clip": 1.05195761, + "balance_loss_mlp": 1.01679397, + "epoch": 0.4601695424758011, + "flos": 17238451224960.0, + "grad_norm": 1.9046528500150963, + "language_loss": 0.69262666, + "learning_rate": 2.3533465674729515e-06, + "loss": 0.71506912, + "num_input_tokens_seen": 82329020, + "step": 3827, + "time_per_iteration": 2.71631121635437 + }, + { + "auxiliary_loss_clip": 0.01189626, + "auxiliary_loss_mlp": 0.01034658, + "balance_loss_clip": 1.05735898, + "balance_loss_mlp": 1.02584898, + "epoch": 0.4602897853664402, + "flos": 15888425529600.0, + "grad_norm": 2.155173923819626, + "language_loss": 0.72716415, + "learning_rate": 2.352579822616895e-06, + "loss": 0.74940699, + "num_input_tokens_seen": 82346455, + "step": 3828, + "time_per_iteration": 2.603438377380371 + }, + { + "auxiliary_loss_clip": 0.01203964, + "auxiliary_loss_mlp": 0.01028893, + "balance_loss_clip": 1.05150259, + "balance_loss_mlp": 1.02074516, + "epoch": 0.4604100282570793, + "flos": 25412617370880.0, + "grad_norm": 1.7644036283540223, + "language_loss": 0.77871788, + "learning_rate": 2.351813024274761e-06, + "loss": 0.80104643, + "num_input_tokens_seen": 82367810, + "step": 3829, + "time_per_iteration": 2.736706018447876 + }, + { + "auxiliary_loss_clip": 0.0121743, + "auxiliary_loss_mlp": 0.01034522, + "balance_loss_clip": 1.05199909, + "balance_loss_mlp": 1.02629042, + "epoch": 0.4605302711477184, + "flos": 27630711048960.0, + "grad_norm": 1.7652884629049768, + "language_loss": 0.73676544, + "learning_rate": 2.3510461725628693e-06, + "loss": 0.75928497, + "num_input_tokens_seen": 82388275, + "step": 3830, + "time_per_iteration": 2.7504844665527344 + }, + { + "auxiliary_loss_clip": 0.01216878, + "auxiliary_loss_mlp": 0.0102548, + "balance_loss_clip": 1.05124235, + "balance_loss_mlp": 1.01694417, + "epoch": 0.4606505140383575, + "flos": 23839657914240.0, + "grad_norm": 1.7782980587642712, + "language_loss": 0.70979989, + "learning_rate": 2.350279267597554e-06, + "loss": 0.73222351, + "num_input_tokens_seen": 82408915, + "step": 3831, + "time_per_iteration": 2.7624588012695312 + }, + { + "auxiliary_loss_clip": 0.01199126, + "auxiliary_loss_mlp": 0.01026054, + "balance_loss_clip": 1.05608892, + "balance_loss_mlp": 1.01717854, + "epoch": 0.46077075692899655, + "flos": 16107013745280.0, + "grad_norm": 2.1605835272897744, + "language_loss": 0.82970113, + "learning_rate": 2.3495123094951515e-06, + "loss": 0.85195291, + "num_input_tokens_seen": 82427260, + "step": 3832, + "time_per_iteration": 3.552168607711792 + }, + { + "auxiliary_loss_clip": 0.0119456, + "auxiliary_loss_mlp": 0.01031485, + "balance_loss_clip": 1.05024779, + "balance_loss_mlp": 1.02266335, + "epoch": 0.46089099981963566, + "flos": 48798147634560.0, + "grad_norm": 2.070824345862937, + "language_loss": 0.75902236, + "learning_rate": 2.34874529837201e-06, + "loss": 0.78128278, + "num_input_tokens_seen": 82450805, + "step": 3833, + "time_per_iteration": 4.728445768356323 + }, + { + "auxiliary_loss_clip": 0.0122338, + "auxiliary_loss_mlp": 0.01034549, + "balance_loss_clip": 1.0459373, + "balance_loss_mlp": 1.02682424, + "epoch": 0.46101124271027477, + "flos": 19099234362240.0, + "grad_norm": 1.9113924095573207, + "language_loss": 0.78822005, + "learning_rate": 2.347978234344483e-06, + "loss": 0.81079936, + "num_input_tokens_seen": 82467010, + "step": 3834, + "time_per_iteration": 2.7612178325653076 + }, + { + "auxiliary_loss_clip": 0.01201148, + "auxiliary_loss_mlp": 0.01035326, + "balance_loss_clip": 1.05595446, + "balance_loss_mlp": 1.02690125, + "epoch": 0.4611314856009138, + "flos": 39347931853440.0, + "grad_norm": 1.8318025451246998, + "language_loss": 0.69062817, + "learning_rate": 2.347211117528935e-06, + "loss": 0.71299291, + "num_input_tokens_seen": 82489310, + "step": 3835, + "time_per_iteration": 2.776599407196045 + }, + { + "auxiliary_loss_clip": 0.01227492, + "auxiliary_loss_mlp": 0.0102743, + "balance_loss_clip": 1.05690193, + "balance_loss_mlp": 1.01943672, + "epoch": 0.46125172849155294, + "flos": 20810772489600.0, + "grad_norm": 2.311613763529573, + "language_loss": 0.71480238, + "learning_rate": 2.3464439480417374e-06, + "loss": 0.73735166, + "num_input_tokens_seen": 82508830, + "step": 3836, + "time_per_iteration": 2.7576661109924316 + }, + { + "auxiliary_loss_clip": 0.01200161, + "auxiliary_loss_mlp": 0.01027497, + "balance_loss_clip": 1.05488932, + "balance_loss_mlp": 1.01867509, + "epoch": 0.46137197138219205, + "flos": 17930808852480.0, + "grad_norm": 3.5115580740446575, + "language_loss": 0.7721622, + "learning_rate": 2.3456767259992676e-06, + "loss": 0.79443872, + "num_input_tokens_seen": 82526475, + "step": 3837, + "time_per_iteration": 2.653097152709961 + }, + { + "auxiliary_loss_clip": 0.01187792, + "auxiliary_loss_mlp": 0.01379254, + "balance_loss_clip": 1.05407, + "balance_loss_mlp": 1.00008452, + "epoch": 0.4614922142728311, + "flos": 16836610798080.0, + "grad_norm": 2.0373739874654038, + "language_loss": 0.8877874, + "learning_rate": 2.3449094515179135e-06, + "loss": 0.91345787, + "num_input_tokens_seen": 82543935, + "step": 3838, + "time_per_iteration": 2.578526496887207 + }, + { + "auxiliary_loss_clip": 0.01209307, + "auxiliary_loss_mlp": 0.01030986, + "balance_loss_clip": 1.05221128, + "balance_loss_mlp": 1.02245688, + "epoch": 0.4616124571634702, + "flos": 26614906427520.0, + "grad_norm": 1.619435636345467, + "language_loss": 0.81813514, + "learning_rate": 2.34414212471407e-06, + "loss": 0.84053802, + "num_input_tokens_seen": 82563730, + "step": 3839, + "time_per_iteration": 3.6854357719421387 + }, + { + "auxiliary_loss_clip": 0.01203795, + "auxiliary_loss_mlp": 0.01036959, + "balance_loss_clip": 1.05477238, + "balance_loss_mlp": 1.0283103, + "epoch": 0.4617327000541093, + "flos": 20340127560960.0, + "grad_norm": 2.136610205994255, + "language_loss": 0.72927433, + "learning_rate": 2.3433747457041394e-06, + "loss": 0.75168192, + "num_input_tokens_seen": 82582435, + "step": 3840, + "time_per_iteration": 2.6587071418762207 + }, + { + "auxiliary_loss_clip": 0.01214633, + "auxiliary_loss_mlp": 0.01026101, + "balance_loss_clip": 1.05413342, + "balance_loss_mlp": 1.01725006, + "epoch": 0.4618529429447484, + "flos": 29570749545600.0, + "grad_norm": 1.6392162045846665, + "language_loss": 0.8509165, + "learning_rate": 2.342607314604533e-06, + "loss": 0.87332386, + "num_input_tokens_seen": 82602185, + "step": 3841, + "time_per_iteration": 2.764845609664917 + }, + { + "auxiliary_loss_clip": 0.01196554, + "auxiliary_loss_mlp": 0.01026264, + "balance_loss_clip": 1.05574441, + "balance_loss_mlp": 1.01743054, + "epoch": 0.4619731858353875, + "flos": 19787030962560.0, + "grad_norm": 1.650423562918002, + "language_loss": 0.84521586, + "learning_rate": 2.3418398315316694e-06, + "loss": 0.86744404, + "num_input_tokens_seen": 82620005, + "step": 3842, + "time_per_iteration": 2.6173105239868164 + }, + { + "auxiliary_loss_clip": 0.01188649, + "auxiliary_loss_mlp": 0.01037297, + "balance_loss_clip": 1.05715609, + "balance_loss_mlp": 1.0286665, + "epoch": 0.4620934287260266, + "flos": 18951138587520.0, + "grad_norm": 2.2932458434944114, + "language_loss": 0.78189003, + "learning_rate": 2.3410722966019755e-06, + "loss": 0.80414951, + "num_input_tokens_seen": 82635120, + "step": 3843, + "time_per_iteration": 2.5884244441986084 + }, + { + "auxiliary_loss_clip": 0.01197169, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 1.05589461, + "balance_loss_mlp": 1.01925063, + "epoch": 0.46221367161666566, + "flos": 37341674634240.0, + "grad_norm": 1.7311424386060705, + "language_loss": 0.65558147, + "learning_rate": 2.3403047099318848e-06, + "loss": 0.67783153, + "num_input_tokens_seen": 82659190, + "step": 3844, + "time_per_iteration": 2.8087539672851562 + }, + { + "auxiliary_loss_clip": 0.0121145, + "auxiliary_loss_mlp": 0.01028457, + "balance_loss_clip": 1.04494929, + "balance_loss_mlp": 1.02019596, + "epoch": 0.46233391450730477, + "flos": 14428549065600.0, + "grad_norm": 2.2512698434873166, + "language_loss": 0.74944097, + "learning_rate": 2.3395370716378405e-06, + "loss": 0.77184004, + "num_input_tokens_seen": 82676635, + "step": 3845, + "time_per_iteration": 2.7717528343200684 + }, + { + "auxiliary_loss_clip": 0.01198821, + "auxiliary_loss_mlp": 0.01026381, + "balance_loss_clip": 1.05296803, + "balance_loss_mlp": 1.01831329, + "epoch": 0.4624541573979438, + "flos": 22493044010880.0, + "grad_norm": 2.37328760231734, + "language_loss": 0.72544062, + "learning_rate": 2.338769381836292e-06, + "loss": 0.7476927, + "num_input_tokens_seen": 82696245, + "step": 3846, + "time_per_iteration": 2.7076408863067627 + }, + { + "auxiliary_loss_clip": 0.01215161, + "auxiliary_loss_mlp": 0.01031599, + "balance_loss_clip": 1.05566072, + "balance_loss_mlp": 1.02281356, + "epoch": 0.46257440028858293, + "flos": 14465070218880.0, + "grad_norm": 2.1408062697918404, + "language_loss": 0.73461711, + "learning_rate": 2.3380016406436984e-06, + "loss": 0.75708467, + "num_input_tokens_seen": 82713725, + "step": 3847, + "time_per_iteration": 2.7252984046936035 + }, + { + "auxiliary_loss_clip": 0.01223226, + "auxiliary_loss_mlp": 0.01030517, + "balance_loss_clip": 1.05391288, + "balance_loss_mlp": 1.02248812, + "epoch": 0.46269464317922204, + "flos": 23332204523520.0, + "grad_norm": 1.8012393179129886, + "language_loss": 0.81355751, + "learning_rate": 2.337233848176524e-06, + "loss": 0.83609492, + "num_input_tokens_seen": 82731495, + "step": 3848, + "time_per_iteration": 2.784196615219116 + }, + { + "auxiliary_loss_clip": 0.0120985, + "auxiliary_loss_mlp": 0.01028021, + "balance_loss_clip": 1.04446363, + "balance_loss_mlp": 1.0193603, + "epoch": 0.4628148860698611, + "flos": 18552027594240.0, + "grad_norm": 1.7885911188238142, + "language_loss": 0.8313669, + "learning_rate": 2.3364660045512435e-06, + "loss": 0.85374558, + "num_input_tokens_seen": 82750255, + "step": 3849, + "time_per_iteration": 2.855077028274536 + }, + { + "auxiliary_loss_clip": 0.01106981, + "auxiliary_loss_mlp": 0.01005329, + "balance_loss_clip": 1.02321875, + "balance_loss_mlp": 1.00405908, + "epoch": 0.4629351289605002, + "flos": 70667569670400.0, + "grad_norm": 0.7422436846305349, + "language_loss": 0.58204216, + "learning_rate": 2.335698109884337e-06, + "loss": 0.60316527, + "num_input_tokens_seen": 82815460, + "step": 3850, + "time_per_iteration": 3.4108567237854004 + }, + { + "auxiliary_loss_clip": 0.01130075, + "auxiliary_loss_mlp": 0.01000021, + "balance_loss_clip": 1.02294886, + "balance_loss_mlp": 0.99874514, + "epoch": 0.4630553718511393, + "flos": 59687200465920.0, + "grad_norm": 0.7871069341823179, + "language_loss": 0.59851515, + "learning_rate": 2.334930164292294e-06, + "loss": 0.61981606, + "num_input_tokens_seen": 82878010, + "step": 3851, + "time_per_iteration": 3.405941963195801 + }, + { + "auxiliary_loss_clip": 0.01210338, + "auxiliary_loss_mlp": 0.01027389, + "balance_loss_clip": 1.04566669, + "balance_loss_mlp": 1.01887178, + "epoch": 0.4631756147417784, + "flos": 15960605909760.0, + "grad_norm": 2.037474881007636, + "language_loss": 0.80312377, + "learning_rate": 2.334162167891612e-06, + "loss": 0.82550108, + "num_input_tokens_seen": 82895275, + "step": 3852, + "time_per_iteration": 2.703423261642456 + }, + { + "auxiliary_loss_clip": 0.01206523, + "auxiliary_loss_mlp": 0.0103122, + "balance_loss_clip": 1.05260038, + "balance_loss_mlp": 1.02258921, + "epoch": 0.4632958576324175, + "flos": 16472907636480.0, + "grad_norm": 1.988792736125431, + "language_loss": 0.7469039, + "learning_rate": 2.333394120798795e-06, + "loss": 0.76928139, + "num_input_tokens_seen": 82914010, + "step": 3853, + "time_per_iteration": 2.6949682235717773 + }, + { + "auxiliary_loss_clip": 0.01204232, + "auxiliary_loss_mlp": 0.01031271, + "balance_loss_clip": 1.0512197, + "balance_loss_mlp": 1.02321231, + "epoch": 0.4634161005230566, + "flos": 22346492520960.0, + "grad_norm": 3.6492523445826617, + "language_loss": 0.71749508, + "learning_rate": 2.3326260231303545e-06, + "loss": 0.73985016, + "num_input_tokens_seen": 82932610, + "step": 3854, + "time_per_iteration": 2.6930580139160156 + }, + { + "auxiliary_loss_clip": 0.01185738, + "auxiliary_loss_mlp": 0.01025602, + "balance_loss_clip": 1.05595303, + "balance_loss_mlp": 1.01741219, + "epoch": 0.46353634341369565, + "flos": 15742233175680.0, + "grad_norm": 1.6496207144703618, + "language_loss": 0.8662914, + "learning_rate": 2.331857875002811e-06, + "loss": 0.88840473, + "num_input_tokens_seen": 82951210, + "step": 3855, + "time_per_iteration": 2.687845230102539 + }, + { + "auxiliary_loss_clip": 0.01206824, + "auxiliary_loss_mlp": 0.01029652, + "balance_loss_clip": 1.05616701, + "balance_loss_mlp": 1.02152801, + "epoch": 0.46365658630433476, + "flos": 28329820433280.0, + "grad_norm": 2.537519016800457, + "language_loss": 0.76168555, + "learning_rate": 2.3310896765326916e-06, + "loss": 0.78405029, + "num_input_tokens_seen": 82972210, + "step": 3856, + "time_per_iteration": 2.7167418003082275 + }, + { + "auxiliary_loss_clip": 0.01210943, + "auxiliary_loss_mlp": 0.01029675, + "balance_loss_clip": 1.05332983, + "balance_loss_mlp": 1.02044201, + "epoch": 0.46377682919497387, + "flos": 24608074590720.0, + "grad_norm": 1.661985004604568, + "language_loss": 0.83952665, + "learning_rate": 2.330321427836531e-06, + "loss": 0.86193275, + "num_input_tokens_seen": 82994080, + "step": 3857, + "time_per_iteration": 2.7596828937530518 + }, + { + "auxiliary_loss_clip": 0.01192238, + "auxiliary_loss_mlp": 0.01029131, + "balance_loss_clip": 1.05402875, + "balance_loss_mlp": 1.02101326, + "epoch": 0.4638970720856129, + "flos": 19060953442560.0, + "grad_norm": 2.413561210382297, + "language_loss": 0.82777071, + "learning_rate": 2.3295531290308733e-06, + "loss": 0.84998441, + "num_input_tokens_seen": 83012230, + "step": 3858, + "time_per_iteration": 3.643411636352539 + }, + { + "auxiliary_loss_clip": 0.01192357, + "auxiliary_loss_mlp": 0.01379634, + "balance_loss_clip": 1.05858552, + "balance_loss_mlp": 1.00010991, + "epoch": 0.46401731497625204, + "flos": 18471012468480.0, + "grad_norm": 3.219460863983722, + "language_loss": 0.75325912, + "learning_rate": 2.3287847802322678e-06, + "loss": 0.77897906, + "num_input_tokens_seen": 83027800, + "step": 3859, + "time_per_iteration": 3.444770097732544 + }, + { + "auxiliary_loss_clip": 0.01212929, + "auxiliary_loss_mlp": 0.01032192, + "balance_loss_clip": 1.05586481, + "balance_loss_mlp": 1.02283978, + "epoch": 0.4641375578668911, + "flos": 26067053214720.0, + "grad_norm": 1.8210293954317347, + "language_loss": 0.84289694, + "learning_rate": 2.3280163815572723e-06, + "loss": 0.8653481, + "num_input_tokens_seen": 83048395, + "step": 3860, + "time_per_iteration": 3.615009307861328 + }, + { + "auxiliary_loss_clip": 0.01192618, + "auxiliary_loss_mlp": 0.01025476, + "balance_loss_clip": 1.04976547, + "balance_loss_mlp": 1.01740599, + "epoch": 0.4642578007575302, + "flos": 19570382081280.0, + "grad_norm": 1.957858283141778, + "language_loss": 0.76768327, + "learning_rate": 2.3272479331224522e-06, + "loss": 0.78986418, + "num_input_tokens_seen": 83065825, + "step": 3861, + "time_per_iteration": 2.676604986190796 + }, + { + "auxiliary_loss_clip": 0.0119021, + "auxiliary_loss_mlp": 0.01033188, + "balance_loss_clip": 1.05600059, + "balance_loss_mlp": 1.0246166, + "epoch": 0.4643780436481693, + "flos": 28186249772160.0, + "grad_norm": 1.6415136497704446, + "language_loss": 0.7858398, + "learning_rate": 2.3264794350443817e-06, + "loss": 0.80807376, + "num_input_tokens_seen": 83087920, + "step": 3862, + "time_per_iteration": 2.640301465988159 + }, + { + "auxiliary_loss_clip": 0.0119686, + "auxiliary_loss_mlp": 0.01032097, + "balance_loss_clip": 1.05223298, + "balance_loss_mlp": 1.02361488, + "epoch": 0.46449828653880837, + "flos": 25375270204800.0, + "grad_norm": 1.873209927043117, + "language_loss": 0.78220332, + "learning_rate": 2.3257108874396396e-06, + "loss": 0.80449295, + "num_input_tokens_seen": 83109015, + "step": 3863, + "time_per_iteration": 2.7010011672973633 + }, + { + "auxiliary_loss_clip": 0.01204437, + "auxiliary_loss_mlp": 0.01031226, + "balance_loss_clip": 1.05108833, + "balance_loss_mlp": 1.02158237, + "epoch": 0.4646185294294475, + "flos": 16034330574720.0, + "grad_norm": 1.952236115782581, + "language_loss": 0.73637831, + "learning_rate": 2.3249422904248152e-06, + "loss": 0.75873494, + "num_input_tokens_seen": 83127450, + "step": 3864, + "time_per_iteration": 2.6262400150299072 + }, + { + "auxiliary_loss_clip": 0.01199628, + "auxiliary_loss_mlp": 0.010254, + "balance_loss_clip": 1.05386269, + "balance_loss_mlp": 1.01697147, + "epoch": 0.4647387723200866, + "flos": 26363101109760.0, + "grad_norm": 1.3156252043234875, + "language_loss": 0.87188363, + "learning_rate": 2.324173644116504e-06, + "loss": 0.89413387, + "num_input_tokens_seen": 83150300, + "step": 3865, + "time_per_iteration": 3.6132705211639404 + }, + { + "auxiliary_loss_clip": 0.01191056, + "auxiliary_loss_mlp": 0.01025275, + "balance_loss_clip": 1.0541271, + "balance_loss_mlp": 1.01695991, + "epoch": 0.46485901521072565, + "flos": 27160209774720.0, + "grad_norm": 1.8544908197236625, + "language_loss": 0.81385756, + "learning_rate": 2.3234049486313087e-06, + "loss": 0.83602083, + "num_input_tokens_seen": 83171750, + "step": 3866, + "time_per_iteration": 2.715825319290161 + }, + { + "auxiliary_loss_clip": 0.01196554, + "auxiliary_loss_mlp": 0.01029842, + "balance_loss_clip": 1.05461276, + "balance_loss_mlp": 1.02204013, + "epoch": 0.46497925810136476, + "flos": 24279851088000.0, + "grad_norm": 1.913340100302243, + "language_loss": 0.75949603, + "learning_rate": 2.322636204085839e-06, + "loss": 0.78175998, + "num_input_tokens_seen": 83191820, + "step": 3867, + "time_per_iteration": 2.632784128189087 + }, + { + "auxiliary_loss_clip": 0.01192533, + "auxiliary_loss_mlp": 0.01026854, + "balance_loss_clip": 1.04875553, + "balance_loss_mlp": 1.01871836, + "epoch": 0.46509950099200387, + "flos": 16253134272000.0, + "grad_norm": 2.2726860440799856, + "language_loss": 0.79042864, + "learning_rate": 2.3218674105967143e-06, + "loss": 0.81262255, + "num_input_tokens_seen": 83210085, + "step": 3868, + "time_per_iteration": 2.686399459838867 + }, + { + "auxiliary_loss_clip": 0.01197736, + "auxiliary_loss_mlp": 0.01035127, + "balance_loss_clip": 1.050717, + "balance_loss_mlp": 1.02705622, + "epoch": 0.4652197438826429, + "flos": 23442270773760.0, + "grad_norm": 1.650242350982332, + "language_loss": 0.83778697, + "learning_rate": 2.3210985682805593e-06, + "loss": 0.86011565, + "num_input_tokens_seen": 83231865, + "step": 3869, + "time_per_iteration": 2.7161362171173096 + }, + { + "auxiliary_loss_clip": 0.0119252, + "auxiliary_loss_mlp": 0.01027672, + "balance_loss_clip": 1.0600605, + "balance_loss_mlp": 1.01899409, + "epoch": 0.46533998677328203, + "flos": 16216397637120.0, + "grad_norm": 2.479718097381875, + "language_loss": 0.68393064, + "learning_rate": 2.320329677254007e-06, + "loss": 0.70613253, + "num_input_tokens_seen": 83249195, + "step": 3870, + "time_per_iteration": 2.5672481060028076 + }, + { + "auxiliary_loss_clip": 0.0118993, + "auxiliary_loss_mlp": 0.01028929, + "balance_loss_clip": 1.05841458, + "balance_loss_mlp": 1.02048326, + "epoch": 0.46546022966392114, + "flos": 21141869080320.0, + "grad_norm": 2.258350547360003, + "language_loss": 0.72385955, + "learning_rate": 2.319560737633697e-06, + "loss": 0.74604809, + "num_input_tokens_seen": 83267915, + "step": 3871, + "time_per_iteration": 2.607752799987793 + }, + { + "auxiliary_loss_clip": 0.01218619, + "auxiliary_loss_mlp": 0.01034497, + "balance_loss_clip": 1.04853952, + "balance_loss_mlp": 1.02565193, + "epoch": 0.4655804725545602, + "flos": 41171942442240.0, + "grad_norm": 1.5240921999002763, + "language_loss": 0.68046445, + "learning_rate": 2.3187917495362775e-06, + "loss": 0.70299566, + "num_input_tokens_seen": 83292325, + "step": 3872, + "time_per_iteration": 2.8714752197265625 + }, + { + "auxiliary_loss_clip": 0.01219882, + "auxiliary_loss_mlp": 0.01031438, + "balance_loss_clip": 1.05035734, + "balance_loss_mlp": 1.02258658, + "epoch": 0.4657007154451993, + "flos": 19570956698880.0, + "grad_norm": 2.2746585289921852, + "language_loss": 0.76358318, + "learning_rate": 2.318022713078403e-06, + "loss": 0.78609633, + "num_input_tokens_seen": 83306905, + "step": 3873, + "time_per_iteration": 2.7562286853790283 + }, + { + "auxiliary_loss_clip": 0.012038, + "auxiliary_loss_mlp": 0.01025685, + "balance_loss_clip": 1.05217481, + "balance_loss_mlp": 1.01711369, + "epoch": 0.4658209583358384, + "flos": 15517826956800.0, + "grad_norm": 2.2035175075308593, + "language_loss": 0.85428607, + "learning_rate": 2.3172536283767354e-06, + "loss": 0.87658095, + "num_input_tokens_seen": 83320665, + "step": 3874, + "time_per_iteration": 2.641903877258301 + }, + { + "auxiliary_loss_clip": 0.0120843, + "auxiliary_loss_mlp": 0.01034323, + "balance_loss_clip": 1.05101013, + "balance_loss_mlp": 1.02553177, + "epoch": 0.4659412012264775, + "flos": 14903180403840.0, + "grad_norm": 2.424904418583078, + "language_loss": 0.80677587, + "learning_rate": 2.3164844955479447e-06, + "loss": 0.82920343, + "num_input_tokens_seen": 83336475, + "step": 3875, + "time_per_iteration": 2.721898317337036 + }, + { + "auxiliary_loss_clip": 0.01233103, + "auxiliary_loss_mlp": 0.0102525, + "balance_loss_clip": 1.05141878, + "balance_loss_mlp": 1.01696467, + "epoch": 0.4660614441171166, + "flos": 24425612478720.0, + "grad_norm": 1.6600168482051587, + "language_loss": 0.70718282, + "learning_rate": 2.3157153147087082e-06, + "loss": 0.72976637, + "num_input_tokens_seen": 83358365, + "step": 3876, + "time_per_iteration": 2.805201768875122 + }, + { + "auxiliary_loss_clip": 0.01232649, + "auxiliary_loss_mlp": 0.01032016, + "balance_loss_clip": 1.05344319, + "balance_loss_mlp": 1.02376103, + "epoch": 0.46618168700775564, + "flos": 22091095843200.0, + "grad_norm": 1.6504707602212925, + "language_loss": 0.83026302, + "learning_rate": 2.314946085975709e-06, + "loss": 0.85290968, + "num_input_tokens_seen": 83377345, + "step": 3877, + "time_per_iteration": 2.754513740539551 + }, + { + "auxiliary_loss_clip": 0.01199435, + "auxiliary_loss_mlp": 0.01031719, + "balance_loss_clip": 1.04850364, + "balance_loss_mlp": 1.0236541, + "epoch": 0.46630192989839475, + "flos": 26176975810560.0, + "grad_norm": 1.686686997944895, + "language_loss": 0.82398742, + "learning_rate": 2.3141768094656393e-06, + "loss": 0.84629899, + "num_input_tokens_seen": 83395920, + "step": 3878, + "time_per_iteration": 2.7721269130706787 + }, + { + "auxiliary_loss_clip": 0.0124081, + "auxiliary_loss_mlp": 0.01026583, + "balance_loss_clip": 1.04201674, + "balance_loss_mlp": 1.01850367, + "epoch": 0.46642217278903386, + "flos": 11509622150400.0, + "grad_norm": 2.7657299866902143, + "language_loss": 0.8349458, + "learning_rate": 2.3134074852951966e-06, + "loss": 0.85761964, + "num_input_tokens_seen": 83412510, + "step": 3879, + "time_per_iteration": 2.839672565460205 + }, + { + "auxiliary_loss_clip": 0.01214446, + "auxiliary_loss_mlp": 0.01030645, + "balance_loss_clip": 1.04757428, + "balance_loss_mlp": 1.02275968, + "epoch": 0.4665424156796729, + "flos": 32306819299200.0, + "grad_norm": 2.7876294247546136, + "language_loss": 0.77861416, + "learning_rate": 2.312638113581088e-06, + "loss": 0.80106509, + "num_input_tokens_seen": 83432995, + "step": 3880, + "time_per_iteration": 3.042231559753418 + }, + { + "auxiliary_loss_clip": 0.01192431, + "auxiliary_loss_mlp": 0.01027231, + "balance_loss_clip": 1.04999781, + "balance_loss_mlp": 1.01902914, + "epoch": 0.46666265857031203, + "flos": 18436179254400.0, + "grad_norm": 2.7838657136011107, + "language_loss": 0.78413439, + "learning_rate": 2.311868694440027e-06, + "loss": 0.80633104, + "num_input_tokens_seen": 83447415, + "step": 3881, + "time_per_iteration": 2.6308724880218506 + }, + { + "auxiliary_loss_clip": 0.01078859, + "auxiliary_loss_mlp": 0.01002113, + "balance_loss_clip": 1.018296, + "balance_loss_mlp": 1.00060487, + "epoch": 0.46678290146095114, + "flos": 68438989221120.0, + "grad_norm": 0.7367841281012419, + "language_loss": 0.62473893, + "learning_rate": 2.3110992279887323e-06, + "loss": 0.64554858, + "num_input_tokens_seen": 83519340, + "step": 3882, + "time_per_iteration": 3.2969276905059814 + }, + { + "auxiliary_loss_clip": 0.0122251, + "auxiliary_loss_mlp": 0.01032836, + "balance_loss_clip": 1.05554044, + "balance_loss_mlp": 1.02348948, + "epoch": 0.4669031443515902, + "flos": 17712507945600.0, + "grad_norm": 2.3613630498771494, + "language_loss": 0.84966701, + "learning_rate": 2.310329714343932e-06, + "loss": 0.87222046, + "num_input_tokens_seen": 83535490, + "step": 3883, + "time_per_iteration": 3.58286452293396 + }, + { + "auxiliary_loss_clip": 0.0119638, + "auxiliary_loss_mlp": 0.01029941, + "balance_loss_clip": 1.05121398, + "balance_loss_mlp": 1.02131605, + "epoch": 0.4670233872422293, + "flos": 23947748916480.0, + "grad_norm": 2.0148838621526814, + "language_loss": 0.81897938, + "learning_rate": 2.309560153622361e-06, + "loss": 0.84124261, + "num_input_tokens_seen": 83552400, + "step": 3884, + "time_per_iteration": 2.7029848098754883 + }, + { + "auxiliary_loss_clip": 0.01211607, + "auxiliary_loss_mlp": 0.01030503, + "balance_loss_clip": 1.05113316, + "balance_loss_mlp": 1.02174747, + "epoch": 0.4671436301328684, + "flos": 28111268131200.0, + "grad_norm": 1.9478365042252956, + "language_loss": 0.74582541, + "learning_rate": 2.3087905459407602e-06, + "loss": 0.76824647, + "num_input_tokens_seen": 83571340, + "step": 3885, + "time_per_iteration": 2.735626459121704 + }, + { + "auxiliary_loss_clip": 0.01092038, + "auxiliary_loss_mlp": 0.01000771, + "balance_loss_clip": 1.01761067, + "balance_loss_mlp": 0.99953097, + "epoch": 0.46726387302350747, + "flos": 69369684566400.0, + "grad_norm": 0.8002896852850424, + "language_loss": 0.62964064, + "learning_rate": 2.3080208914158795e-06, + "loss": 0.65056872, + "num_input_tokens_seen": 83634340, + "step": 3886, + "time_per_iteration": 4.1714842319488525 + }, + { + "auxiliary_loss_clip": 0.0120282, + "auxiliary_loss_mlp": 0.01031753, + "balance_loss_clip": 1.05505919, + "balance_loss_mlp": 1.02355182, + "epoch": 0.4673841159141466, + "flos": 25519666878720.0, + "grad_norm": 3.3420123724741786, + "language_loss": 0.72094274, + "learning_rate": 2.3072511901644753e-06, + "loss": 0.7432884, + "num_input_tokens_seen": 83653410, + "step": 3887, + "time_per_iteration": 2.6971869468688965 + }, + { + "auxiliary_loss_clip": 0.01184462, + "auxiliary_loss_mlp": 0.01030369, + "balance_loss_clip": 1.05551314, + "balance_loss_mlp": 1.02234018, + "epoch": 0.4675043588047857, + "flos": 24499265316480.0, + "grad_norm": 2.4744526552245447, + "language_loss": 0.81009763, + "learning_rate": 2.306481442303309e-06, + "loss": 0.83224595, + "num_input_tokens_seen": 83672985, + "step": 3888, + "time_per_iteration": 2.738474130630493 + }, + { + "auxiliary_loss_clip": 0.01196826, + "auxiliary_loss_mlp": 0.01028908, + "balance_loss_clip": 1.05247736, + "balance_loss_mlp": 1.01978827, + "epoch": 0.46762460169542475, + "flos": 20960771685120.0, + "grad_norm": 1.8174146756870837, + "language_loss": 0.7320261, + "learning_rate": 2.3057116479491515e-06, + "loss": 0.75428343, + "num_input_tokens_seen": 83692395, + "step": 3889, + "time_per_iteration": 2.6743357181549072 + }, + { + "auxiliary_loss_clip": 0.01187909, + "auxiliary_loss_mlp": 0.01028192, + "balance_loss_clip": 1.04952931, + "balance_loss_mlp": 1.01972198, + "epoch": 0.46774484458606386, + "flos": 19171666137600.0, + "grad_norm": 1.9554944127246285, + "language_loss": 0.76057088, + "learning_rate": 2.30494180721878e-06, + "loss": 0.78273195, + "num_input_tokens_seen": 83709735, + "step": 3890, + "time_per_iteration": 2.627511978149414 + }, + { + "auxiliary_loss_clip": 0.01195143, + "auxiliary_loss_mlp": 0.01034608, + "balance_loss_clip": 1.05468702, + "balance_loss_mlp": 1.02626324, + "epoch": 0.4678650874767029, + "flos": 17967689141760.0, + "grad_norm": 1.9567350276830058, + "language_loss": 0.89745212, + "learning_rate": 2.3041719202289794e-06, + "loss": 0.91974956, + "num_input_tokens_seen": 83725910, + "step": 3891, + "time_per_iteration": 3.5575923919677734 + }, + { + "auxiliary_loss_clip": 0.01197099, + "auxiliary_loss_mlp": 0.01024193, + "balance_loss_clip": 1.05423009, + "balance_loss_mlp": 1.01665866, + "epoch": 0.467985330367342, + "flos": 21360816432000.0, + "grad_norm": 2.7181778350544294, + "language_loss": 0.80440545, + "learning_rate": 2.30340198709654e-06, + "loss": 0.82661843, + "num_input_tokens_seen": 83745745, + "step": 3892, + "time_per_iteration": 2.6282036304473877 + }, + { + "auxiliary_loss_clip": 0.01204601, + "auxiliary_loss_mlp": 0.01034025, + "balance_loss_clip": 1.04897833, + "balance_loss_mlp": 1.02499461, + "epoch": 0.46810557325798113, + "flos": 20521835487360.0, + "grad_norm": 2.3229393694797373, + "language_loss": 0.74217224, + "learning_rate": 2.3026320079382605e-06, + "loss": 0.76455843, + "num_input_tokens_seen": 83762680, + "step": 3893, + "time_per_iteration": 2.63836932182312 + }, + { + "auxiliary_loss_clip": 0.01186989, + "auxiliary_loss_mlp": 0.010297, + "balance_loss_clip": 1.05694556, + "balance_loss_mlp": 1.02107525, + "epoch": 0.4682258161486202, + "flos": 30117848572800.0, + "grad_norm": 2.9684329508910077, + "language_loss": 0.76326621, + "learning_rate": 2.3018619828709454e-06, + "loss": 0.78543317, + "num_input_tokens_seen": 83784220, + "step": 3894, + "time_per_iteration": 2.6700332164764404 + }, + { + "auxiliary_loss_clip": 0.01193277, + "auxiliary_loss_mlp": 0.01379418, + "balance_loss_clip": 1.05578542, + "balance_loss_mlp": 1.00011218, + "epoch": 0.4683460590392593, + "flos": 25293357239040.0, + "grad_norm": 2.0248760817431997, + "language_loss": 0.82078266, + "learning_rate": 2.3010919120114084e-06, + "loss": 0.84650964, + "num_input_tokens_seen": 83800750, + "step": 3895, + "time_per_iteration": 2.673952341079712 + }, + { + "auxiliary_loss_clip": 0.01189227, + "auxiliary_loss_mlp": 0.01026612, + "balance_loss_clip": 1.04963136, + "balance_loss_mlp": 1.01787353, + "epoch": 0.4684663019298984, + "flos": 15368330551680.0, + "grad_norm": 2.222493110568683, + "language_loss": 0.65890878, + "learning_rate": 2.3003217954764672e-06, + "loss": 0.68106723, + "num_input_tokens_seen": 83815455, + "step": 3896, + "time_per_iteration": 2.56559157371521 + }, + { + "auxiliary_loss_clip": 0.01196354, + "auxiliary_loss_mlp": 0.01033143, + "balance_loss_clip": 1.05093062, + "balance_loss_mlp": 1.0244348, + "epoch": 0.46858654482053747, + "flos": 27778842737280.0, + "grad_norm": 1.5787980358556089, + "language_loss": 0.79531878, + "learning_rate": 2.299551633382949e-06, + "loss": 0.81761378, + "num_input_tokens_seen": 83835765, + "step": 3897, + "time_per_iteration": 2.688716411590576 + }, + { + "auxiliary_loss_clip": 0.01195947, + "auxiliary_loss_mlp": 0.01032199, + "balance_loss_clip": 1.0507468, + "balance_loss_mlp": 1.02366972, + "epoch": 0.4687067877111766, + "flos": 18040623707520.0, + "grad_norm": 1.8080187699570842, + "language_loss": 0.85596347, + "learning_rate": 2.2987814258476854e-06, + "loss": 0.87824494, + "num_input_tokens_seen": 83853565, + "step": 3898, + "time_per_iteration": 2.642289638519287 + }, + { + "auxiliary_loss_clip": 0.01224941, + "auxiliary_loss_mlp": 0.01028781, + "balance_loss_clip": 1.0479424, + "balance_loss_mlp": 1.01948309, + "epoch": 0.4688270306018157, + "flos": 16977380198400.0, + "grad_norm": 2.7389954436222252, + "language_loss": 0.68030602, + "learning_rate": 2.2980111729875177e-06, + "loss": 0.70284325, + "num_input_tokens_seen": 83869815, + "step": 3899, + "time_per_iteration": 2.753114700317383 + }, + { + "auxiliary_loss_clip": 0.01199938, + "auxiliary_loss_mlp": 0.01027281, + "balance_loss_clip": 1.05335999, + "balance_loss_mlp": 1.01866221, + "epoch": 0.46894727349245474, + "flos": 17821640442240.0, + "grad_norm": 1.6215063314109939, + "language_loss": 0.82114351, + "learning_rate": 2.2972408749192917e-06, + "loss": 0.84341568, + "num_input_tokens_seen": 83887545, + "step": 3900, + "time_per_iteration": 2.661888360977173 + }, + { + "auxiliary_loss_clip": 0.01193093, + "auxiliary_loss_mlp": 0.01378633, + "balance_loss_clip": 1.0545342, + "balance_loss_mlp": 1.00004911, + "epoch": 0.46906751638309385, + "flos": 21471349559040.0, + "grad_norm": 2.0658124853859094, + "language_loss": 0.6672653, + "learning_rate": 2.296470531759861e-06, + "loss": 0.69298255, + "num_input_tokens_seen": 83905645, + "step": 3901, + "time_per_iteration": 2.648991346359253 + }, + { + "auxiliary_loss_clip": 0.01206855, + "auxiliary_loss_mlp": 0.01027047, + "balance_loss_clip": 1.04923415, + "balance_loss_mlp": 1.01846409, + "epoch": 0.46918775927373296, + "flos": 20337829090560.0, + "grad_norm": 1.8681039579106036, + "language_loss": 0.79540616, + "learning_rate": 2.2957001436260866e-06, + "loss": 0.81774521, + "num_input_tokens_seen": 83922705, + "step": 3902, + "time_per_iteration": 2.748497247695923 + }, + { + "auxiliary_loss_clip": 0.01201326, + "auxiliary_loss_mlp": 0.01034668, + "balance_loss_clip": 1.05287075, + "balance_loss_mlp": 1.0254885, + "epoch": 0.469308002164372, + "flos": 18403249461120.0, + "grad_norm": 1.6150049430088373, + "language_loss": 0.73243904, + "learning_rate": 2.294929710634836e-06, + "loss": 0.75479901, + "num_input_tokens_seen": 83940795, + "step": 3903, + "time_per_iteration": 2.6758100986480713 + }, + { + "auxiliary_loss_clip": 0.01195306, + "auxiliary_loss_mlp": 0.01032511, + "balance_loss_clip": 1.05329978, + "balance_loss_mlp": 1.0233916, + "epoch": 0.46942824505501113, + "flos": 37962067363200.0, + "grad_norm": 1.7256028503486858, + "language_loss": 0.61001438, + "learning_rate": 2.2941592329029823e-06, + "loss": 0.63229251, + "num_input_tokens_seen": 83961900, + "step": 3904, + "time_per_iteration": 2.828697443008423 + }, + { + "auxiliary_loss_clip": 0.01188738, + "auxiliary_loss_mlp": 0.01030689, + "balance_loss_clip": 1.05140996, + "balance_loss_mlp": 1.02202868, + "epoch": 0.46954848794565024, + "flos": 21872507627520.0, + "grad_norm": 1.8365856137082337, + "language_loss": 0.78901517, + "learning_rate": 2.2933887105474067e-06, + "loss": 0.8112095, + "num_input_tokens_seen": 83980075, + "step": 3905, + "time_per_iteration": 2.655167818069458 + }, + { + "auxiliary_loss_clip": 0.01194973, + "auxiliary_loss_mlp": 0.01027491, + "balance_loss_clip": 1.05556285, + "balance_loss_mlp": 1.01949763, + "epoch": 0.4696687308362893, + "flos": 22016545165440.0, + "grad_norm": 1.8149596337414597, + "language_loss": 0.81613433, + "learning_rate": 2.2926181436849974e-06, + "loss": 0.838359, + "num_input_tokens_seen": 83999430, + "step": 3906, + "time_per_iteration": 2.6683034896850586 + }, + { + "auxiliary_loss_clip": 0.01192426, + "auxiliary_loss_mlp": 0.01031861, + "balance_loss_clip": 1.05221152, + "balance_loss_mlp": 1.0235287, + "epoch": 0.4697889737269284, + "flos": 21613663244160.0, + "grad_norm": 2.0690005810388774, + "language_loss": 0.73019606, + "learning_rate": 2.2918475324326478e-06, + "loss": 0.7524389, + "num_input_tokens_seen": 84019150, + "step": 3907, + "time_per_iteration": 2.6147570610046387 + }, + { + "auxiliary_loss_clip": 0.01199834, + "auxiliary_loss_mlp": 0.01379519, + "balance_loss_clip": 1.05606616, + "balance_loss_mlp": 1.00015199, + "epoch": 0.46990921661756746, + "flos": 25228323665280.0, + "grad_norm": 3.827046717276994, + "language_loss": 0.91763622, + "learning_rate": 2.2910768769072603e-06, + "loss": 0.94342971, + "num_input_tokens_seen": 84037930, + "step": 3908, + "time_per_iteration": 2.681093215942383 + }, + { + "auxiliary_loss_clip": 0.01186821, + "auxiliary_loss_mlp": 0.01030042, + "balance_loss_clip": 1.05156362, + "balance_loss_mlp": 1.02209675, + "epoch": 0.47002945950820657, + "flos": 13844031045120.0, + "grad_norm": 2.7611236851994723, + "language_loss": 0.75992364, + "learning_rate": 2.2903061772257417e-06, + "loss": 0.78209227, + "num_input_tokens_seen": 84055915, + "step": 3909, + "time_per_iteration": 3.5736706256866455 + }, + { + "auxiliary_loss_clip": 0.01196487, + "auxiliary_loss_mlp": 0.01029092, + "balance_loss_clip": 1.05531693, + "balance_loss_mlp": 1.02047348, + "epoch": 0.4701497023988457, + "flos": 26247001374720.0, + "grad_norm": 1.521474711261687, + "language_loss": 0.78736842, + "learning_rate": 2.289535433505007e-06, + "loss": 0.8096242, + "num_input_tokens_seen": 84077270, + "step": 3910, + "time_per_iteration": 2.723522424697876 + }, + { + "auxiliary_loss_clip": 0.01204116, + "auxiliary_loss_mlp": 0.01028274, + "balance_loss_clip": 1.05076659, + "balance_loss_mlp": 1.01995301, + "epoch": 0.47026994528948474, + "flos": 25629517647360.0, + "grad_norm": 1.833575960212562, + "language_loss": 0.63910389, + "learning_rate": 2.2887646458619767e-06, + "loss": 0.66142786, + "num_input_tokens_seen": 84098635, + "step": 3911, + "time_per_iteration": 2.749619483947754 + }, + { + "auxiliary_loss_clip": 0.01217858, + "auxiliary_loss_mlp": 0.01025985, + "balance_loss_clip": 1.0515784, + "balance_loss_mlp": 1.01752162, + "epoch": 0.47039018818012385, + "flos": 20554406144640.0, + "grad_norm": 1.7744084436135954, + "language_loss": 0.76501215, + "learning_rate": 2.2879938144135797e-06, + "loss": 0.78745061, + "num_input_tokens_seen": 84114740, + "step": 3912, + "time_per_iteration": 4.403128147125244 + }, + { + "auxiliary_loss_clip": 0.01206772, + "auxiliary_loss_mlp": 0.01378859, + "balance_loss_clip": 1.04741514, + "balance_loss_mlp": 1.00012469, + "epoch": 0.47051043107076296, + "flos": 21577249831680.0, + "grad_norm": 1.5063906745978428, + "language_loss": 0.75234365, + "learning_rate": 2.2872229392767496e-06, + "loss": 0.77819991, + "num_input_tokens_seen": 84134845, + "step": 3913, + "time_per_iteration": 2.777387857437134 + }, + { + "auxiliary_loss_clip": 0.01200784, + "auxiliary_loss_mlp": 0.01024694, + "balance_loss_clip": 1.05619764, + "balance_loss_mlp": 1.01664758, + "epoch": 0.470630673961402, + "flos": 18953185662720.0, + "grad_norm": 1.949048476711205, + "language_loss": 0.74436414, + "learning_rate": 2.286452020568428e-06, + "loss": 0.76661891, + "num_input_tokens_seen": 84152920, + "step": 3914, + "time_per_iteration": 2.628553867340088 + }, + { + "auxiliary_loss_clip": 0.01189952, + "auxiliary_loss_mlp": 0.01024857, + "balance_loss_clip": 1.05454135, + "balance_loss_mlp": 1.01591015, + "epoch": 0.4707509168520411, + "flos": 19938969492480.0, + "grad_norm": 1.6592808985784278, + "language_loss": 0.73235291, + "learning_rate": 2.2856810584055637e-06, + "loss": 0.75450099, + "num_input_tokens_seen": 84170455, + "step": 3915, + "time_per_iteration": 2.6027941703796387 + }, + { + "auxiliary_loss_clip": 0.01198833, + "auxiliary_loss_mlp": 0.01032486, + "balance_loss_clip": 1.05585051, + "balance_loss_mlp": 1.02408218, + "epoch": 0.47087115974268023, + "flos": 40118754741120.0, + "grad_norm": 1.5081859749612958, + "language_loss": 0.67891926, + "learning_rate": 2.2849100529051085e-06, + "loss": 0.70123243, + "num_input_tokens_seen": 84197390, + "step": 3916, + "time_per_iteration": 2.8903160095214844 + }, + { + "auxiliary_loss_clip": 0.01184845, + "auxiliary_loss_mlp": 0.01025725, + "balance_loss_clip": 1.05539346, + "balance_loss_mlp": 1.0176959, + "epoch": 0.4709914026333193, + "flos": 13552723745280.0, + "grad_norm": 2.437392422097086, + "language_loss": 0.80112386, + "learning_rate": 2.284139004184026e-06, + "loss": 0.82322955, + "num_input_tokens_seen": 84214620, + "step": 3917, + "time_per_iteration": 3.483503580093384 + }, + { + "auxiliary_loss_clip": 0.01189199, + "auxiliary_loss_mlp": 0.01030206, + "balance_loss_clip": 1.05774307, + "balance_loss_mlp": 1.02099705, + "epoch": 0.4711116455239584, + "flos": 19974628719360.0, + "grad_norm": 2.102243235771973, + "language_loss": 0.74603295, + "learning_rate": 2.2833679123592814e-06, + "loss": 0.76822698, + "num_input_tokens_seen": 84231880, + "step": 3918, + "time_per_iteration": 2.712010145187378 + }, + { + "auxiliary_loss_clip": 0.0120115, + "auxiliary_loss_mlp": 0.01036037, + "balance_loss_clip": 1.05270731, + "balance_loss_mlp": 1.02742386, + "epoch": 0.4712318884145975, + "flos": 32124824064000.0, + "grad_norm": 1.6111044608121452, + "language_loss": 0.63605702, + "learning_rate": 2.2825967775478508e-06, + "loss": 0.65842891, + "num_input_tokens_seen": 84252980, + "step": 3919, + "time_per_iteration": 2.7660961151123047 + }, + { + "auxiliary_loss_clip": 0.01189384, + "auxiliary_loss_mlp": 0.01032511, + "balance_loss_clip": 1.05672121, + "balance_loss_mlp": 1.0240829, + "epoch": 0.47135213130523657, + "flos": 20047850593920.0, + "grad_norm": 2.1417112185652503, + "language_loss": 0.839414, + "learning_rate": 2.2818255998667135e-06, + "loss": 0.86163294, + "num_input_tokens_seen": 84271490, + "step": 3920, + "time_per_iteration": 2.5765039920806885 + }, + { + "auxiliary_loss_clip": 0.01196335, + "auxiliary_loss_mlp": 0.01031459, + "balance_loss_clip": 1.05549288, + "balance_loss_mlp": 1.02331114, + "epoch": 0.4714723741958757, + "flos": 19426990988160.0, + "grad_norm": 1.5816582323213613, + "language_loss": 0.79183686, + "learning_rate": 2.2810543794328566e-06, + "loss": 0.81411481, + "num_input_tokens_seen": 84290525, + "step": 3921, + "time_per_iteration": 2.7177865505218506 + }, + { + "auxiliary_loss_clip": 0.0119886, + "auxiliary_loss_mlp": 0.01025597, + "balance_loss_clip": 1.05448306, + "balance_loss_mlp": 1.01742542, + "epoch": 0.4715926170865148, + "flos": 20373883367040.0, + "grad_norm": 2.1493263508789724, + "language_loss": 0.82246184, + "learning_rate": 2.2802831163632735e-06, + "loss": 0.84470642, + "num_input_tokens_seen": 84309245, + "step": 3922, + "time_per_iteration": 2.6596460342407227 + }, + { + "auxiliary_loss_clip": 0.01231798, + "auxiliary_loss_mlp": 0.01028739, + "balance_loss_clip": 1.04888213, + "balance_loss_mlp": 1.01973271, + "epoch": 0.47171285997715384, + "flos": 22672884430080.0, + "grad_norm": 1.6213040768207352, + "language_loss": 0.74489474, + "learning_rate": 2.279511810774965e-06, + "loss": 0.76750016, + "num_input_tokens_seen": 84330775, + "step": 3923, + "time_per_iteration": 2.834542751312256 + }, + { + "auxiliary_loss_clip": 0.0118777, + "auxiliary_loss_mlp": 0.0102451, + "balance_loss_clip": 1.05523443, + "balance_loss_mlp": 1.01666594, + "epoch": 0.47183310286779295, + "flos": 21105419754240.0, + "grad_norm": 1.9870450121336884, + "language_loss": 0.714284, + "learning_rate": 2.2787404627849364e-06, + "loss": 0.7364068, + "num_input_tokens_seen": 84349985, + "step": 3924, + "time_per_iteration": 2.642305850982666 + }, + { + "auxiliary_loss_clip": 0.01201264, + "auxiliary_loss_mlp": 0.01024828, + "balance_loss_clip": 1.05212367, + "balance_loss_mlp": 1.01657891, + "epoch": 0.471953345758432, + "flos": 21726566668800.0, + "grad_norm": 1.7670035079306472, + "language_loss": 0.78950548, + "learning_rate": 2.277969072510202e-06, + "loss": 0.81176633, + "num_input_tokens_seen": 84368965, + "step": 3925, + "time_per_iteration": 2.7439370155334473 + }, + { + "auxiliary_loss_clip": 0.0120402, + "auxiliary_loss_mlp": 0.0103062, + "balance_loss_clip": 1.05183673, + "balance_loss_mlp": 1.02253199, + "epoch": 0.4720735886490711, + "flos": 19861078849920.0, + "grad_norm": 1.6640547682762925, + "language_loss": 0.81512868, + "learning_rate": 2.2771976400677803e-06, + "loss": 0.83747506, + "num_input_tokens_seen": 84387795, + "step": 3926, + "time_per_iteration": 2.820774555206299 + }, + { + "auxiliary_loss_clip": 0.01208551, + "auxiliary_loss_mlp": 0.01027029, + "balance_loss_clip": 1.04764366, + "balance_loss_mlp": 1.01904833, + "epoch": 0.47219383153971023, + "flos": 19171809792000.0, + "grad_norm": 2.43072767747785, + "language_loss": 0.78712201, + "learning_rate": 2.2764261655746965e-06, + "loss": 0.80947781, + "num_input_tokens_seen": 84405290, + "step": 3927, + "time_per_iteration": 2.790609359741211 + }, + { + "auxiliary_loss_clip": 0.01208684, + "auxiliary_loss_mlp": 0.0103241, + "balance_loss_clip": 1.0502305, + "balance_loss_mlp": 1.02358818, + "epoch": 0.4723140744303493, + "flos": 23224005780480.0, + "grad_norm": 1.7402412272012913, + "language_loss": 0.75971615, + "learning_rate": 2.2756546491479832e-06, + "loss": 0.78212714, + "num_input_tokens_seen": 84426205, + "step": 3928, + "time_per_iteration": 2.7192537784576416 + }, + { + "auxiliary_loss_clip": 0.01187038, + "auxiliary_loss_mlp": 0.01379066, + "balance_loss_clip": 1.05469263, + "balance_loss_mlp": 1.00015187, + "epoch": 0.4724343173209884, + "flos": 18223265387520.0, + "grad_norm": 2.5083621806811287, + "language_loss": 0.80537343, + "learning_rate": 2.274883090904679e-06, + "loss": 0.83103448, + "num_input_tokens_seen": 84443970, + "step": 3929, + "time_per_iteration": 2.723403215408325 + }, + { + "auxiliary_loss_clip": 0.01193355, + "auxiliary_loss_mlp": 0.01028509, + "balance_loss_clip": 1.06009817, + "balance_loss_mlp": 1.02000988, + "epoch": 0.4725545602116275, + "flos": 21251037490560.0, + "grad_norm": 2.709552304393561, + "language_loss": 0.68042815, + "learning_rate": 2.2741114909618283e-06, + "loss": 0.70264679, + "num_input_tokens_seen": 84459865, + "step": 3930, + "time_per_iteration": 2.588667154312134 + }, + { + "auxiliary_loss_clip": 0.01212687, + "auxiliary_loss_mlp": 0.01026292, + "balance_loss_clip": 1.05201328, + "balance_loss_mlp": 1.01819193, + "epoch": 0.47267480310226656, + "flos": 21434002392960.0, + "grad_norm": 1.676884089787824, + "language_loss": 0.72072697, + "learning_rate": 2.2733398494364828e-06, + "loss": 0.74311674, + "num_input_tokens_seen": 84479110, + "step": 3931, + "time_per_iteration": 2.733689069747925 + }, + { + "auxiliary_loss_clip": 0.01202188, + "auxiliary_loss_mlp": 0.01027681, + "balance_loss_clip": 1.0564388, + "balance_loss_mlp": 1.01934242, + "epoch": 0.47279504599290567, + "flos": 18770508069120.0, + "grad_norm": 2.1533759032540036, + "language_loss": 0.84574497, + "learning_rate": 2.272568166445699e-06, + "loss": 0.8680436, + "num_input_tokens_seen": 84497675, + "step": 3932, + "time_per_iteration": 2.6663224697113037 + }, + { + "auxiliary_loss_clip": 0.01197267, + "auxiliary_loss_mlp": 0.01024337, + "balance_loss_clip": 1.05468595, + "balance_loss_mlp": 1.01578331, + "epoch": 0.4729152888835448, + "flos": 21105742976640.0, + "grad_norm": 3.2323042101765282, + "language_loss": 0.64439201, + "learning_rate": 2.271796442106541e-06, + "loss": 0.66660798, + "num_input_tokens_seen": 84517030, + "step": 3933, + "time_per_iteration": 2.6592607498168945 + }, + { + "auxiliary_loss_clip": 0.01114737, + "auxiliary_loss_mlp": 0.01003928, + "balance_loss_clip": 1.02016091, + "balance_loss_mlp": 1.00267637, + "epoch": 0.47303553177418384, + "flos": 70201877840640.0, + "grad_norm": 0.804680485966181, + "language_loss": 0.56556255, + "learning_rate": 2.271024676536079e-06, + "loss": 0.5867492, + "num_input_tokens_seen": 84577290, + "step": 3934, + "time_per_iteration": 3.2786622047424316 + }, + { + "auxiliary_loss_clip": 0.01213651, + "auxiliary_loss_mlp": 0.01033877, + "balance_loss_clip": 1.05870461, + "balance_loss_mlp": 1.02471018, + "epoch": 0.47315577466482295, + "flos": 22455122227200.0, + "grad_norm": 1.712410456795444, + "language_loss": 0.73383707, + "learning_rate": 2.2702528698513894e-06, + "loss": 0.75631237, + "num_input_tokens_seen": 84598415, + "step": 3935, + "time_per_iteration": 3.598966360092163 + }, + { + "auxiliary_loss_clip": 0.0119998, + "auxiliary_loss_mlp": 0.01027461, + "balance_loss_clip": 1.04837811, + "balance_loss_mlp": 1.01901531, + "epoch": 0.47327601755546206, + "flos": 24352857480960.0, + "grad_norm": 1.7973287829556122, + "language_loss": 0.78625035, + "learning_rate": 2.269481022169554e-06, + "loss": 0.80852485, + "num_input_tokens_seen": 84617010, + "step": 3936, + "time_per_iteration": 2.7467665672302246 + }, + { + "auxiliary_loss_clip": 0.01212328, + "auxiliary_loss_mlp": 0.01025879, + "balance_loss_clip": 1.05349851, + "balance_loss_mlp": 1.01750445, + "epoch": 0.4733962604461011, + "flos": 22926772736640.0, + "grad_norm": 2.3597696363263223, + "language_loss": 0.80190253, + "learning_rate": 2.2687091336076614e-06, + "loss": 0.82428455, + "num_input_tokens_seen": 84636350, + "step": 3937, + "time_per_iteration": 3.653892993927002 + }, + { + "auxiliary_loss_clip": 0.01194404, + "auxiliary_loss_mlp": 0.01029332, + "balance_loss_clip": 1.05435467, + "balance_loss_mlp": 1.02058148, + "epoch": 0.4735165033367402, + "flos": 18327369980160.0, + "grad_norm": 2.489160679757962, + "language_loss": 0.79930592, + "learning_rate": 2.267937204282807e-06, + "loss": 0.82154322, + "num_input_tokens_seen": 84653490, + "step": 3938, + "time_per_iteration": 3.569016695022583 + }, + { + "auxiliary_loss_clip": 0.01202559, + "auxiliary_loss_mlp": 0.0103039, + "balance_loss_clip": 1.05607891, + "balance_loss_mlp": 1.02141929, + "epoch": 0.4736367462273793, + "flos": 23037018554880.0, + "grad_norm": 3.4803686872611745, + "language_loss": 0.79101121, + "learning_rate": 2.2671652343120926e-06, + "loss": 0.81334066, + "num_input_tokens_seen": 84673965, + "step": 3939, + "time_per_iteration": 2.66329288482666 + }, + { + "auxiliary_loss_clip": 0.01187073, + "auxiliary_loss_mlp": 0.01032414, + "balance_loss_clip": 1.05670631, + "balance_loss_mlp": 1.02428412, + "epoch": 0.4737569891180184, + "flos": 25374336451200.0, + "grad_norm": 1.577735475297266, + "language_loss": 0.80473018, + "learning_rate": 2.2663932238126236e-06, + "loss": 0.82692504, + "num_input_tokens_seen": 84692525, + "step": 3940, + "time_per_iteration": 2.679069995880127 + }, + { + "auxiliary_loss_clip": 0.01190932, + "auxiliary_loss_mlp": 0.01030638, + "balance_loss_clip": 1.05123889, + "balance_loss_mlp": 1.02180457, + "epoch": 0.4738772320086575, + "flos": 25849326925440.0, + "grad_norm": 1.37916225923516, + "language_loss": 0.8030268, + "learning_rate": 2.265621172901515e-06, + "loss": 0.82524252, + "num_input_tokens_seen": 84715640, + "step": 3941, + "time_per_iteration": 2.7348101139068604 + }, + { + "auxiliary_loss_clip": 0.01192193, + "auxiliary_loss_mlp": 0.01036108, + "balance_loss_clip": 1.06044984, + "balance_loss_mlp": 1.02764988, + "epoch": 0.47399747489929656, + "flos": 27564420499200.0, + "grad_norm": 2.042463864562623, + "language_loss": 0.71043915, + "learning_rate": 2.2648490816958854e-06, + "loss": 0.73272216, + "num_input_tokens_seen": 84736635, + "step": 3942, + "time_per_iteration": 2.704622268676758 + }, + { + "auxiliary_loss_clip": 0.01192894, + "auxiliary_loss_mlp": 0.01028186, + "balance_loss_clip": 1.05188966, + "balance_loss_mlp": 1.01910806, + "epoch": 0.47411771778993567, + "flos": 24863650836480.0, + "grad_norm": 2.183063000554017, + "language_loss": 0.73425984, + "learning_rate": 2.264076950312861e-06, + "loss": 0.75647056, + "num_input_tokens_seen": 84755445, + "step": 3943, + "time_per_iteration": 3.5991692543029785 + }, + { + "auxiliary_loss_clip": 0.01210555, + "auxiliary_loss_mlp": 0.01029753, + "balance_loss_clip": 1.05379629, + "balance_loss_mlp": 1.02043653, + "epoch": 0.4742379606805748, + "flos": 22748009725440.0, + "grad_norm": 2.100036312524487, + "language_loss": 0.82321, + "learning_rate": 2.2633047788695727e-06, + "loss": 0.84561306, + "num_input_tokens_seen": 84775750, + "step": 3944, + "time_per_iteration": 2.7507524490356445 + }, + { + "auxiliary_loss_clip": 0.01206965, + "auxiliary_loss_mlp": 0.01033423, + "balance_loss_clip": 1.0550344, + "balance_loss_mlp": 1.02468491, + "epoch": 0.47435820357121383, + "flos": 19681130689920.0, + "grad_norm": 1.7551349200526625, + "language_loss": 0.64033568, + "learning_rate": 2.262532567483159e-06, + "loss": 0.66273957, + "num_input_tokens_seen": 84794310, + "step": 3945, + "time_per_iteration": 2.655146837234497 + }, + { + "auxiliary_loss_clip": 0.01191836, + "auxiliary_loss_mlp": 0.01379284, + "balance_loss_clip": 1.05916119, + "balance_loss_mlp": 1.00010455, + "epoch": 0.47447844646185294, + "flos": 25228718714880.0, + "grad_norm": 2.2801040973988136, + "language_loss": 0.80334085, + "learning_rate": 2.2617603162707635e-06, + "loss": 0.82905203, + "num_input_tokens_seen": 84814720, + "step": 3946, + "time_per_iteration": 2.6631171703338623 + }, + { + "auxiliary_loss_clip": 0.01188555, + "auxiliary_loss_mlp": 0.01027968, + "balance_loss_clip": 1.05798876, + "balance_loss_mlp": 1.01940298, + "epoch": 0.47459868935249205, + "flos": 24570619683840.0, + "grad_norm": 2.1442291821824435, + "language_loss": 0.82622117, + "learning_rate": 2.2609880253495363e-06, + "loss": 0.84838641, + "num_input_tokens_seen": 84834355, + "step": 3947, + "time_per_iteration": 2.637608766555786 + }, + { + "auxiliary_loss_clip": 0.01224217, + "auxiliary_loss_mlp": 0.01033188, + "balance_loss_clip": 1.05259454, + "balance_loss_mlp": 1.02474248, + "epoch": 0.4747189322431311, + "flos": 20558500295040.0, + "grad_norm": 1.889461491684202, + "language_loss": 0.86573029, + "learning_rate": 2.260215694836633e-06, + "loss": 0.88830435, + "num_input_tokens_seen": 84853530, + "step": 3948, + "time_per_iteration": 2.802537679672241 + }, + { + "auxiliary_loss_clip": 0.01220416, + "auxiliary_loss_mlp": 0.01379489, + "balance_loss_clip": 1.04614782, + "balance_loss_mlp": 1.00006366, + "epoch": 0.4748391751337702, + "flos": 25995231970560.0, + "grad_norm": 2.075881972540693, + "language_loss": 0.65073204, + "learning_rate": 2.2594433248492157e-06, + "loss": 0.67673117, + "num_input_tokens_seen": 84872505, + "step": 3949, + "time_per_iteration": 2.826805591583252 + }, + { + "auxiliary_loss_clip": 0.0120273, + "auxiliary_loss_mlp": 0.01029017, + "balance_loss_clip": 1.05546355, + "balance_loss_mlp": 1.02055955, + "epoch": 0.47495941802440933, + "flos": 22821052032000.0, + "grad_norm": 1.672247362017447, + "language_loss": 0.80280662, + "learning_rate": 2.2586709155044527e-06, + "loss": 0.82512403, + "num_input_tokens_seen": 84893105, + "step": 3950, + "time_per_iteration": 2.704745292663574 + }, + { + "auxiliary_loss_clip": 0.01189849, + "auxiliary_loss_mlp": 0.01029495, + "balance_loss_clip": 1.05708373, + "balance_loss_mlp": 1.02146006, + "epoch": 0.4750796609150484, + "flos": 27891782075520.0, + "grad_norm": 1.6783668643575047, + "language_loss": 0.75945526, + "learning_rate": 2.2578984669195167e-06, + "loss": 0.78164876, + "num_input_tokens_seen": 84914070, + "step": 3951, + "time_per_iteration": 2.669102907180786 + }, + { + "auxiliary_loss_clip": 0.01192934, + "auxiliary_loss_mlp": 0.01030543, + "balance_loss_clip": 1.05199325, + "balance_loss_mlp": 1.0225172, + "epoch": 0.4751999038056875, + "flos": 35660085471360.0, + "grad_norm": 1.8981050129435553, + "language_loss": 0.68055713, + "learning_rate": 2.2571259792115887e-06, + "loss": 0.70279193, + "num_input_tokens_seen": 84935290, + "step": 3952, + "time_per_iteration": 2.749060869216919 + }, + { + "auxiliary_loss_clip": 0.01192878, + "auxiliary_loss_mlp": 0.0102794, + "balance_loss_clip": 1.05674613, + "balance_loss_mlp": 1.02026856, + "epoch": 0.4753201466963266, + "flos": 22090880361600.0, + "grad_norm": 1.8158815163772777, + "language_loss": 0.79527682, + "learning_rate": 2.2563534524978544e-06, + "loss": 0.81748497, + "num_input_tokens_seen": 84952760, + "step": 3953, + "time_per_iteration": 2.75516676902771 + }, + { + "auxiliary_loss_clip": 0.01211407, + "auxiliary_loss_mlp": 0.01026038, + "balance_loss_clip": 1.05441999, + "balance_loss_mlp": 1.01823568, + "epoch": 0.47544038958696566, + "flos": 30190854965760.0, + "grad_norm": 1.816836657925347, + "language_loss": 0.70745325, + "learning_rate": 2.2555808868955052e-06, + "loss": 0.7298277, + "num_input_tokens_seen": 84974890, + "step": 3954, + "time_per_iteration": 2.7840991020202637 + }, + { + "auxiliary_loss_clip": 0.01223487, + "auxiliary_loss_mlp": 0.01025328, + "balance_loss_clip": 1.05080414, + "balance_loss_mlp": 1.01647031, + "epoch": 0.47556063247760477, + "flos": 23472219738240.0, + "grad_norm": 2.084843932088766, + "language_loss": 0.73823988, + "learning_rate": 2.254808282521738e-06, + "loss": 0.760728, + "num_input_tokens_seen": 84993640, + "step": 3955, + "time_per_iteration": 2.828917980194092 + }, + { + "auxiliary_loss_clip": 0.01214042, + "auxiliary_loss_mlp": 0.01378971, + "balance_loss_clip": 1.04917085, + "balance_loss_mlp": 1.00006676, + "epoch": 0.4756808753682438, + "flos": 25155209531520.0, + "grad_norm": 1.9336657000345157, + "language_loss": 0.81274009, + "learning_rate": 2.2540356394937573e-06, + "loss": 0.83867025, + "num_input_tokens_seen": 85012340, + "step": 3956, + "time_per_iteration": 2.7513856887817383 + }, + { + "auxiliary_loss_clip": 0.01217253, + "auxiliary_loss_mlp": 0.0103051, + "balance_loss_clip": 1.05079889, + "balance_loss_mlp": 1.02170658, + "epoch": 0.47580111825888294, + "flos": 15669729573120.0, + "grad_norm": 2.0307636389791828, + "language_loss": 0.84151632, + "learning_rate": 2.253262957928772e-06, + "loss": 0.86399388, + "num_input_tokens_seen": 85029225, + "step": 3957, + "time_per_iteration": 2.651515483856201 + }, + { + "auxiliary_loss_clip": 0.01193125, + "auxiliary_loss_mlp": 0.01030016, + "balance_loss_clip": 1.04786682, + "balance_loss_mlp": 1.02148652, + "epoch": 0.47592136114952205, + "flos": 17636556637440.0, + "grad_norm": 1.5921981806496168, + "language_loss": 0.72373247, + "learning_rate": 2.2524902379439976e-06, + "loss": 0.74596393, + "num_input_tokens_seen": 85047895, + "step": 3958, + "time_per_iteration": 2.725433588027954 + }, + { + "auxiliary_loss_clip": 0.01136591, + "auxiliary_loss_mlp": 0.01006227, + "balance_loss_clip": 1.0203805, + "balance_loss_mlp": 1.00476098, + "epoch": 0.4760416040401611, + "flos": 61417159292160.0, + "grad_norm": 0.7405076928949336, + "language_loss": 0.63634658, + "learning_rate": 2.251717479656655e-06, + "loss": 0.65777469, + "num_input_tokens_seen": 85112690, + "step": 3959, + "time_per_iteration": 3.439265489578247 + }, + { + "auxiliary_loss_clip": 0.01190745, + "auxiliary_loss_mlp": 0.0103139, + "balance_loss_clip": 1.0570513, + "balance_loss_mlp": 1.02314115, + "epoch": 0.4761618469308002, + "flos": 18405871153920.0, + "grad_norm": 2.7511977525599507, + "language_loss": 0.76537216, + "learning_rate": 2.2509446831839704e-06, + "loss": 0.7875936, + "num_input_tokens_seen": 85132130, + "step": 3960, + "time_per_iteration": 2.8824398517608643 + }, + { + "auxiliary_loss_clip": 0.01208809, + "auxiliary_loss_mlp": 0.01029736, + "balance_loss_clip": 1.05311596, + "balance_loss_mlp": 1.02137339, + "epoch": 0.4762820898214393, + "flos": 18040911016320.0, + "grad_norm": 2.378072911069231, + "language_loss": 0.82353091, + "learning_rate": 2.250171848643177e-06, + "loss": 0.84591639, + "num_input_tokens_seen": 85149420, + "step": 3961, + "time_per_iteration": 3.610095500946045 + }, + { + "auxiliary_loss_clip": 0.01200942, + "auxiliary_loss_mlp": 0.01028565, + "balance_loss_clip": 1.05350447, + "balance_loss_mlp": 1.0204289, + "epoch": 0.4764023327120784, + "flos": 19318253541120.0, + "grad_norm": 1.8307902725825902, + "language_loss": 0.85890031, + "learning_rate": 2.249398976151513e-06, + "loss": 0.88119543, + "num_input_tokens_seen": 85166970, + "step": 3962, + "time_per_iteration": 2.664104461669922 + }, + { + "auxiliary_loss_clip": 0.01184518, + "auxiliary_loss_mlp": 0.01025056, + "balance_loss_clip": 1.05494237, + "balance_loss_mlp": 1.01663041, + "epoch": 0.4765225756027175, + "flos": 22747255539840.0, + "grad_norm": 2.8371742730208465, + "language_loss": 0.7823717, + "learning_rate": 2.248626065826223e-06, + "loss": 0.80446744, + "num_input_tokens_seen": 85185175, + "step": 3963, + "time_per_iteration": 3.9514071941375732 + }, + { + "auxiliary_loss_clip": 0.01081648, + "auxiliary_loss_mlp": 0.01003809, + "balance_loss_clip": 1.02039027, + "balance_loss_mlp": 1.00242615, + "epoch": 0.4766428184933566, + "flos": 65933392106880.0, + "grad_norm": 0.7576928650127284, + "language_loss": 0.62566817, + "learning_rate": 2.2478531177845564e-06, + "loss": 0.6465227, + "num_input_tokens_seen": 85246170, + "step": 3964, + "time_per_iteration": 4.033729076385498 + }, + { + "auxiliary_loss_clip": 0.01210406, + "auxiliary_loss_mlp": 0.0103145, + "balance_loss_clip": 1.0558157, + "balance_loss_mlp": 1.0228852, + "epoch": 0.47676306138399566, + "flos": 24136495908480.0, + "grad_norm": 1.6761149693579391, + "language_loss": 0.85078132, + "learning_rate": 2.247080132143769e-06, + "loss": 0.87319988, + "num_input_tokens_seen": 85268525, + "step": 3965, + "time_per_iteration": 2.775510311126709 + }, + { + "auxiliary_loss_clip": 0.01211091, + "auxiliary_loss_mlp": 0.01035803, + "balance_loss_clip": 1.04827845, + "balance_loss_mlp": 1.02701688, + "epoch": 0.47688330427463477, + "flos": 12604322995200.0, + "grad_norm": 2.3598144007822457, + "language_loss": 0.69229114, + "learning_rate": 2.246307109021121e-06, + "loss": 0.71476007, + "num_input_tokens_seen": 85285930, + "step": 3966, + "time_per_iteration": 2.7246809005737305 + }, + { + "auxiliary_loss_clip": 0.01198297, + "auxiliary_loss_mlp": 0.01028717, + "balance_loss_clip": 1.05005443, + "balance_loss_mlp": 1.02064061, + "epoch": 0.4770035471652739, + "flos": 21390585828480.0, + "grad_norm": 1.720214876602117, + "language_loss": 0.82306051, + "learning_rate": 2.2455340485338817e-06, + "loss": 0.8453306, + "num_input_tokens_seen": 85303565, + "step": 3967, + "time_per_iteration": 2.649258613586426 + }, + { + "auxiliary_loss_clip": 0.01197663, + "auxiliary_loss_mlp": 0.01025646, + "balance_loss_clip": 1.05421638, + "balance_loss_mlp": 1.01764703, + "epoch": 0.47712379005591293, + "flos": 25156251025920.0, + "grad_norm": 1.7167097658109014, + "language_loss": 0.67759925, + "learning_rate": 2.244760950799322e-06, + "loss": 0.69983232, + "num_input_tokens_seen": 85321835, + "step": 3968, + "time_per_iteration": 2.7050259113311768 + }, + { + "auxiliary_loss_clip": 0.01194537, + "auxiliary_loss_mlp": 0.01027156, + "balance_loss_clip": 1.04887843, + "balance_loss_mlp": 1.01913345, + "epoch": 0.47724403294655204, + "flos": 22054323294720.0, + "grad_norm": 2.2778573165756493, + "language_loss": 0.72778553, + "learning_rate": 2.2439878159347203e-06, + "loss": 0.7500025, + "num_input_tokens_seen": 85341260, + "step": 3969, + "time_per_iteration": 3.6221234798431396 + }, + { + "auxiliary_loss_clip": 0.01080525, + "auxiliary_loss_mlp": 0.01001009, + "balance_loss_clip": 1.01944184, + "balance_loss_mlp": 0.99955493, + "epoch": 0.4773642758371911, + "flos": 70229387658240.0, + "grad_norm": 0.7260634283178538, + "language_loss": 0.55294144, + "learning_rate": 2.2432146440573616e-06, + "loss": 0.57375681, + "num_input_tokens_seen": 85407220, + "step": 3970, + "time_per_iteration": 3.2875587940216064 + }, + { + "auxiliary_loss_clip": 0.01208342, + "auxiliary_loss_mlp": 0.01027766, + "balance_loss_clip": 1.05566919, + "balance_loss_mlp": 1.01931977, + "epoch": 0.4774845187278302, + "flos": 23548602009600.0, + "grad_norm": 1.6337820445138225, + "language_loss": 0.66503251, + "learning_rate": 2.242441435284534e-06, + "loss": 0.68739361, + "num_input_tokens_seen": 85426095, + "step": 3971, + "time_per_iteration": 2.6873204708099365 + }, + { + "auxiliary_loss_clip": 0.01197459, + "auxiliary_loss_mlp": 0.0102987, + "balance_loss_clip": 1.05465174, + "balance_loss_mlp": 1.0207026, + "epoch": 0.4776047616184693, + "flos": 23075371301760.0, + "grad_norm": 2.522162245854249, + "language_loss": 0.85265672, + "learning_rate": 2.2416681897335337e-06, + "loss": 0.87493002, + "num_input_tokens_seen": 85444245, + "step": 3972, + "time_per_iteration": 2.624922275543213 + }, + { + "auxiliary_loss_clip": 0.01222156, + "auxiliary_loss_mlp": 0.01033689, + "balance_loss_clip": 1.05257559, + "balance_loss_mlp": 1.02504039, + "epoch": 0.4777250045091084, + "flos": 31898119374720.0, + "grad_norm": 1.784817408023579, + "language_loss": 0.66967154, + "learning_rate": 2.240894907521661e-06, + "loss": 0.69222999, + "num_input_tokens_seen": 85463325, + "step": 3973, + "time_per_iteration": 2.853975534439087 + }, + { + "auxiliary_loss_clip": 0.0120539, + "auxiliary_loss_mlp": 0.01031979, + "balance_loss_clip": 1.05275428, + "balance_loss_mlp": 1.02346754, + "epoch": 0.4778452473997475, + "flos": 24278163148800.0, + "grad_norm": 1.6285735694429637, + "language_loss": 0.63528162, + "learning_rate": 2.240121588766223e-06, + "loss": 0.65765536, + "num_input_tokens_seen": 85483375, + "step": 3974, + "time_per_iteration": 2.771697759628296 + }, + { + "auxiliary_loss_clip": 0.0119667, + "auxiliary_loss_mlp": 0.01026367, + "balance_loss_clip": 1.05112362, + "balance_loss_mlp": 1.01867485, + "epoch": 0.4779654902903866, + "flos": 31575031516800.0, + "grad_norm": 1.8127296364816117, + "language_loss": 0.71490669, + "learning_rate": 2.239348233584531e-06, + "loss": 0.73713696, + "num_input_tokens_seen": 85504230, + "step": 3975, + "time_per_iteration": 2.7348899841308594 + }, + { + "auxiliary_loss_clip": 0.01199813, + "auxiliary_loss_mlp": 0.01027626, + "balance_loss_clip": 1.05476534, + "balance_loss_mlp": 1.01923347, + "epoch": 0.47808573318102565, + "flos": 19500428344320.0, + "grad_norm": 1.7911694351334644, + "language_loss": 0.8065691, + "learning_rate": 2.2385748420939013e-06, + "loss": 0.82884347, + "num_input_tokens_seen": 85523425, + "step": 3976, + "time_per_iteration": 2.6516335010528564 + }, + { + "auxiliary_loss_clip": 0.01185122, + "auxiliary_loss_mlp": 0.01029426, + "balance_loss_clip": 1.05763805, + "balance_loss_mlp": 1.02151036, + "epoch": 0.47820597607166476, + "flos": 22601135013120.0, + "grad_norm": 1.718712730377452, + "language_loss": 0.72704256, + "learning_rate": 2.2378014144116583e-06, + "loss": 0.74918795, + "num_input_tokens_seen": 85542235, + "step": 3977, + "time_per_iteration": 2.647268772125244 + }, + { + "auxiliary_loss_clip": 0.01193932, + "auxiliary_loss_mlp": 0.01027946, + "balance_loss_clip": 1.0593189, + "balance_loss_mlp": 1.01941049, + "epoch": 0.4783262189623039, + "flos": 23003011353600.0, + "grad_norm": 1.830522371201697, + "language_loss": 0.79795146, + "learning_rate": 2.23702795065513e-06, + "loss": 0.82017016, + "num_input_tokens_seen": 85561815, + "step": 3978, + "time_per_iteration": 2.6253790855407715 + }, + { + "auxiliary_loss_clip": 0.01093078, + "auxiliary_loss_mlp": 0.01001411, + "balance_loss_clip": 1.01817346, + "balance_loss_mlp": 0.99995655, + "epoch": 0.47844646185294293, + "flos": 49772801226240.0, + "grad_norm": 0.9871351721153075, + "language_loss": 0.67509449, + "learning_rate": 2.2362544509416493e-06, + "loss": 0.69603938, + "num_input_tokens_seen": 85613930, + "step": 3979, + "time_per_iteration": 3.058589458465576 + }, + { + "auxiliary_loss_clip": 0.01195826, + "auxiliary_loss_mlp": 0.01022557, + "balance_loss_clip": 1.05061769, + "balance_loss_mlp": 1.01467109, + "epoch": 0.47856670474358204, + "flos": 20229558520320.0, + "grad_norm": 2.147668144312348, + "language_loss": 0.82903314, + "learning_rate": 2.2354809153885572e-06, + "loss": 0.85121703, + "num_input_tokens_seen": 85631000, + "step": 3980, + "time_per_iteration": 2.688519239425659 + }, + { + "auxiliary_loss_clip": 0.0119476, + "auxiliary_loss_mlp": 0.01032136, + "balance_loss_clip": 1.05374455, + "balance_loss_mlp": 1.02355909, + "epoch": 0.47868694763422115, + "flos": 20990936131200.0, + "grad_norm": 2.773624800008385, + "language_loss": 0.83086783, + "learning_rate": 2.234707344113197e-06, + "loss": 0.85313678, + "num_input_tokens_seen": 85649095, + "step": 3981, + "time_per_iteration": 2.645754098892212 + }, + { + "auxiliary_loss_clip": 0.01188004, + "auxiliary_loss_mlp": 0.01033111, + "balance_loss_clip": 1.05783355, + "balance_loss_mlp": 1.02505851, + "epoch": 0.4788071905248602, + "flos": 19026551191680.0, + "grad_norm": 2.256695127288715, + "language_loss": 0.77716064, + "learning_rate": 2.233933737232919e-06, + "loss": 0.79937184, + "num_input_tokens_seen": 85666875, + "step": 3982, + "time_per_iteration": 2.629584789276123 + }, + { + "auxiliary_loss_clip": 0.01210523, + "auxiliary_loss_mlp": 0.01379028, + "balance_loss_clip": 1.04663062, + "balance_loss_mlp": 1.00000739, + "epoch": 0.4789274334154993, + "flos": 23002221254400.0, + "grad_norm": 1.6175908930129337, + "language_loss": 0.78382152, + "learning_rate": 2.2331600948650793e-06, + "loss": 0.809717, + "num_input_tokens_seen": 85687020, + "step": 3983, + "time_per_iteration": 2.805901050567627 + }, + { + "auxiliary_loss_clip": 0.01198924, + "auxiliary_loss_mlp": 0.01379901, + "balance_loss_clip": 1.05035639, + "balance_loss_mlp": 1.00009727, + "epoch": 0.4790476763061384, + "flos": 23075586783360.0, + "grad_norm": 1.6400145855169137, + "language_loss": 0.80308604, + "learning_rate": 2.2323864171270386e-06, + "loss": 0.82887435, + "num_input_tokens_seen": 85708290, + "step": 3984, + "time_per_iteration": 2.807756185531616 + }, + { + "auxiliary_loss_clip": 0.01215135, + "auxiliary_loss_mlp": 0.01028076, + "balance_loss_clip": 1.04875827, + "balance_loss_mlp": 1.01920056, + "epoch": 0.4791679191967775, + "flos": 21179288073600.0, + "grad_norm": 2.0764683259899006, + "language_loss": 0.72866297, + "learning_rate": 2.231612704136164e-06, + "loss": 0.75109512, + "num_input_tokens_seen": 85728660, + "step": 3985, + "time_per_iteration": 2.736104965209961 + }, + { + "auxiliary_loss_clip": 0.01189098, + "auxiliary_loss_mlp": 0.01033643, + "balance_loss_clip": 1.05203259, + "balance_loss_mlp": 1.02490461, + "epoch": 0.4792881620874166, + "flos": 22301495758080.0, + "grad_norm": 3.3745988338488067, + "language_loss": 0.75255907, + "learning_rate": 2.2308389560098253e-06, + "loss": 0.77478647, + "num_input_tokens_seen": 85745035, + "step": 3986, + "time_per_iteration": 2.646017074584961 + }, + { + "auxiliary_loss_clip": 0.01216881, + "auxiliary_loss_mlp": 0.01027712, + "balance_loss_clip": 1.0525341, + "balance_loss_mlp": 1.01906919, + "epoch": 0.47940840497805565, + "flos": 17420877423360.0, + "grad_norm": 2.208903680300497, + "language_loss": 0.77137017, + "learning_rate": 2.2300651728654008e-06, + "loss": 0.79381609, + "num_input_tokens_seen": 85760295, + "step": 3987, + "time_per_iteration": 3.7155263423919678 + }, + { + "auxiliary_loss_clip": 0.01086319, + "auxiliary_loss_mlp": 0.01374184, + "balance_loss_clip": 1.01886082, + "balance_loss_mlp": 0.99984664, + "epoch": 0.47952864786869476, + "flos": 65358175708800.0, + "grad_norm": 0.7342146157766606, + "language_loss": 0.60218132, + "learning_rate": 2.229291354820272e-06, + "loss": 0.62678635, + "num_input_tokens_seen": 85821305, + "step": 3988, + "time_per_iteration": 3.215177297592163 + }, + { + "auxiliary_loss_clip": 0.01193956, + "auxiliary_loss_mlp": 0.01028223, + "balance_loss_clip": 1.05334496, + "balance_loss_mlp": 1.01890707, + "epoch": 0.47964889075933387, + "flos": 16799802336000.0, + "grad_norm": 2.128300496322706, + "language_loss": 0.75638789, + "learning_rate": 2.228517501991828e-06, + "loss": 0.77860969, + "num_input_tokens_seen": 85840105, + "step": 3989, + "time_per_iteration": 2.654399871826172 + }, + { + "auxiliary_loss_clip": 0.01101189, + "auxiliary_loss_mlp": 0.01005345, + "balance_loss_clip": 1.02020943, + "balance_loss_mlp": 1.00387311, + "epoch": 0.4797691336499729, + "flos": 70079244808320.0, + "grad_norm": 0.8231372385461946, + "language_loss": 0.61076432, + "learning_rate": 2.22774361449746e-06, + "loss": 0.63182968, + "num_input_tokens_seen": 85896585, + "step": 3990, + "time_per_iteration": 4.956457614898682 + }, + { + "auxiliary_loss_clip": 0.01232046, + "auxiliary_loss_mlp": 0.01030829, + "balance_loss_clip": 1.0485785, + "balance_loss_mlp": 1.02264547, + "epoch": 0.47988937654061203, + "flos": 18953329317120.0, + "grad_norm": 2.537925413508939, + "language_loss": 0.7060712, + "learning_rate": 2.2269696924545668e-06, + "loss": 0.72869998, + "num_input_tokens_seen": 85914415, + "step": 3991, + "time_per_iteration": 2.7462825775146484 + }, + { + "auxiliary_loss_clip": 0.01212781, + "auxiliary_loss_mlp": 0.01031942, + "balance_loss_clip": 1.0533042, + "balance_loss_mlp": 1.02416396, + "epoch": 0.48000961943125114, + "flos": 14461981649280.0, + "grad_norm": 2.2743325867350275, + "language_loss": 0.78152508, + "learning_rate": 2.2261957359805523e-06, + "loss": 0.80397224, + "num_input_tokens_seen": 85931650, + "step": 3992, + "time_per_iteration": 2.7044103145599365 + }, + { + "auxiliary_loss_clip": 0.01190202, + "auxiliary_loss_mlp": 0.01027672, + "balance_loss_clip": 1.05702305, + "balance_loss_mlp": 1.01839781, + "epoch": 0.4801298623218902, + "flos": 27051149105280.0, + "grad_norm": 1.8086487908713962, + "language_loss": 0.73974001, + "learning_rate": 2.225421745192823e-06, + "loss": 0.76191878, + "num_input_tokens_seen": 85951805, + "step": 3993, + "time_per_iteration": 2.6321301460266113 + }, + { + "auxiliary_loss_clip": 0.01194098, + "auxiliary_loss_mlp": 0.0102342, + "balance_loss_clip": 1.05465794, + "balance_loss_mlp": 1.01462245, + "epoch": 0.4802501052125293, + "flos": 26355236031360.0, + "grad_norm": 2.647850263273101, + "language_loss": 0.78180397, + "learning_rate": 2.2246477202087955e-06, + "loss": 0.80397916, + "num_input_tokens_seen": 85972485, + "step": 3994, + "time_per_iteration": 2.7007360458374023 + }, + { + "auxiliary_loss_clip": 0.01205186, + "auxiliary_loss_mlp": 0.01025088, + "balance_loss_clip": 1.05247343, + "balance_loss_mlp": 1.01605749, + "epoch": 0.4803703481031684, + "flos": 20993916960000.0, + "grad_norm": 1.58886417366864, + "language_loss": 0.83182085, + "learning_rate": 2.223873661145887e-06, + "loss": 0.85412365, + "num_input_tokens_seen": 85992540, + "step": 3995, + "time_per_iteration": 3.6473886966705322 + }, + { + "auxiliary_loss_clip": 0.01203223, + "auxiliary_loss_mlp": 0.01378754, + "balance_loss_clip": 1.0545361, + "balance_loss_mlp": 1.0000248, + "epoch": 0.4804905909938075, + "flos": 20703722981760.0, + "grad_norm": 1.543749694183363, + "language_loss": 0.71391106, + "learning_rate": 2.2230995681215226e-06, + "loss": 0.73973083, + "num_input_tokens_seen": 86012065, + "step": 3996, + "time_per_iteration": 2.6710197925567627 + }, + { + "auxiliary_loss_clip": 0.01207252, + "auxiliary_loss_mlp": 0.01024728, + "balance_loss_clip": 1.04908156, + "balance_loss_mlp": 1.0168488, + "epoch": 0.4806108338844466, + "flos": 16654831044480.0, + "grad_norm": 2.422138348966122, + "language_loss": 0.77453429, + "learning_rate": 2.2223254412531305e-06, + "loss": 0.79685414, + "num_input_tokens_seen": 86029435, + "step": 3997, + "time_per_iteration": 2.7243261337280273 + }, + { + "auxiliary_loss_clip": 0.01187663, + "auxiliary_loss_mlp": 0.01022333, + "balance_loss_clip": 1.04904556, + "balance_loss_mlp": 1.01472187, + "epoch": 0.4807310767750857, + "flos": 20011329440640.0, + "grad_norm": 1.7285589610568215, + "language_loss": 0.82179821, + "learning_rate": 2.221551280658146e-06, + "loss": 0.84389818, + "num_input_tokens_seen": 86048495, + "step": 3998, + "time_per_iteration": 2.6611218452453613 + }, + { + "auxiliary_loss_clip": 0.01217552, + "auxiliary_loss_mlp": 0.01027627, + "balance_loss_clip": 1.04870236, + "balance_loss_mlp": 1.01877558, + "epoch": 0.48085131966572475, + "flos": 23185257984000.0, + "grad_norm": 1.6738669009864975, + "language_loss": 0.74391139, + "learning_rate": 2.2207770864540085e-06, + "loss": 0.76636314, + "num_input_tokens_seen": 86067470, + "step": 3999, + "time_per_iteration": 2.786311149597168 + }, + { + "auxiliary_loss_clip": 0.01194013, + "auxiliary_loss_mlp": 0.01021731, + "balance_loss_clip": 1.05103755, + "balance_loss_mlp": 1.01413763, + "epoch": 0.48097156255636386, + "flos": 20558643949440.0, + "grad_norm": 1.8560472378516728, + "language_loss": 0.73239964, + "learning_rate": 2.220002858758162e-06, + "loss": 0.75455707, + "num_input_tokens_seen": 86085460, + "step": 4000, + "time_per_iteration": 2.6336822509765625 + }, + { + "auxiliary_loss_clip": 0.01091269, + "auxiliary_loss_mlp": 0.01000771, + "balance_loss_clip": 1.01818359, + "balance_loss_mlp": 0.99946576, + "epoch": 0.481091805447003, + "flos": 70511608817280.0, + "grad_norm": 0.8774923693143801, + "language_loss": 0.60866898, + "learning_rate": 2.2192285976880573e-06, + "loss": 0.62958944, + "num_input_tokens_seen": 86149715, + "step": 4001, + "time_per_iteration": 3.238496780395508 + }, + { + "auxiliary_loss_clip": 0.0121469, + "auxiliary_loss_mlp": 0.01378506, + "balance_loss_clip": 1.05063593, + "balance_loss_mlp": 1.00005984, + "epoch": 0.48121204833764203, + "flos": 36428214839040.0, + "grad_norm": 1.697096450426754, + "language_loss": 0.81054717, + "learning_rate": 2.2184543033611485e-06, + "loss": 0.83647919, + "num_input_tokens_seen": 86170795, + "step": 4002, + "time_per_iteration": 2.8754026889801025 + }, + { + "auxiliary_loss_clip": 0.01197194, + "auxiliary_loss_mlp": 0.01024953, + "balance_loss_clip": 1.05293202, + "balance_loss_mlp": 1.01690638, + "epoch": 0.48133229122828114, + "flos": 27490264871040.0, + "grad_norm": 2.029948022053734, + "language_loss": 0.81614566, + "learning_rate": 2.2176799758948957e-06, + "loss": 0.83836716, + "num_input_tokens_seen": 86190955, + "step": 4003, + "time_per_iteration": 2.6826062202453613 + }, + { + "auxiliary_loss_clip": 0.01198398, + "auxiliary_loss_mlp": 0.01034715, + "balance_loss_clip": 1.05351686, + "balance_loss_mlp": 1.02650714, + "epoch": 0.4814525341189202, + "flos": 43072802179200.0, + "grad_norm": 1.7924484962428302, + "language_loss": 0.7310046, + "learning_rate": 2.2169056154067635e-06, + "loss": 0.75333577, + "num_input_tokens_seen": 86214875, + "step": 4004, + "time_per_iteration": 2.8711371421813965 + }, + { + "auxiliary_loss_clip": 0.01195304, + "auxiliary_loss_mlp": 0.01379185, + "balance_loss_clip": 1.05523038, + "balance_loss_mlp": 1.00005102, + "epoch": 0.4815727770095593, + "flos": 24236901400320.0, + "grad_norm": 1.6783724062834156, + "language_loss": 0.82333022, + "learning_rate": 2.216131222014222e-06, + "loss": 0.84907508, + "num_input_tokens_seen": 86232950, + "step": 4005, + "time_per_iteration": 2.6278460025787354 + }, + { + "auxiliary_loss_clip": 0.01208093, + "auxiliary_loss_mlp": 0.0102956, + "balance_loss_clip": 1.05131602, + "balance_loss_mlp": 1.02091765, + "epoch": 0.4816930199001984, + "flos": 18113630100480.0, + "grad_norm": 2.1564502982904257, + "language_loss": 0.8036381, + "learning_rate": 2.2153567958347455e-06, + "loss": 0.82601464, + "num_input_tokens_seen": 86249160, + "step": 4006, + "time_per_iteration": 2.7005972862243652 + }, + { + "auxiliary_loss_clip": 0.01203252, + "auxiliary_loss_mlp": 0.0102787, + "balance_loss_clip": 1.05320334, + "balance_loss_mlp": 1.01966882, + "epoch": 0.48181326279083747, + "flos": 17274720983040.0, + "grad_norm": 3.1756232631486356, + "language_loss": 0.79735655, + "learning_rate": 2.214582336985815e-06, + "loss": 0.81966782, + "num_input_tokens_seen": 86267060, + "step": 4007, + "time_per_iteration": 2.641435146331787 + }, + { + "auxiliary_loss_clip": 0.0119426, + "auxiliary_loss_mlp": 0.01034473, + "balance_loss_clip": 1.05128372, + "balance_loss_mlp": 1.02546108, + "epoch": 0.4819335056814766, + "flos": 14903252231040.0, + "grad_norm": 2.6215342947165348, + "language_loss": 0.66028088, + "learning_rate": 2.2138078455849142e-06, + "loss": 0.68256819, + "num_input_tokens_seen": 86285055, + "step": 4008, + "time_per_iteration": 2.670074939727783 + }, + { + "auxiliary_loss_clip": 0.01201191, + "auxiliary_loss_mlp": 0.01031458, + "balance_loss_clip": 1.0548898, + "balance_loss_mlp": 1.02337527, + "epoch": 0.4820537485721157, + "flos": 19244888012160.0, + "grad_norm": 2.024071169521747, + "language_loss": 0.78426749, + "learning_rate": 2.2130333217495334e-06, + "loss": 0.80659395, + "num_input_tokens_seen": 86304225, + "step": 4009, + "time_per_iteration": 2.628000497817993 + }, + { + "auxiliary_loss_clip": 0.0119792, + "auxiliary_loss_mlp": 0.01027693, + "balance_loss_clip": 1.05086315, + "balance_loss_mlp": 1.01920545, + "epoch": 0.48217399146275475, + "flos": 16033791870720.0, + "grad_norm": 2.551426429254802, + "language_loss": 0.67703754, + "learning_rate": 2.2122587655971665e-06, + "loss": 0.69929361, + "num_input_tokens_seen": 86319170, + "step": 4010, + "time_per_iteration": 2.6392502784729004 + }, + { + "auxiliary_loss_clip": 0.01201165, + "auxiliary_loss_mlp": 0.01027344, + "balance_loss_clip": 1.04869199, + "balance_loss_mlp": 1.01912463, + "epoch": 0.48229423435339386, + "flos": 24134197438080.0, + "grad_norm": 1.5238830963253591, + "language_loss": 0.64132857, + "learning_rate": 2.211484177245314e-06, + "loss": 0.66361362, + "num_input_tokens_seen": 86338760, + "step": 4011, + "time_per_iteration": 2.7328648567199707 + }, + { + "auxiliary_loss_clip": 0.01188439, + "auxiliary_loss_mlp": 0.01029077, + "balance_loss_clip": 1.05626225, + "balance_loss_mlp": 1.02050626, + "epoch": 0.48241447724403297, + "flos": 23805435231360.0, + "grad_norm": 2.249692564713587, + "language_loss": 0.71601021, + "learning_rate": 2.21070955681148e-06, + "loss": 0.73818535, + "num_input_tokens_seen": 86357865, + "step": 4012, + "time_per_iteration": 2.6249351501464844 + }, + { + "auxiliary_loss_clip": 0.0120361, + "auxiliary_loss_mlp": 0.01031312, + "balance_loss_clip": 1.05035579, + "balance_loss_mlp": 1.02339697, + "epoch": 0.482534720134672, + "flos": 23110312256640.0, + "grad_norm": 1.5429654307303222, + "language_loss": 0.77784818, + "learning_rate": 2.209934904413174e-06, + "loss": 0.80019742, + "num_input_tokens_seen": 86379470, + "step": 4013, + "time_per_iteration": 3.686549186706543 + }, + { + "auxiliary_loss_clip": 0.01219373, + "auxiliary_loss_mlp": 0.01029036, + "balance_loss_clip": 1.03996408, + "balance_loss_mlp": 1.02014887, + "epoch": 0.48265496302531113, + "flos": 20923819568640.0, + "grad_norm": 3.1982698730553394, + "language_loss": 0.71290529, + "learning_rate": 2.2091602201679095e-06, + "loss": 0.73538935, + "num_input_tokens_seen": 86399080, + "step": 4014, + "time_per_iteration": 2.78442120552063 + }, + { + "auxiliary_loss_clip": 0.0121451, + "auxiliary_loss_mlp": 0.01027345, + "balance_loss_clip": 1.05036879, + "balance_loss_mlp": 1.01943851, + "epoch": 0.48277520591595025, + "flos": 15231152511360.0, + "grad_norm": 2.2998983710140934, + "language_loss": 0.83391929, + "learning_rate": 2.208385504193206e-06, + "loss": 0.85633779, + "num_input_tokens_seen": 86416580, + "step": 4015, + "time_per_iteration": 4.629103422164917 + }, + { + "auxiliary_loss_clip": 0.01186361, + "auxiliary_loss_mlp": 0.01030871, + "balance_loss_clip": 1.05372739, + "balance_loss_mlp": 1.02268076, + "epoch": 0.4828954488065893, + "flos": 17858664385920.0, + "grad_norm": 2.1489642060671987, + "language_loss": 0.81327736, + "learning_rate": 2.2076107566065873e-06, + "loss": 0.8354497, + "num_input_tokens_seen": 86434365, + "step": 4016, + "time_per_iteration": 2.607558488845825 + }, + { + "auxiliary_loss_clip": 0.01196891, + "auxiliary_loss_mlp": 0.01026787, + "balance_loss_clip": 1.0537138, + "balance_loss_mlp": 1.0190506, + "epoch": 0.4830156916972284, + "flos": 32087405070720.0, + "grad_norm": 2.0834038647690964, + "language_loss": 0.75481296, + "learning_rate": 2.2068359775255816e-06, + "loss": 0.77704972, + "num_input_tokens_seen": 86452675, + "step": 4017, + "time_per_iteration": 2.7287943363189697 + }, + { + "auxiliary_loss_clip": 0.01213305, + "auxiliary_loss_mlp": 0.01023054, + "balance_loss_clip": 1.04703927, + "balance_loss_mlp": 1.01531744, + "epoch": 0.48313593458786747, + "flos": 21871717528320.0, + "grad_norm": 2.723973587254907, + "language_loss": 0.78538835, + "learning_rate": 2.206061167067723e-06, + "loss": 0.80775201, + "num_input_tokens_seen": 86470785, + "step": 4018, + "time_per_iteration": 2.808082103729248 + }, + { + "auxiliary_loss_clip": 0.01208207, + "auxiliary_loss_mlp": 0.01026417, + "balance_loss_clip": 1.04752612, + "balance_loss_mlp": 1.01800704, + "epoch": 0.4832561774785066, + "flos": 22601206840320.0, + "grad_norm": 2.294909182388478, + "language_loss": 0.79553688, + "learning_rate": 2.205286325350549e-06, + "loss": 0.81788307, + "num_input_tokens_seen": 86489850, + "step": 4019, + "time_per_iteration": 2.7502071857452393 + }, + { + "auxiliary_loss_clip": 0.01220498, + "auxiliary_loss_mlp": 0.01031995, + "balance_loss_clip": 1.04963207, + "balance_loss_mlp": 1.02276182, + "epoch": 0.4833764203691457, + "flos": 13437342282240.0, + "grad_norm": 1.97380515778772, + "language_loss": 0.72835469, + "learning_rate": 2.204511452491603e-06, + "loss": 0.75087965, + "num_input_tokens_seen": 86506475, + "step": 4020, + "time_per_iteration": 2.7511558532714844 + }, + { + "auxiliary_loss_clip": 0.01184453, + "auxiliary_loss_mlp": 0.01026801, + "balance_loss_clip": 1.05683744, + "balance_loss_mlp": 1.01835549, + "epoch": 0.48349666325978474, + "flos": 44128036955520.0, + "grad_norm": 1.5084749371573944, + "language_loss": 0.74988472, + "learning_rate": 2.2037365486084316e-06, + "loss": 0.77199727, + "num_input_tokens_seen": 86529715, + "step": 4021, + "time_per_iteration": 3.8004562854766846 + }, + { + "auxiliary_loss_clip": 0.01216739, + "auxiliary_loss_mlp": 0.01026404, + "balance_loss_clip": 1.04647899, + "balance_loss_mlp": 1.01832795, + "epoch": 0.48361690615042385, + "flos": 26028377245440.0, + "grad_norm": 1.8437160870287188, + "language_loss": 0.77944672, + "learning_rate": 2.2029616138185886e-06, + "loss": 0.80187809, + "num_input_tokens_seen": 86548715, + "step": 4022, + "time_per_iteration": 2.7321720123291016 + }, + { + "auxiliary_loss_clip": 0.01210862, + "auxiliary_loss_mlp": 0.01030481, + "balance_loss_clip": 1.05410647, + "balance_loss_mlp": 1.02179635, + "epoch": 0.48373714904106296, + "flos": 22273306560000.0, + "grad_norm": 2.555151538521395, + "language_loss": 0.82731223, + "learning_rate": 2.202186648239629e-06, + "loss": 0.8497256, + "num_input_tokens_seen": 86568650, + "step": 4023, + "time_per_iteration": 2.698425531387329 + }, + { + "auxiliary_loss_clip": 0.01193046, + "auxiliary_loss_mlp": 0.01027447, + "balance_loss_clip": 1.05390024, + "balance_loss_mlp": 1.01908445, + "epoch": 0.483857391931702, + "flos": 28292293699200.0, + "grad_norm": 1.7362202598638143, + "language_loss": 0.71475494, + "learning_rate": 2.201411651989117e-06, + "loss": 0.73695982, + "num_input_tokens_seen": 86590630, + "step": 4024, + "time_per_iteration": 2.702458143234253 + }, + { + "auxiliary_loss_clip": 0.01200353, + "auxiliary_loss_mlp": 0.01378819, + "balance_loss_clip": 1.05271769, + "balance_loss_mlp": 0.99999237, + "epoch": 0.48397763482234113, + "flos": 27418048577280.0, + "grad_norm": 6.862433825855763, + "language_loss": 0.78051823, + "learning_rate": 2.2006366251846167e-06, + "loss": 0.80630994, + "num_input_tokens_seen": 86611270, + "step": 4025, + "time_per_iteration": 2.753336191177368 + }, + { + "auxiliary_loss_clip": 0.01203315, + "auxiliary_loss_mlp": 0.01025532, + "balance_loss_clip": 1.05509746, + "balance_loss_mlp": 1.01805472, + "epoch": 0.48409787771298024, + "flos": 16797252470400.0, + "grad_norm": 1.791195939922571, + "language_loss": 0.75770813, + "learning_rate": 2.1998615679436997e-06, + "loss": 0.77999663, + "num_input_tokens_seen": 86628810, + "step": 4026, + "time_per_iteration": 2.640428304672241 + }, + { + "auxiliary_loss_clip": 0.012091, + "auxiliary_loss_mlp": 0.01031947, + "balance_loss_clip": 1.05167007, + "balance_loss_mlp": 1.02385294, + "epoch": 0.4842181206036193, + "flos": 25083496028160.0, + "grad_norm": 4.38156785803652, + "language_loss": 0.76713395, + "learning_rate": 2.199086480383942e-06, + "loss": 0.78954446, + "num_input_tokens_seen": 86648185, + "step": 4027, + "time_per_iteration": 2.7067136764526367 + }, + { + "auxiliary_loss_clip": 0.01212364, + "auxiliary_loss_mlp": 0.01032636, + "balance_loss_clip": 1.0525887, + "balance_loss_mlp": 1.02358222, + "epoch": 0.4843383634942584, + "flos": 30372311496960.0, + "grad_norm": 4.296613325722817, + "language_loss": 0.6763072, + "learning_rate": 2.1983113626229234e-06, + "loss": 0.69875717, + "num_input_tokens_seen": 86667435, + "step": 4028, + "time_per_iteration": 2.7620105743408203 + }, + { + "auxiliary_loss_clip": 0.01203159, + "auxiliary_loss_mlp": 0.01378939, + "balance_loss_clip": 1.04808438, + "balance_loss_mlp": 1.00005579, + "epoch": 0.4844586063848975, + "flos": 20413564917120.0, + "grad_norm": 1.6469740969060458, + "language_loss": 0.78427577, + "learning_rate": 2.1975362147782293e-06, + "loss": 0.81009674, + "num_input_tokens_seen": 86686630, + "step": 4029, + "time_per_iteration": 2.7586495876312256 + }, + { + "auxiliary_loss_clip": 0.01099507, + "auxiliary_loss_mlp": 0.01001815, + "balance_loss_clip": 1.01953959, + "balance_loss_mlp": 1.00040841, + "epoch": 0.48457884927553657, + "flos": 70303722854400.0, + "grad_norm": 0.690061126690983, + "language_loss": 0.54182088, + "learning_rate": 2.196761036967448e-06, + "loss": 0.56283414, + "num_input_tokens_seen": 86754595, + "step": 4030, + "time_per_iteration": 3.3436391353607178 + }, + { + "auxiliary_loss_clip": 0.01188157, + "auxiliary_loss_mlp": 0.01021983, + "balance_loss_clip": 1.05143094, + "balance_loss_mlp": 1.01462793, + "epoch": 0.4846990921661757, + "flos": 19934516206080.0, + "grad_norm": 1.6711270962323121, + "language_loss": 0.77687287, + "learning_rate": 2.1959858293081743e-06, + "loss": 0.79897428, + "num_input_tokens_seen": 86773730, + "step": 4031, + "time_per_iteration": 2.6954457759857178 + }, + { + "auxiliary_loss_clip": 0.0121048, + "auxiliary_loss_mlp": 0.0103123, + "balance_loss_clip": 1.05349123, + "balance_loss_mlp": 1.02351403, + "epoch": 0.4848193350568148, + "flos": 23075945919360.0, + "grad_norm": 1.6019148405361823, + "language_loss": 0.75845945, + "learning_rate": 2.1952105919180056e-06, + "loss": 0.78087652, + "num_input_tokens_seen": 86792985, + "step": 4032, + "time_per_iteration": 2.754845380783081 + }, + { + "auxiliary_loss_clip": 0.01204149, + "auxiliary_loss_mlp": 0.01030818, + "balance_loss_clip": 1.05513263, + "balance_loss_mlp": 1.0228188, + "epoch": 0.48493957794745385, + "flos": 22455481363200.0, + "grad_norm": 2.672710053050955, + "language_loss": 0.67632365, + "learning_rate": 2.1944353249145456e-06, + "loss": 0.69867331, + "num_input_tokens_seen": 86812095, + "step": 4033, + "time_per_iteration": 2.707043409347534 + }, + { + "auxiliary_loss_clip": 0.01185925, + "auxiliary_loss_mlp": 0.01026363, + "balance_loss_clip": 1.05747843, + "balance_loss_mlp": 1.01869214, + "epoch": 0.48505982083809296, + "flos": 25046112948480.0, + "grad_norm": 1.634656841966451, + "language_loss": 0.74724239, + "learning_rate": 2.193660028415401e-06, + "loss": 0.76936531, + "num_input_tokens_seen": 86832875, + "step": 4034, + "time_per_iteration": 2.6452691555023193 + }, + { + "auxiliary_loss_clip": 0.01190963, + "auxiliary_loss_mlp": 0.01033083, + "balance_loss_clip": 1.04960871, + "balance_loss_mlp": 1.02506661, + "epoch": 0.485180063728732, + "flos": 26761386090240.0, + "grad_norm": 1.807776600470654, + "language_loss": 0.82226121, + "learning_rate": 2.1928847025381852e-06, + "loss": 0.84450173, + "num_input_tokens_seen": 86853480, + "step": 4035, + "time_per_iteration": 2.6600446701049805 + }, + { + "auxiliary_loss_clip": 0.0118963, + "auxiliary_loss_mlp": 0.01026512, + "balance_loss_clip": 1.04969382, + "balance_loss_mlp": 1.01809573, + "epoch": 0.4853003066193711, + "flos": 24059143969920.0, + "grad_norm": 1.6663789790666785, + "language_loss": 0.83996052, + "learning_rate": 2.192109347400512e-06, + "loss": 0.86212194, + "num_input_tokens_seen": 86873695, + "step": 4036, + "time_per_iteration": 2.6686551570892334 + }, + { + "auxiliary_loss_clip": 0.01203657, + "auxiliary_loss_mlp": 0.010335, + "balance_loss_clip": 1.05010676, + "balance_loss_mlp": 1.02431488, + "epoch": 0.48542054951001024, + "flos": 23076376882560.0, + "grad_norm": 3.2991944666981627, + "language_loss": 0.79026443, + "learning_rate": 2.191333963120004e-06, + "loss": 0.81263602, + "num_input_tokens_seen": 86892675, + "step": 4037, + "time_per_iteration": 2.6492738723754883 + }, + { + "auxiliary_loss_clip": 0.01200517, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.05020928, + "balance_loss_mlp": 1.01924324, + "epoch": 0.4855407924006493, + "flos": 25664889565440.0, + "grad_norm": 2.128219744048363, + "language_loss": 0.70302814, + "learning_rate": 2.190558549814286e-06, + "loss": 0.72530735, + "num_input_tokens_seen": 86912835, + "step": 4038, + "time_per_iteration": 2.7697198390960693 + }, + { + "auxiliary_loss_clip": 0.01197198, + "auxiliary_loss_mlp": 0.01028655, + "balance_loss_clip": 1.04948711, + "balance_loss_mlp": 1.02051258, + "epoch": 0.4856610352912884, + "flos": 23987933256960.0, + "grad_norm": 1.793407949379741, + "language_loss": 0.79416662, + "learning_rate": 2.1897831076009872e-06, + "loss": 0.81642514, + "num_input_tokens_seen": 86932475, + "step": 4039, + "time_per_iteration": 3.642223834991455 + }, + { + "auxiliary_loss_clip": 0.01193578, + "auxiliary_loss_mlp": 0.01028284, + "balance_loss_clip": 1.05296242, + "balance_loss_mlp": 1.02071452, + "epoch": 0.4857812781819275, + "flos": 24096814358400.0, + "grad_norm": 1.6968537006415914, + "language_loss": 0.79842401, + "learning_rate": 2.1890076365977426e-06, + "loss": 0.82064265, + "num_input_tokens_seen": 86952300, + "step": 4040, + "time_per_iteration": 2.6263539791107178 + }, + { + "auxiliary_loss_clip": 0.01099205, + "auxiliary_loss_mlp": 0.01003799, + "balance_loss_clip": 1.01732326, + "balance_loss_mlp": 1.00272584, + "epoch": 0.48590152107256657, + "flos": 56266635185280.0, + "grad_norm": 0.856228729591407, + "language_loss": 0.52794152, + "learning_rate": 2.188232136922189e-06, + "loss": 0.54897153, + "num_input_tokens_seen": 87010420, + "step": 4041, + "time_per_iteration": 4.987253189086914 + }, + { + "auxiliary_loss_clip": 0.01223129, + "auxiliary_loss_mlp": 0.01027033, + "balance_loss_clip": 1.04676914, + "balance_loss_mlp": 1.0189625, + "epoch": 0.4860217639632057, + "flos": 20046988667520.0, + "grad_norm": 2.262384524001389, + "language_loss": 0.75513095, + "learning_rate": 2.187456608691971e-06, + "loss": 0.77763259, + "num_input_tokens_seen": 87029295, + "step": 4042, + "time_per_iteration": 2.8197128772735596 + }, + { + "auxiliary_loss_clip": 0.01215876, + "auxiliary_loss_mlp": 0.01026544, + "balance_loss_clip": 1.05270052, + "balance_loss_mlp": 1.0185982, + "epoch": 0.4861420068538448, + "flos": 17822143232640.0, + "grad_norm": 1.784289884533091, + "language_loss": 0.87823129, + "learning_rate": 2.1866810520247334e-06, + "loss": 0.90065551, + "num_input_tokens_seen": 87048165, + "step": 4043, + "time_per_iteration": 2.7449793815612793 + }, + { + "auxiliary_loss_clip": 0.01195823, + "auxiliary_loss_mlp": 0.01028694, + "balance_loss_clip": 1.05187154, + "balance_loss_mlp": 1.01996231, + "epoch": 0.48626224974448384, + "flos": 26250125857920.0, + "grad_norm": 1.88667810682891, + "language_loss": 0.64679825, + "learning_rate": 2.185905467038129e-06, + "loss": 0.66904348, + "num_input_tokens_seen": 87067070, + "step": 4044, + "time_per_iteration": 2.6914145946502686 + }, + { + "auxiliary_loss_clip": 0.01183066, + "auxiliary_loss_mlp": 0.01032219, + "balance_loss_clip": 1.05593348, + "balance_loss_mlp": 1.02503061, + "epoch": 0.48638249263512295, + "flos": 22054502862720.0, + "grad_norm": 1.629298175081544, + "language_loss": 0.77358115, + "learning_rate": 2.1851298538498127e-06, + "loss": 0.79573399, + "num_input_tokens_seen": 87086785, + "step": 4045, + "time_per_iteration": 2.711026906967163 + }, + { + "auxiliary_loss_clip": 0.01199822, + "auxiliary_loss_mlp": 0.01379244, + "balance_loss_clip": 1.05504608, + "balance_loss_mlp": 0.99999839, + "epoch": 0.48650273552576206, + "flos": 25119945354240.0, + "grad_norm": 1.9731891536012145, + "language_loss": 0.80151939, + "learning_rate": 2.184354212577446e-06, + "loss": 0.82731009, + "num_input_tokens_seen": 87107090, + "step": 4046, + "time_per_iteration": 2.705065965652466 + }, + { + "auxiliary_loss_clip": 0.01186857, + "auxiliary_loss_mlp": 0.01024958, + "balance_loss_clip": 1.05434465, + "balance_loss_mlp": 1.01651764, + "epoch": 0.4866229784164011, + "flos": 17456931699840.0, + "grad_norm": 3.2661541443212863, + "language_loss": 0.62442249, + "learning_rate": 2.1835785433386907e-06, + "loss": 0.64654064, + "num_input_tokens_seen": 87125905, + "step": 4047, + "time_per_iteration": 3.5088858604431152 + }, + { + "auxiliary_loss_clip": 0.01203397, + "auxiliary_loss_mlp": 0.01043161, + "balance_loss_clip": 1.05270052, + "balance_loss_mlp": 1.03436971, + "epoch": 0.48674322130704023, + "flos": 23331127115520.0, + "grad_norm": 2.188629230459578, + "language_loss": 0.65258968, + "learning_rate": 2.182802846251216e-06, + "loss": 0.67505527, + "num_input_tokens_seen": 87146175, + "step": 4048, + "time_per_iteration": 2.708760976791382 + }, + { + "auxiliary_loss_clip": 0.01211842, + "auxiliary_loss_mlp": 0.01026151, + "balance_loss_clip": 1.04855871, + "balance_loss_mlp": 1.0181936, + "epoch": 0.4868634641976793, + "flos": 28804344030720.0, + "grad_norm": 2.5947255357079277, + "language_loss": 0.72296858, + "learning_rate": 2.182027121432696e-06, + "loss": 0.74534857, + "num_input_tokens_seen": 87166800, + "step": 4049, + "time_per_iteration": 2.7688403129577637 + }, + { + "auxiliary_loss_clip": 0.01187518, + "auxiliary_loss_mlp": 0.01030184, + "balance_loss_clip": 1.05488575, + "balance_loss_mlp": 1.02139223, + "epoch": 0.4869837070883184, + "flos": 19025976574080.0, + "grad_norm": 1.8345304909133686, + "language_loss": 0.82258731, + "learning_rate": 2.1812513690008054e-06, + "loss": 0.84476435, + "num_input_tokens_seen": 87185920, + "step": 4050, + "time_per_iteration": 2.7076950073242188 + }, + { + "auxiliary_loss_clip": 0.01199631, + "auxiliary_loss_mlp": 0.01026529, + "balance_loss_clip": 1.05272567, + "balance_loss_mlp": 1.01831555, + "epoch": 0.4871039499789575, + "flos": 15121409483520.0, + "grad_norm": 2.0758915532105213, + "language_loss": 0.79744381, + "learning_rate": 2.180475589073227e-06, + "loss": 0.81970543, + "num_input_tokens_seen": 87203620, + "step": 4051, + "time_per_iteration": 2.660015344619751 + }, + { + "auxiliary_loss_clip": 0.01181618, + "auxiliary_loss_mlp": 0.01029784, + "balance_loss_clip": 1.05064464, + "balance_loss_mlp": 1.02161181, + "epoch": 0.48722419286959656, + "flos": 26174066808960.0, + "grad_norm": 1.570991456871379, + "language_loss": 0.73414755, + "learning_rate": 2.1796997817676456e-06, + "loss": 0.75626153, + "num_input_tokens_seen": 87224630, + "step": 4052, + "time_per_iteration": 2.80029296875 + }, + { + "auxiliary_loss_clip": 0.01196886, + "auxiliary_loss_mlp": 0.01378546, + "balance_loss_clip": 1.05536962, + "balance_loss_mlp": 1.00001967, + "epoch": 0.4873444357602357, + "flos": 24026142349440.0, + "grad_norm": 1.5766947815651626, + "language_loss": 0.67305022, + "learning_rate": 2.1789239472017494e-06, + "loss": 0.69880456, + "num_input_tokens_seen": 87246280, + "step": 4053, + "time_per_iteration": 2.741328001022339 + }, + { + "auxiliary_loss_clip": 0.01208712, + "auxiliary_loss_mlp": 0.01029237, + "balance_loss_clip": 1.04988325, + "balance_loss_mlp": 1.02164352, + "epoch": 0.4874646786508748, + "flos": 22820441500800.0, + "grad_norm": 1.904974225179331, + "language_loss": 0.7264266, + "learning_rate": 2.1781480854932326e-06, + "loss": 0.74880612, + "num_input_tokens_seen": 87266045, + "step": 4054, + "time_per_iteration": 2.745450735092163 + }, + { + "auxiliary_loss_clip": 0.0122135, + "auxiliary_loss_mlp": 0.01029369, + "balance_loss_clip": 1.05054092, + "balance_loss_mlp": 1.02176976, + "epoch": 0.48758492154151384, + "flos": 21287594557440.0, + "grad_norm": 2.199116782815037, + "language_loss": 0.79285288, + "learning_rate": 2.1773721967597933e-06, + "loss": 0.81536007, + "num_input_tokens_seen": 87284495, + "step": 4055, + "time_per_iteration": 2.719606399536133 + }, + { + "auxiliary_loss_clip": 0.01094924, + "auxiliary_loss_mlp": 0.01006044, + "balance_loss_clip": 1.01867211, + "balance_loss_mlp": 1.00476825, + "epoch": 0.48770516443215295, + "flos": 62244109180800.0, + "grad_norm": 0.8454982804152404, + "language_loss": 0.57395917, + "learning_rate": 2.1765962811191322e-06, + "loss": 0.59496886, + "num_input_tokens_seen": 87338960, + "step": 4056, + "time_per_iteration": 3.1759209632873535 + }, + { + "auxiliary_loss_clip": 0.01132433, + "auxiliary_loss_mlp": 0.01002782, + "balance_loss_clip": 1.02900457, + "balance_loss_mlp": 1.00154269, + "epoch": 0.48782540732279206, + "flos": 66133451882880.0, + "grad_norm": 0.8326160700352394, + "language_loss": 0.6213733, + "learning_rate": 2.1758203386889566e-06, + "loss": 0.64272547, + "num_input_tokens_seen": 87401730, + "step": 4057, + "time_per_iteration": 3.3095898628234863 + }, + { + "auxiliary_loss_clip": 0.01210005, + "auxiliary_loss_mlp": 0.01379086, + "balance_loss_clip": 1.04927659, + "balance_loss_mlp": 1.00006616, + "epoch": 0.4879456502134311, + "flos": 14607922608000.0, + "grad_norm": 1.8435295038057615, + "language_loss": 0.84214836, + "learning_rate": 2.1750443695869746e-06, + "loss": 0.86803925, + "num_input_tokens_seen": 87417300, + "step": 4058, + "time_per_iteration": 2.7194571495056152 + }, + { + "auxiliary_loss_clip": 0.01196839, + "auxiliary_loss_mlp": 0.01029146, + "balance_loss_clip": 1.0530107, + "balance_loss_mlp": 1.02090871, + "epoch": 0.4880658931040702, + "flos": 19500464257920.0, + "grad_norm": 1.67840353232838, + "language_loss": 0.85867202, + "learning_rate": 2.174268373930901e-06, + "loss": 0.88093185, + "num_input_tokens_seen": 87434815, + "step": 4059, + "time_per_iteration": 2.610496997833252 + }, + { + "auxiliary_loss_clip": 0.01203889, + "auxiliary_loss_mlp": 0.01378911, + "balance_loss_clip": 1.05036378, + "balance_loss_mlp": 1.00001097, + "epoch": 0.48818613599470934, + "flos": 16723060928640.0, + "grad_norm": 2.0801904951075443, + "language_loss": 0.79959095, + "learning_rate": 2.1734923518384537e-06, + "loss": 0.82541901, + "num_input_tokens_seen": 87451420, + "step": 4060, + "time_per_iteration": 2.7056198120117188 + }, + { + "auxiliary_loss_clip": 0.01195806, + "auxiliary_loss_mlp": 0.01026427, + "balance_loss_clip": 1.05147159, + "balance_loss_mlp": 1.01805878, + "epoch": 0.4883063788853484, + "flos": 26756932803840.0, + "grad_norm": 8.990649178548908, + "language_loss": 0.82206154, + "learning_rate": 2.1727163034273547e-06, + "loss": 0.84428388, + "num_input_tokens_seen": 87469585, + "step": 4061, + "time_per_iteration": 2.756807565689087 + }, + { + "auxiliary_loss_clip": 0.01194816, + "auxiliary_loss_mlp": 0.01029832, + "balance_loss_clip": 1.0517633, + "balance_loss_mlp": 1.02150559, + "epoch": 0.4884266217759875, + "flos": 16763388923520.0, + "grad_norm": 2.135692142074251, + "language_loss": 0.78630811, + "learning_rate": 2.17194022881533e-06, + "loss": 0.80855465, + "num_input_tokens_seen": 87485675, + "step": 4062, + "time_per_iteration": 2.6087701320648193 + }, + { + "auxiliary_loss_clip": 0.01202126, + "auxiliary_loss_mlp": 0.01027492, + "balance_loss_clip": 1.05034399, + "balance_loss_mlp": 1.01912975, + "epoch": 0.4885468646666266, + "flos": 24207132003840.0, + "grad_norm": 2.0496488411872233, + "language_loss": 0.67352211, + "learning_rate": 2.1711641281201092e-06, + "loss": 0.69581831, + "num_input_tokens_seen": 87505605, + "step": 4063, + "time_per_iteration": 2.716963052749634 + }, + { + "auxiliary_loss_clip": 0.0119465, + "auxiliary_loss_mlp": 0.01027506, + "balance_loss_clip": 1.05479574, + "balance_loss_mlp": 1.0191555, + "epoch": 0.48866710755726567, + "flos": 14610795696000.0, + "grad_norm": 2.402211242382241, + "language_loss": 0.79488128, + "learning_rate": 2.1703880014594264e-06, + "loss": 0.81710285, + "num_input_tokens_seen": 87523195, + "step": 4064, + "time_per_iteration": 2.6127371788024902 + }, + { + "auxiliary_loss_clip": 0.01211906, + "auxiliary_loss_mlp": 0.01033291, + "balance_loss_clip": 1.04949331, + "balance_loss_mlp": 1.02546453, + "epoch": 0.4887873504479048, + "flos": 28804451771520.0, + "grad_norm": 1.9443206042911543, + "language_loss": 0.7380656, + "learning_rate": 2.1696118489510182e-06, + "loss": 0.7605176, + "num_input_tokens_seen": 87544125, + "step": 4065, + "time_per_iteration": 3.7384510040283203 + }, + { + "auxiliary_loss_clip": 0.01216014, + "auxiliary_loss_mlp": 0.01379038, + "balance_loss_clip": 1.05080724, + "balance_loss_mlp": 0.99997985, + "epoch": 0.48890759333854383, + "flos": 22784387224320.0, + "grad_norm": 1.809832150395393, + "language_loss": 0.72162104, + "learning_rate": 2.1688356707126286e-06, + "loss": 0.74757159, + "num_input_tokens_seen": 87563745, + "step": 4066, + "time_per_iteration": 2.7227325439453125 + }, + { + "auxiliary_loss_clip": 0.0120605, + "auxiliary_loss_mlp": 0.01033978, + "balance_loss_clip": 1.05106831, + "balance_loss_mlp": 1.02473307, + "epoch": 0.48902783622918294, + "flos": 17786088956160.0, + "grad_norm": 1.8458617553369894, + "language_loss": 0.69720352, + "learning_rate": 2.168059466862001e-06, + "loss": 0.71960372, + "num_input_tokens_seen": 87581895, + "step": 4067, + "time_per_iteration": 3.6311304569244385 + }, + { + "auxiliary_loss_clip": 0.01202617, + "auxiliary_loss_mlp": 0.01028789, + "balance_loss_clip": 1.04872918, + "balance_loss_mlp": 1.02105832, + "epoch": 0.48914807911982205, + "flos": 22310294590080.0, + "grad_norm": 2.0066941199440675, + "language_loss": 0.81903481, + "learning_rate": 2.167283237516887e-06, + "loss": 0.84134889, + "num_input_tokens_seen": 87600170, + "step": 4068, + "time_per_iteration": 3.5883328914642334 + }, + { + "auxiliary_loss_clip": 0.01204387, + "auxiliary_loss_mlp": 0.0103187, + "balance_loss_clip": 1.05070913, + "balance_loss_mlp": 1.02367401, + "epoch": 0.4892683220104611, + "flos": 16363020954240.0, + "grad_norm": 1.8927564662459002, + "language_loss": 0.74917841, + "learning_rate": 2.1665069827950383e-06, + "loss": 0.771541, + "num_input_tokens_seen": 87617455, + "step": 4069, + "time_per_iteration": 2.693417549133301 + }, + { + "auxiliary_loss_clip": 0.01200501, + "auxiliary_loss_mlp": 0.01027075, + "balance_loss_clip": 1.05177402, + "balance_loss_mlp": 1.0189867, + "epoch": 0.4893885649011002, + "flos": 15739144606080.0, + "grad_norm": 1.9741420707667676, + "language_loss": 0.86440051, + "learning_rate": 2.1657307028142126e-06, + "loss": 0.88667619, + "num_input_tokens_seen": 87634995, + "step": 4070, + "time_per_iteration": 2.6951730251312256 + }, + { + "auxiliary_loss_clip": 0.01202839, + "auxiliary_loss_mlp": 0.01025953, + "balance_loss_clip": 1.0518899, + "balance_loss_mlp": 1.01815033, + "epoch": 0.48950880779173933, + "flos": 28581984887040.0, + "grad_norm": 2.1412896517844593, + "language_loss": 0.67564034, + "learning_rate": 2.164954397692171e-06, + "loss": 0.69792825, + "num_input_tokens_seen": 87654420, + "step": 4071, + "time_per_iteration": 2.725050449371338 + }, + { + "auxiliary_loss_clip": 0.01102074, + "auxiliary_loss_mlp": 0.01003549, + "balance_loss_clip": 1.01579642, + "balance_loss_mlp": 1.00217235, + "epoch": 0.4896290506823784, + "flos": 66186310746240.0, + "grad_norm": 1.472498886739176, + "language_loss": 0.77435112, + "learning_rate": 2.164178067546678e-06, + "loss": 0.79540741, + "num_input_tokens_seen": 87713585, + "step": 4072, + "time_per_iteration": 3.299318313598633 + }, + { + "auxiliary_loss_clip": 0.01208235, + "auxiliary_loss_mlp": 0.01028755, + "balance_loss_clip": 1.04998064, + "balance_loss_mlp": 1.0201, + "epoch": 0.4897492935730175, + "flos": 12531065207040.0, + "grad_norm": 1.8118939305453186, + "language_loss": 0.90700293, + "learning_rate": 2.163401712495504e-06, + "loss": 0.92937279, + "num_input_tokens_seen": 87731280, + "step": 4073, + "time_per_iteration": 3.586562395095825 + }, + { + "auxiliary_loss_clip": 0.01224508, + "auxiliary_loss_mlp": 0.01027114, + "balance_loss_clip": 1.05001342, + "balance_loss_mlp": 1.01860225, + "epoch": 0.4898695364636566, + "flos": 23476816679040.0, + "grad_norm": 1.7143501070846574, + "language_loss": 0.7911377, + "learning_rate": 2.1626253326564194e-06, + "loss": 0.81365395, + "num_input_tokens_seen": 87750230, + "step": 4074, + "time_per_iteration": 2.7047998905181885 + }, + { + "auxiliary_loss_clip": 0.01202587, + "auxiliary_loss_mlp": 0.01030363, + "balance_loss_clip": 1.05095506, + "balance_loss_mlp": 1.02132094, + "epoch": 0.48998977935429566, + "flos": 27160209774720.0, + "grad_norm": 1.7117630259622476, + "language_loss": 0.76886457, + "learning_rate": 2.161848928147201e-06, + "loss": 0.79119408, + "num_input_tokens_seen": 87770500, + "step": 4075, + "time_per_iteration": 2.7957990169525146 + }, + { + "auxiliary_loss_clip": 0.01194808, + "auxiliary_loss_mlp": 0.01033242, + "balance_loss_clip": 1.05475473, + "balance_loss_mlp": 1.02443814, + "epoch": 0.4901100222449348, + "flos": 20339588856960.0, + "grad_norm": 3.0545926256683655, + "language_loss": 0.80387211, + "learning_rate": 2.161072499085629e-06, + "loss": 0.82615262, + "num_input_tokens_seen": 87789495, + "step": 4076, + "time_per_iteration": 2.6655232906341553 + }, + { + "auxiliary_loss_clip": 0.01214092, + "auxiliary_loss_mlp": 0.01034599, + "balance_loss_clip": 1.05103874, + "balance_loss_mlp": 1.02620697, + "epoch": 0.4902302651355739, + "flos": 30446359384320.0, + "grad_norm": 1.7725506956290058, + "language_loss": 0.83046299, + "learning_rate": 2.160296045589487e-06, + "loss": 0.85294986, + "num_input_tokens_seen": 87812955, + "step": 4077, + "time_per_iteration": 2.8435654640197754 + }, + { + "auxiliary_loss_clip": 0.01193929, + "auxiliary_loss_mlp": 0.01028086, + "balance_loss_clip": 1.05476999, + "balance_loss_mlp": 1.01925254, + "epoch": 0.49035050802621294, + "flos": 19174180089600.0, + "grad_norm": 1.7266874139881254, + "language_loss": 0.6990701, + "learning_rate": 2.159519567776562e-06, + "loss": 0.72129023, + "num_input_tokens_seen": 87832605, + "step": 4078, + "time_per_iteration": 2.629971742630005 + }, + { + "auxiliary_loss_clip": 0.01223081, + "auxiliary_loss_mlp": 0.01030335, + "balance_loss_clip": 1.04680467, + "balance_loss_mlp": 1.02218139, + "epoch": 0.49047075091685205, + "flos": 22228489365120.0, + "grad_norm": 3.076420683820157, + "language_loss": 0.71278173, + "learning_rate": 2.1587430657646463e-06, + "loss": 0.73531592, + "num_input_tokens_seen": 87846040, + "step": 4079, + "time_per_iteration": 2.6840767860412598 + }, + { + "auxiliary_loss_clip": 0.01206761, + "auxiliary_loss_mlp": 0.01027346, + "balance_loss_clip": 1.05587971, + "balance_loss_mlp": 1.01924586, + "epoch": 0.4905909938074911, + "flos": 20156516213760.0, + "grad_norm": 2.0114517217032173, + "language_loss": 0.77899945, + "learning_rate": 2.157966539671533e-06, + "loss": 0.80134046, + "num_input_tokens_seen": 87865680, + "step": 4080, + "time_per_iteration": 2.731659173965454 + }, + { + "auxiliary_loss_clip": 0.01216092, + "auxiliary_loss_mlp": 0.010283, + "balance_loss_clip": 1.0510006, + "balance_loss_mlp": 1.02045035, + "epoch": 0.4907112366981302, + "flos": 17202217380480.0, + "grad_norm": 1.765375731464323, + "language_loss": 0.67636889, + "learning_rate": 2.157189989615021e-06, + "loss": 0.69881284, + "num_input_tokens_seen": 87884270, + "step": 4081, + "time_per_iteration": 2.674849271774292 + }, + { + "auxiliary_loss_clip": 0.01197566, + "auxiliary_loss_mlp": 0.01379902, + "balance_loss_clip": 1.05402148, + "balance_loss_mlp": 1.00006318, + "epoch": 0.4908314795887693, + "flos": 21688968107520.0, + "grad_norm": 2.1483197999608374, + "language_loss": 0.75308537, + "learning_rate": 2.156413415712913e-06, + "loss": 0.77886009, + "num_input_tokens_seen": 87906320, + "step": 4082, + "time_per_iteration": 2.6825852394104004 + }, + { + "auxiliary_loss_clip": 0.01209304, + "auxiliary_loss_mlp": 0.01379195, + "balance_loss_clip": 1.05275059, + "balance_loss_mlp": 1.00004911, + "epoch": 0.4909517224794084, + "flos": 26213676531840.0, + "grad_norm": 1.5283164808705194, + "language_loss": 0.78433383, + "learning_rate": 2.155636818083014e-06, + "loss": 0.81021881, + "num_input_tokens_seen": 87927690, + "step": 4083, + "time_per_iteration": 2.6813859939575195 + }, + { + "auxiliary_loss_clip": 0.01196554, + "auxiliary_loss_mlp": 0.0102789, + "balance_loss_clip": 1.0521785, + "balance_loss_mlp": 1.02033198, + "epoch": 0.4910719653700475, + "flos": 23148377694720.0, + "grad_norm": 1.7520665071479924, + "language_loss": 0.84217536, + "learning_rate": 2.154860196843134e-06, + "loss": 0.86441982, + "num_input_tokens_seen": 87946885, + "step": 4084, + "time_per_iteration": 2.6570565700531006 + }, + { + "auxiliary_loss_clip": 0.01186599, + "auxiliary_loss_mlp": 0.01031521, + "balance_loss_clip": 1.05505753, + "balance_loss_mlp": 1.02287841, + "epoch": 0.4911922082606866, + "flos": 23331845387520.0, + "grad_norm": 1.97381130561222, + "language_loss": 0.76638067, + "learning_rate": 2.154083552111085e-06, + "loss": 0.78856182, + "num_input_tokens_seen": 87966055, + "step": 4085, + "time_per_iteration": 2.5868475437164307 + }, + { + "auxiliary_loss_clip": 0.01187577, + "auxiliary_loss_mlp": 0.01033914, + "balance_loss_clip": 1.05424213, + "balance_loss_mlp": 1.02502084, + "epoch": 0.49131245115132566, + "flos": 29203239542400.0, + "grad_norm": 2.204239402316025, + "language_loss": 0.81669545, + "learning_rate": 2.1533068840046834e-06, + "loss": 0.83891034, + "num_input_tokens_seen": 87986320, + "step": 4086, + "time_per_iteration": 2.645230531692505 + }, + { + "auxiliary_loss_clip": 0.01194105, + "auxiliary_loss_mlp": 0.01379533, + "balance_loss_clip": 1.05027342, + "balance_loss_mlp": 1.00002944, + "epoch": 0.49143269404196477, + "flos": 20147465986560.0, + "grad_norm": 2.4042687797998687, + "language_loss": 0.61470103, + "learning_rate": 2.152530192641749e-06, + "loss": 0.64043748, + "num_input_tokens_seen": 88001230, + "step": 4087, + "time_per_iteration": 2.6679961681365967 + }, + { + "auxiliary_loss_clip": 0.01198244, + "auxiliary_loss_mlp": 0.01028993, + "balance_loss_clip": 1.0534122, + "balance_loss_mlp": 1.021119, + "epoch": 0.4915529369326039, + "flos": 24389809597440.0, + "grad_norm": 1.6423627557939, + "language_loss": 0.72361732, + "learning_rate": 2.1517534781401068e-06, + "loss": 0.74588966, + "num_input_tokens_seen": 88019110, + "step": 4088, + "time_per_iteration": 2.7487401962280273 + }, + { + "auxiliary_loss_clip": 0.01195911, + "auxiliary_loss_mlp": 0.01026862, + "balance_loss_clip": 1.0540123, + "balance_loss_mlp": 1.01867282, + "epoch": 0.49167317982324293, + "flos": 10524305197440.0, + "grad_norm": 2.096015685451598, + "language_loss": 0.69681334, + "learning_rate": 2.150976740617581e-06, + "loss": 0.71904111, + "num_input_tokens_seen": 88035670, + "step": 4089, + "time_per_iteration": 2.602532386779785 + }, + { + "auxiliary_loss_clip": 0.01207771, + "auxiliary_loss_mlp": 0.01028316, + "balance_loss_clip": 1.05337501, + "balance_loss_mlp": 1.02074659, + "epoch": 0.49179342271388204, + "flos": 25593427457280.0, + "grad_norm": 2.1655305590203375, + "language_loss": 0.71590698, + "learning_rate": 2.150199980192006e-06, + "loss": 0.7382679, + "num_input_tokens_seen": 88054790, + "step": 4090, + "time_per_iteration": 2.6982667446136475 + }, + { + "auxiliary_loss_clip": 0.01197482, + "auxiliary_loss_mlp": 0.01033968, + "balance_loss_clip": 1.0506289, + "balance_loss_mlp": 1.0259037, + "epoch": 0.49191366560452116, + "flos": 21102043875840.0, + "grad_norm": 1.6805436150578754, + "language_loss": 0.8072055, + "learning_rate": 2.1494231969812114e-06, + "loss": 0.82951999, + "num_input_tokens_seen": 88073780, + "step": 4091, + "time_per_iteration": 3.5481083393096924 + }, + { + "auxiliary_loss_clip": 0.01220131, + "auxiliary_loss_mlp": 0.01025554, + "balance_loss_clip": 1.05360484, + "balance_loss_mlp": 1.01822281, + "epoch": 0.4920339084951602, + "flos": 26067520091520.0, + "grad_norm": 2.4663788877782045, + "language_loss": 0.81128067, + "learning_rate": 2.1486463911030372e-06, + "loss": 0.83373749, + "num_input_tokens_seen": 88094430, + "step": 4092, + "time_per_iteration": 2.69838809967041 + }, + { + "auxiliary_loss_clip": 0.01200349, + "auxiliary_loss_mlp": 0.0103178, + "balance_loss_clip": 1.04994154, + "balance_loss_mlp": 1.02373934, + "epoch": 0.4921541513857993, + "flos": 25081269384960.0, + "grad_norm": 1.7329758263146982, + "language_loss": 0.74445045, + "learning_rate": 2.147869562675324e-06, + "loss": 0.76677167, + "num_input_tokens_seen": 88113400, + "step": 4093, + "time_per_iteration": 3.5853514671325684 + }, + { + "auxiliary_loss_clip": 0.01193611, + "auxiliary_loss_mlp": 0.01026595, + "balance_loss_clip": 1.05336809, + "balance_loss_mlp": 1.01797009, + "epoch": 0.49227439427643843, + "flos": 24389809597440.0, + "grad_norm": 2.6418848890731774, + "language_loss": 0.72412014, + "learning_rate": 2.147092711815915e-06, + "loss": 0.74632221, + "num_input_tokens_seen": 88132750, + "step": 4094, + "time_per_iteration": 3.621061325073242 + }, + { + "auxiliary_loss_clip": 0.01214282, + "auxiliary_loss_mlp": 0.01031976, + "balance_loss_clip": 1.054847, + "balance_loss_mlp": 1.0236969, + "epoch": 0.4923946371670775, + "flos": 11363753018880.0, + "grad_norm": 3.015569938379485, + "language_loss": 0.86746848, + "learning_rate": 2.1463158386426593e-06, + "loss": 0.88993108, + "num_input_tokens_seen": 88150560, + "step": 4095, + "time_per_iteration": 2.6778886318206787 + }, + { + "auxiliary_loss_clip": 0.01209396, + "auxiliary_loss_mlp": 0.01030316, + "balance_loss_clip": 1.05370462, + "balance_loss_mlp": 1.02154839, + "epoch": 0.4925148800577166, + "flos": 30445964334720.0, + "grad_norm": 2.1791231322794387, + "language_loss": 0.77183694, + "learning_rate": 2.145538943273407e-06, + "loss": 0.79423404, + "num_input_tokens_seen": 88170835, + "step": 4096, + "time_per_iteration": 2.7587337493896484 + }, + { + "auxiliary_loss_clip": 0.01191266, + "auxiliary_loss_mlp": 0.01030489, + "balance_loss_clip": 1.05841184, + "balance_loss_mlp": 1.02225208, + "epoch": 0.49263512294835565, + "flos": 20850454039680.0, + "grad_norm": 1.7640585621564882, + "language_loss": 0.71900547, + "learning_rate": 2.144762025826013e-06, + "loss": 0.74122298, + "num_input_tokens_seen": 88189925, + "step": 4097, + "time_per_iteration": 2.5823607444763184 + }, + { + "auxiliary_loss_clip": 0.01199317, + "auxiliary_loss_mlp": 0.01031152, + "balance_loss_clip": 1.0538795, + "balance_loss_mlp": 1.02219367, + "epoch": 0.49275536583899476, + "flos": 23767477534080.0, + "grad_norm": 2.0830297692840993, + "language_loss": 0.86902505, + "learning_rate": 2.143985086418334e-06, + "loss": 0.89132977, + "num_input_tokens_seen": 88205105, + "step": 4098, + "time_per_iteration": 2.728403329849243 + }, + { + "auxiliary_loss_clip": 0.0120868, + "auxiliary_loss_mlp": 0.01027597, + "balance_loss_clip": 1.05244434, + "balance_loss_mlp": 1.01983905, + "epoch": 0.4928756087296339, + "flos": 22273522041600.0, + "grad_norm": 1.3776857968533511, + "language_loss": 0.76558888, + "learning_rate": 2.1432081251682324e-06, + "loss": 0.78795159, + "num_input_tokens_seen": 88225475, + "step": 4099, + "time_per_iteration": 3.6642649173736572 + }, + { + "auxiliary_loss_clip": 0.01195535, + "auxiliary_loss_mlp": 0.01033042, + "balance_loss_clip": 1.05675614, + "balance_loss_mlp": 1.02486467, + "epoch": 0.49299585162027293, + "flos": 19645471463040.0, + "grad_norm": 1.8273621908022266, + "language_loss": 0.8725276, + "learning_rate": 2.142431142193572e-06, + "loss": 0.89481342, + "num_input_tokens_seen": 88243255, + "step": 4100, + "time_per_iteration": 2.5900557041168213 + }, + { + "auxiliary_loss_clip": 0.01183553, + "auxiliary_loss_mlp": 0.01029466, + "balance_loss_clip": 1.05482936, + "balance_loss_mlp": 1.02150929, + "epoch": 0.49311609451091204, + "flos": 38837138497920.0, + "grad_norm": 3.587552756398963, + "language_loss": 0.71459937, + "learning_rate": 2.1416541376122207e-06, + "loss": 0.73672956, + "num_input_tokens_seen": 88263435, + "step": 4101, + "time_per_iteration": 2.7433743476867676 + }, + { + "auxiliary_loss_clip": 0.0118449, + "auxiliary_loss_mlp": 0.01026796, + "balance_loss_clip": 1.05255604, + "balance_loss_mlp": 1.01795065, + "epoch": 0.49323633740155115, + "flos": 28329102161280.0, + "grad_norm": 2.0133103875439775, + "language_loss": 0.72989953, + "learning_rate": 2.1408771115420496e-06, + "loss": 0.75201237, + "num_input_tokens_seen": 88283295, + "step": 4102, + "time_per_iteration": 2.6262929439544678 + }, + { + "auxiliary_loss_clip": 0.01229248, + "auxiliary_loss_mlp": 0.01033305, + "balance_loss_clip": 1.05444741, + "balance_loss_mlp": 1.02495456, + "epoch": 0.4933565802921902, + "flos": 21135584200320.0, + "grad_norm": 1.9394901128646307, + "language_loss": 0.64613557, + "learning_rate": 2.140100064100932e-06, + "loss": 0.66876113, + "num_input_tokens_seen": 88299270, + "step": 4103, + "time_per_iteration": 2.7790772914886475 + }, + { + "auxiliary_loss_clip": 0.01192478, + "auxiliary_loss_mlp": 0.01031802, + "balance_loss_clip": 1.05284178, + "balance_loss_mlp": 1.02336216, + "epoch": 0.4934768231828293, + "flos": 18039007595520.0, + "grad_norm": 2.3110166860001713, + "language_loss": 0.76161474, + "learning_rate": 2.139322995406746e-06, + "loss": 0.78385752, + "num_input_tokens_seen": 88316905, + "step": 4104, + "time_per_iteration": 2.619035243988037 + }, + { + "auxiliary_loss_clip": 0.01189552, + "auxiliary_loss_mlp": 0.01032283, + "balance_loss_clip": 1.05944383, + "balance_loss_mlp": 1.02359867, + "epoch": 0.4935970660734684, + "flos": 23469957181440.0, + "grad_norm": 2.4978626880813173, + "language_loss": 0.7934767, + "learning_rate": 2.1385459055773727e-06, + "loss": 0.81569505, + "num_input_tokens_seen": 88335095, + "step": 4105, + "time_per_iteration": 2.6143131256103516 + }, + { + "auxiliary_loss_clip": 0.01201345, + "auxiliary_loss_mlp": 0.01378859, + "balance_loss_clip": 1.04512393, + "balance_loss_mlp": 1.00007486, + "epoch": 0.4937173089641075, + "flos": 64479258840960.0, + "grad_norm": 2.040283250287074, + "language_loss": 0.74135411, + "learning_rate": 2.137768794730696e-06, + "loss": 0.76715612, + "num_input_tokens_seen": 88358545, + "step": 4106, + "time_per_iteration": 3.1091487407684326 + }, + { + "auxiliary_loss_clip": 0.01209946, + "auxiliary_loss_mlp": 0.01035009, + "balance_loss_clip": 1.05647492, + "balance_loss_mlp": 1.02579391, + "epoch": 0.4938375518547466, + "flos": 22346025644160.0, + "grad_norm": 1.9010934892682936, + "language_loss": 0.80135524, + "learning_rate": 2.1369916629846026e-06, + "loss": 0.82380474, + "num_input_tokens_seen": 88378295, + "step": 4107, + "time_per_iteration": 2.679724931716919 + }, + { + "auxiliary_loss_clip": 0.01202447, + "auxiliary_loss_mlp": 0.01028303, + "balance_loss_clip": 1.05055642, + "balance_loss_mlp": 1.02016664, + "epoch": 0.4939577947453857, + "flos": 17858700299520.0, + "grad_norm": 1.784397066898016, + "language_loss": 0.75278598, + "learning_rate": 2.136214510456983e-06, + "loss": 0.77509344, + "num_input_tokens_seen": 88396750, + "step": 4108, + "time_per_iteration": 2.6857285499572754 + }, + { + "auxiliary_loss_clip": 0.0113326, + "auxiliary_loss_mlp": 0.01374404, + "balance_loss_clip": 1.02028203, + "balance_loss_mlp": 0.99973756, + "epoch": 0.49407803763602476, + "flos": 70066746875520.0, + "grad_norm": 0.9141912857026216, + "language_loss": 0.6312862, + "learning_rate": 2.1354373372657296e-06, + "loss": 0.65636277, + "num_input_tokens_seen": 88455190, + "step": 4109, + "time_per_iteration": 3.4064900875091553 + }, + { + "auxiliary_loss_clip": 0.01187192, + "auxiliary_loss_mlp": 0.01030359, + "balance_loss_clip": 1.05773497, + "balance_loss_mlp": 1.02247906, + "epoch": 0.49419828052666387, + "flos": 24317485562880.0, + "grad_norm": 1.4527617732688187, + "language_loss": 0.70996487, + "learning_rate": 2.1346601435287404e-06, + "loss": 0.73214042, + "num_input_tokens_seen": 88477460, + "step": 4110, + "time_per_iteration": 2.757537841796875 + }, + { + "auxiliary_loss_clip": 0.01201499, + "auxiliary_loss_mlp": 0.01027745, + "balance_loss_clip": 1.04998672, + "balance_loss_mlp": 1.01966858, + "epoch": 0.494318523417303, + "flos": 29386060790400.0, + "grad_norm": 1.7806356754659232, + "language_loss": 0.80038512, + "learning_rate": 2.1338829293639144e-06, + "loss": 0.82267761, + "num_input_tokens_seen": 88497820, + "step": 4111, + "time_per_iteration": 2.7443673610687256 + }, + { + "auxiliary_loss_clip": 0.01217422, + "auxiliary_loss_mlp": 0.01034618, + "balance_loss_clip": 1.04800105, + "balance_loss_mlp": 1.02604699, + "epoch": 0.49443876630794203, + "flos": 15268284195840.0, + "grad_norm": 2.9702893528555014, + "language_loss": 0.82503688, + "learning_rate": 2.1331056948891547e-06, + "loss": 0.84755731, + "num_input_tokens_seen": 88514920, + "step": 4112, + "time_per_iteration": 2.7566633224487305 + }, + { + "auxiliary_loss_clip": 0.01202933, + "auxiliary_loss_mlp": 0.01030629, + "balance_loss_clip": 1.05424213, + "balance_loss_mlp": 1.02164018, + "epoch": 0.49455900919858115, + "flos": 12347453859840.0, + "grad_norm": 2.756435426271533, + "language_loss": 0.76978791, + "learning_rate": 2.1323284402223666e-06, + "loss": 0.7921235, + "num_input_tokens_seen": 88530910, + "step": 4113, + "time_per_iteration": 2.647545576095581 + }, + { + "auxiliary_loss_clip": 0.01185531, + "auxiliary_loss_mlp": 0.01378595, + "balance_loss_clip": 1.05879068, + "balance_loss_mlp": 1.00003648, + "epoch": 0.4946792520892202, + "flos": 22779610715520.0, + "grad_norm": 1.8370313499065385, + "language_loss": 0.88045889, + "learning_rate": 2.1315511654814597e-06, + "loss": 0.90610015, + "num_input_tokens_seen": 88549320, + "step": 4114, + "time_per_iteration": 2.6235127449035645 + }, + { + "auxiliary_loss_clip": 0.01196657, + "auxiliary_loss_mlp": 0.01025699, + "balance_loss_clip": 1.05268955, + "balance_loss_mlp": 1.01787877, + "epoch": 0.4947994949798593, + "flos": 23148126299520.0, + "grad_norm": 1.8929233090596678, + "language_loss": 0.78285384, + "learning_rate": 2.1307738707843456e-06, + "loss": 0.80507731, + "num_input_tokens_seen": 88568985, + "step": 4115, + "time_per_iteration": 2.6435940265655518 + }, + { + "auxiliary_loss_clip": 0.01201793, + "auxiliary_loss_mlp": 0.01032976, + "balance_loss_clip": 1.05583525, + "balance_loss_mlp": 1.02409506, + "epoch": 0.4949197378704984, + "flos": 23659997063040.0, + "grad_norm": 1.916878009706913, + "language_loss": 0.69479543, + "learning_rate": 2.1299965562489385e-06, + "loss": 0.71714312, + "num_input_tokens_seen": 88588790, + "step": 4116, + "time_per_iteration": 2.6589534282684326 + }, + { + "auxiliary_loss_clip": 0.01191149, + "auxiliary_loss_mlp": 0.01023548, + "balance_loss_clip": 1.05237842, + "balance_loss_mlp": 1.01523948, + "epoch": 0.4950399807611375, + "flos": 26911493026560.0, + "grad_norm": 1.3830724448274434, + "language_loss": 0.78926814, + "learning_rate": 2.129219221993158e-06, + "loss": 0.8114152, + "num_input_tokens_seen": 88613575, + "step": 4117, + "time_per_iteration": 2.665494441986084 + }, + { + "auxiliary_loss_clip": 0.01118404, + "auxiliary_loss_mlp": 0.01002651, + "balance_loss_clip": 1.0248127, + "balance_loss_mlp": 1.00137508, + "epoch": 0.4951602236517766, + "flos": 67315270187520.0, + "grad_norm": 0.7839380653714775, + "language_loss": 0.59918964, + "learning_rate": 2.128441868134924e-06, + "loss": 0.62040019, + "num_input_tokens_seen": 88675510, + "step": 4118, + "time_per_iteration": 4.386830806732178 + }, + { + "auxiliary_loss_clip": 0.01212936, + "auxiliary_loss_mlp": 0.01028039, + "balance_loss_clip": 1.04901958, + "balance_loss_mlp": 1.02029693, + "epoch": 0.4952804665424157, + "flos": 19901442758400.0, + "grad_norm": 2.0520577260937083, + "language_loss": 0.82883656, + "learning_rate": 2.1276644947921606e-06, + "loss": 0.85124636, + "num_input_tokens_seen": 88694425, + "step": 4119, + "time_per_iteration": 3.5982189178466797 + }, + { + "auxiliary_loss_clip": 0.01195689, + "auxiliary_loss_mlp": 0.01032755, + "balance_loss_clip": 1.05392325, + "balance_loss_mlp": 1.02377534, + "epoch": 0.49540070943305475, + "flos": 18806813740800.0, + "grad_norm": 2.8706683149201107, + "language_loss": 0.82448876, + "learning_rate": 2.126887102082795e-06, + "loss": 0.84677321, + "num_input_tokens_seen": 88714450, + "step": 4120, + "time_per_iteration": 3.762740135192871 + }, + { + "auxiliary_loss_clip": 0.01204558, + "auxiliary_loss_mlp": 0.01031392, + "balance_loss_clip": 1.04631448, + "balance_loss_mlp": 1.02303553, + "epoch": 0.49552095232369386, + "flos": 24934179191040.0, + "grad_norm": 1.6101132999862995, + "language_loss": 0.70397842, + "learning_rate": 2.126109690124757e-06, + "loss": 0.72633791, + "num_input_tokens_seen": 88735265, + "step": 4121, + "time_per_iteration": 2.741595983505249 + }, + { + "auxiliary_loss_clip": 0.01221536, + "auxiliary_loss_mlp": 0.01031528, + "balance_loss_clip": 1.04919732, + "balance_loss_mlp": 1.0233202, + "epoch": 0.495641195214333, + "flos": 22857249962880.0, + "grad_norm": 1.690297368764054, + "language_loss": 0.71245909, + "learning_rate": 2.1253322590359786e-06, + "loss": 0.73498976, + "num_input_tokens_seen": 88754600, + "step": 4122, + "time_per_iteration": 2.877530097961426 + }, + { + "auxiliary_loss_clip": 0.01187784, + "auxiliary_loss_mlp": 0.01033882, + "balance_loss_clip": 1.05080283, + "balance_loss_mlp": 1.02587676, + "epoch": 0.49576143810497203, + "flos": 25769748343680.0, + "grad_norm": 1.7672441249944542, + "language_loss": 0.73964554, + "learning_rate": 2.124554808934397e-06, + "loss": 0.76186216, + "num_input_tokens_seen": 88775180, + "step": 4123, + "time_per_iteration": 2.663856267929077 + }, + { + "auxiliary_loss_clip": 0.01207151, + "auxiliary_loss_mlp": 0.01030325, + "balance_loss_clip": 1.04375613, + "balance_loss_mlp": 1.0222367, + "epoch": 0.49588168099561114, + "flos": 22128838058880.0, + "grad_norm": 1.846752949688217, + "language_loss": 0.72977376, + "learning_rate": 2.1237773399379496e-06, + "loss": 0.75214857, + "num_input_tokens_seen": 88796145, + "step": 4124, + "time_per_iteration": 2.771059989929199 + }, + { + "auxiliary_loss_clip": 0.01205545, + "auxiliary_loss_mlp": 0.01027802, + "balance_loss_clip": 1.04883623, + "balance_loss_mlp": 1.01968956, + "epoch": 0.49600192388625025, + "flos": 24387331559040.0, + "grad_norm": 1.75654615273294, + "language_loss": 0.86816323, + "learning_rate": 2.122999852164578e-06, + "loss": 0.89049673, + "num_input_tokens_seen": 88816765, + "step": 4125, + "time_per_iteration": 3.6723387241363525 + }, + { + "auxiliary_loss_clip": 0.01216171, + "auxiliary_loss_mlp": 0.01026268, + "balance_loss_clip": 1.0485543, + "balance_loss_mlp": 1.01785159, + "epoch": 0.4961221667768893, + "flos": 22857429530880.0, + "grad_norm": 2.119752038072088, + "language_loss": 0.58555323, + "learning_rate": 2.122222345732227e-06, + "loss": 0.60797763, + "num_input_tokens_seen": 88836680, + "step": 4126, + "time_per_iteration": 2.716444253921509 + }, + { + "auxiliary_loss_clip": 0.0121319, + "auxiliary_loss_mlp": 0.01030045, + "balance_loss_clip": 1.05070245, + "balance_loss_mlp": 1.02091932, + "epoch": 0.4962424096675284, + "flos": 17858089768320.0, + "grad_norm": 2.072453219457427, + "language_loss": 0.83159792, + "learning_rate": 2.121444820758843e-06, + "loss": 0.85403025, + "num_input_tokens_seen": 88855320, + "step": 4127, + "time_per_iteration": 2.714040756225586 + }, + { + "auxiliary_loss_clip": 0.01221252, + "auxiliary_loss_mlp": 0.01031844, + "balance_loss_clip": 1.05188632, + "balance_loss_mlp": 1.02304626, + "epoch": 0.49636265255816747, + "flos": 21793611404160.0, + "grad_norm": 5.606498170784628, + "language_loss": 0.7898519, + "learning_rate": 2.120667277362376e-06, + "loss": 0.81238288, + "num_input_tokens_seen": 88874035, + "step": 4128, + "time_per_iteration": 2.7045094966888428 + }, + { + "auxiliary_loss_clip": 0.01186766, + "auxiliary_loss_mlp": 0.01028445, + "balance_loss_clip": 1.05600691, + "balance_loss_mlp": 1.01991582, + "epoch": 0.4964828954488066, + "flos": 16358603581440.0, + "grad_norm": 2.8572294753512284, + "language_loss": 0.85265487, + "learning_rate": 2.1198897156607796e-06, + "loss": 0.874807, + "num_input_tokens_seen": 88891390, + "step": 4129, + "time_per_iteration": 2.619225263595581 + }, + { + "auxiliary_loss_clip": 0.01196212, + "auxiliary_loss_mlp": 0.01028814, + "balance_loss_clip": 1.05213451, + "balance_loss_mlp": 1.02031422, + "epoch": 0.4966031383394457, + "flos": 24711101775360.0, + "grad_norm": 1.921275795109575, + "language_loss": 0.74370277, + "learning_rate": 2.1191121357720085e-06, + "loss": 0.765953, + "num_input_tokens_seen": 88909450, + "step": 4130, + "time_per_iteration": 2.6443333625793457 + }, + { + "auxiliary_loss_clip": 0.01207711, + "auxiliary_loss_mlp": 0.01030411, + "balance_loss_clip": 1.0448637, + "balance_loss_mlp": 1.02236462, + "epoch": 0.49672338123008475, + "flos": 22930615491840.0, + "grad_norm": 5.6620623687632365, + "language_loss": 0.74590755, + "learning_rate": 2.1183345378140206e-06, + "loss": 0.76828873, + "num_input_tokens_seen": 88929195, + "step": 4131, + "time_per_iteration": 2.7540040016174316 + }, + { + "auxiliary_loss_clip": 0.0108841, + "auxiliary_loss_mlp": 0.01003351, + "balance_loss_clip": 1.0168848, + "balance_loss_mlp": 1.00199151, + "epoch": 0.49684362412072386, + "flos": 65976736844160.0, + "grad_norm": 0.877873550019601, + "language_loss": 0.62018478, + "learning_rate": 2.1175569219047783e-06, + "loss": 0.64110237, + "num_input_tokens_seen": 88990635, + "step": 4132, + "time_per_iteration": 3.316274404525757 + }, + { + "auxiliary_loss_clip": 0.01185088, + "auxiliary_loss_mlp": 0.01030137, + "balance_loss_clip": 1.05440879, + "balance_loss_mlp": 1.02177489, + "epoch": 0.49696386701136297, + "flos": 19971288754560.0, + "grad_norm": 1.7928929227968031, + "language_loss": 0.73356557, + "learning_rate": 2.1167792881622437e-06, + "loss": 0.75571775, + "num_input_tokens_seen": 89009655, + "step": 4133, + "time_per_iteration": 2.592923164367676 + }, + { + "auxiliary_loss_clip": 0.01202551, + "auxiliary_loss_mlp": 0.01031344, + "balance_loss_clip": 1.05494988, + "balance_loss_mlp": 1.02304101, + "epoch": 0.497084109902002, + "flos": 24750819239040.0, + "grad_norm": 1.800783449942406, + "language_loss": 0.80996013, + "learning_rate": 2.116001636704384e-06, + "loss": 0.83229911, + "num_input_tokens_seen": 89030040, + "step": 4134, + "time_per_iteration": 2.9168925285339355 + }, + { + "auxiliary_loss_clip": 0.01226613, + "auxiliary_loss_mlp": 0.01028723, + "balance_loss_clip": 1.04840803, + "balance_loss_mlp": 1.02091742, + "epoch": 0.49720435279264114, + "flos": 21871825269120.0, + "grad_norm": 1.8604904756924816, + "language_loss": 0.8039639, + "learning_rate": 2.1152239676491685e-06, + "loss": 0.82651722, + "num_input_tokens_seen": 89048145, + "step": 4135, + "time_per_iteration": 2.772793769836426 + }, + { + "auxiliary_loss_clip": 0.01204402, + "auxiliary_loss_mlp": 0.01029496, + "balance_loss_clip": 1.0500052, + "balance_loss_mlp": 1.02200413, + "epoch": 0.49732459568328025, + "flos": 23805794367360.0, + "grad_norm": 1.717645925596415, + "language_loss": 0.7326166, + "learning_rate": 2.114446281114569e-06, + "loss": 0.75495559, + "num_input_tokens_seen": 89067165, + "step": 4136, + "time_per_iteration": 2.7236850261688232 + }, + { + "auxiliary_loss_clip": 0.01192082, + "auxiliary_loss_mlp": 0.01028488, + "balance_loss_clip": 1.05005789, + "balance_loss_mlp": 1.02070928, + "epoch": 0.4974448385739193, + "flos": 20047742853120.0, + "grad_norm": 1.8601666324604884, + "language_loss": 0.76008475, + "learning_rate": 2.1136685772185587e-06, + "loss": 0.78229046, + "num_input_tokens_seen": 89086190, + "step": 4137, + "time_per_iteration": 2.6581485271453857 + }, + { + "auxiliary_loss_clip": 0.01201427, + "auxiliary_loss_mlp": 0.01379232, + "balance_loss_clip": 1.04727972, + "balance_loss_mlp": 1.00004411, + "epoch": 0.4975650814645584, + "flos": 24821347593600.0, + "grad_norm": 1.786738174091656, + "language_loss": 0.78049123, + "learning_rate": 2.1128908560791163e-06, + "loss": 0.80629784, + "num_input_tokens_seen": 89106020, + "step": 4138, + "time_per_iteration": 2.6800360679626465 + }, + { + "auxiliary_loss_clip": 0.0118399, + "auxiliary_loss_mlp": 0.01026741, + "balance_loss_clip": 1.05376792, + "balance_loss_mlp": 1.01845646, + "epoch": 0.4976853243551975, + "flos": 19829477859840.0, + "grad_norm": 2.171050278855522, + "language_loss": 0.78199661, + "learning_rate": 2.1121131178142203e-06, + "loss": 0.80410385, + "num_input_tokens_seen": 89125385, + "step": 4139, + "time_per_iteration": 2.5700595378875732 + }, + { + "auxiliary_loss_clip": 0.01200396, + "auxiliary_loss_mlp": 0.01027769, + "balance_loss_clip": 1.04962707, + "balance_loss_mlp": 1.01916194, + "epoch": 0.4978055672458366, + "flos": 23142990654720.0, + "grad_norm": 1.479575902469834, + "language_loss": 0.82424593, + "learning_rate": 2.1113353625418544e-06, + "loss": 0.84652758, + "num_input_tokens_seen": 89143935, + "step": 4140, + "time_per_iteration": 2.680753469467163 + }, + { + "auxiliary_loss_clip": 0.01187534, + "auxiliary_loss_mlp": 0.01031295, + "balance_loss_clip": 1.05490685, + "balance_loss_mlp": 1.02386272, + "epoch": 0.4979258101364757, + "flos": 15559914718080.0, + "grad_norm": 1.654214703447195, + "language_loss": 0.78865451, + "learning_rate": 2.1105575903800017e-06, + "loss": 0.81084275, + "num_input_tokens_seen": 89162655, + "step": 4141, + "time_per_iteration": 2.5808866024017334 + }, + { + "auxiliary_loss_clip": 0.0119674, + "auxiliary_loss_mlp": 0.01034608, + "balance_loss_clip": 1.05170691, + "balance_loss_mlp": 1.02684188, + "epoch": 0.4980460530271148, + "flos": 26356169784960.0, + "grad_norm": 1.8904107766289813, + "language_loss": 0.85143882, + "learning_rate": 2.1097798014466502e-06, + "loss": 0.87375236, + "num_input_tokens_seen": 89182255, + "step": 4142, + "time_per_iteration": 2.6865124702453613 + }, + { + "auxiliary_loss_clip": 0.01197522, + "auxiliary_loss_mlp": 0.01029448, + "balance_loss_clip": 1.05312204, + "balance_loss_mlp": 1.02098441, + "epoch": 0.49816629591775385, + "flos": 17274541415040.0, + "grad_norm": 2.1095995699638346, + "language_loss": 0.58818245, + "learning_rate": 2.109001995859791e-06, + "loss": 0.61045212, + "num_input_tokens_seen": 89201155, + "step": 4143, + "time_per_iteration": 3.584520101547241 + }, + { + "auxiliary_loss_clip": 0.01099527, + "auxiliary_loss_mlp": 0.01003867, + "balance_loss_clip": 1.0180074, + "balance_loss_mlp": 1.00261521, + "epoch": 0.49828653880839296, + "flos": 64930947344640.0, + "grad_norm": 0.7950445490191317, + "language_loss": 0.6006965, + "learning_rate": 2.108224173737415e-06, + "loss": 0.62173045, + "num_input_tokens_seen": 89264455, + "step": 4144, + "time_per_iteration": 3.199357748031616 + }, + { + "auxiliary_loss_clip": 0.01197584, + "auxiliary_loss_mlp": 0.0103159, + "balance_loss_clip": 1.04821706, + "balance_loss_mlp": 1.02312338, + "epoch": 0.498406781699032, + "flos": 27484806003840.0, + "grad_norm": 2.049056827073844, + "language_loss": 0.7616325, + "learning_rate": 2.1074463351975183e-06, + "loss": 0.78392422, + "num_input_tokens_seen": 89283340, + "step": 4145, + "time_per_iteration": 4.578942775726318 + }, + { + "auxiliary_loss_clip": 0.01215951, + "auxiliary_loss_mlp": 0.01029614, + "balance_loss_clip": 1.04998302, + "balance_loss_mlp": 1.0222466, + "epoch": 0.49852702458967113, + "flos": 31499870307840.0, + "grad_norm": 3.7039765845465937, + "language_loss": 0.71319455, + "learning_rate": 2.106668480358098e-06, + "loss": 0.73565024, + "num_input_tokens_seen": 89303565, + "step": 4146, + "time_per_iteration": 2.757716178894043 + }, + { + "auxiliary_loss_clip": 0.01219499, + "auxiliary_loss_mlp": 0.010287, + "balance_loss_clip": 1.04831815, + "balance_loss_mlp": 1.01995885, + "epoch": 0.49864726748031024, + "flos": 22852868503680.0, + "grad_norm": 1.9162188558935542, + "language_loss": 0.71283138, + "learning_rate": 2.105890609337154e-06, + "loss": 0.73531342, + "num_input_tokens_seen": 89322080, + "step": 4147, + "time_per_iteration": 2.6820857524871826 + }, + { + "auxiliary_loss_clip": 0.01076738, + "auxiliary_loss_mlp": 0.01003526, + "balance_loss_clip": 1.01825309, + "balance_loss_mlp": 1.00227439, + "epoch": 0.4987675103709493, + "flos": 70405708544640.0, + "grad_norm": 0.7001097478581746, + "language_loss": 0.63798225, + "learning_rate": 2.1051127222526883e-06, + "loss": 0.65878487, + "num_input_tokens_seen": 89394195, + "step": 4148, + "time_per_iteration": 3.2663002014160156 + }, + { + "auxiliary_loss_clip": 0.01190768, + "auxiliary_loss_mlp": 0.01029111, + "balance_loss_clip": 1.05467534, + "balance_loss_mlp": 1.0213685, + "epoch": 0.4988877532615884, + "flos": 28767571482240.0, + "grad_norm": 1.6969694638616297, + "language_loss": 0.80918145, + "learning_rate": 2.1043348192227067e-06, + "loss": 0.83138031, + "num_input_tokens_seen": 89414565, + "step": 4149, + "time_per_iteration": 2.670266628265381 + }, + { + "auxiliary_loss_clip": 0.01197183, + "auxiliary_loss_mlp": 0.01030017, + "balance_loss_clip": 1.0488627, + "balance_loss_mlp": 1.02228642, + "epoch": 0.4990079961522275, + "flos": 16872700988160.0, + "grad_norm": 1.900913773424562, + "language_loss": 0.61869848, + "learning_rate": 2.1035569003652156e-06, + "loss": 0.64097047, + "num_input_tokens_seen": 89433195, + "step": 4150, + "time_per_iteration": 2.653658390045166 + }, + { + "auxiliary_loss_clip": 0.01213671, + "auxiliary_loss_mlp": 0.01031495, + "balance_loss_clip": 1.04665446, + "balance_loss_mlp": 1.02172017, + "epoch": 0.4991282390428666, + "flos": 13291042187520.0, + "grad_norm": 1.9240670295602251, + "language_loss": 0.82155621, + "learning_rate": 2.1027789657982255e-06, + "loss": 0.84400785, + "num_input_tokens_seen": 89447410, + "step": 4151, + "time_per_iteration": 3.5947303771972656 + }, + { + "auxiliary_loss_clip": 0.01221058, + "auxiliary_loss_mlp": 0.01028331, + "balance_loss_clip": 1.04985797, + "balance_loss_mlp": 1.02049327, + "epoch": 0.4992484819335057, + "flos": 21537496454400.0, + "grad_norm": 2.130740952887998, + "language_loss": 0.77365756, + "learning_rate": 2.1020010156397482e-06, + "loss": 0.7961514, + "num_input_tokens_seen": 89464630, + "step": 4152, + "time_per_iteration": 2.8064076900482178 + }, + { + "auxiliary_loss_clip": 0.01195771, + "auxiliary_loss_mlp": 0.01028312, + "balance_loss_clip": 1.0539099, + "balance_loss_mlp": 1.01981878, + "epoch": 0.4993687248241448, + "flos": 24860095390080.0, + "grad_norm": 1.537147154043513, + "language_loss": 0.77546537, + "learning_rate": 2.101223050007797e-06, + "loss": 0.79770625, + "num_input_tokens_seen": 89483180, + "step": 4153, + "time_per_iteration": 2.7286603450775146 + }, + { + "auxiliary_loss_clip": 0.01074201, + "auxiliary_loss_mlp": 0.01002961, + "balance_loss_clip": 1.01575398, + "balance_loss_mlp": 1.00170934, + "epoch": 0.49948896771478385, + "flos": 62941602453120.0, + "grad_norm": 0.8264342144769382, + "language_loss": 0.53817439, + "learning_rate": 2.1004450690203904e-06, + "loss": 0.55894601, + "num_input_tokens_seen": 89539260, + "step": 4154, + "time_per_iteration": 3.2810051441192627 + }, + { + "auxiliary_loss_clip": 0.01073761, + "auxiliary_loss_mlp": 0.01001636, + "balance_loss_clip": 1.01540554, + "balance_loss_mlp": 1.00034285, + "epoch": 0.49960921060542296, + "flos": 68284213516800.0, + "grad_norm": 0.8735426965207918, + "language_loss": 0.63309169, + "learning_rate": 2.099667072795546e-06, + "loss": 0.65384567, + "num_input_tokens_seen": 89601380, + "step": 4155, + "time_per_iteration": 3.2312095165252686 + }, + { + "auxiliary_loss_clip": 0.01192438, + "auxiliary_loss_mlp": 0.01029005, + "balance_loss_clip": 1.05081105, + "balance_loss_mlp": 1.02066076, + "epoch": 0.49972945349606207, + "flos": 23659350618240.0, + "grad_norm": 2.7617982560022623, + "language_loss": 0.79551947, + "learning_rate": 2.0988890614512864e-06, + "loss": 0.81773388, + "num_input_tokens_seen": 89621270, + "step": 4156, + "time_per_iteration": 2.8203470706939697 + }, + { + "auxiliary_loss_clip": 0.01204121, + "auxiliary_loss_mlp": 0.01027448, + "balance_loss_clip": 1.05314136, + "balance_loss_mlp": 1.01897788, + "epoch": 0.4998496963867011, + "flos": 19755825022080.0, + "grad_norm": 2.0277482728909866, + "language_loss": 0.8398459, + "learning_rate": 2.098111035105635e-06, + "loss": 0.86216152, + "num_input_tokens_seen": 89639695, + "step": 4157, + "time_per_iteration": 2.636061906814575 + }, + { + "auxiliary_loss_clip": 0.0121834, + "auxiliary_loss_mlp": 0.01032221, + "balance_loss_clip": 1.05151582, + "balance_loss_mlp": 1.02353024, + "epoch": 0.49996993927734024, + "flos": 22265728790400.0, + "grad_norm": 1.7392748049271256, + "language_loss": 0.73203611, + "learning_rate": 2.0973329938766176e-06, + "loss": 0.75454175, + "num_input_tokens_seen": 89657125, + "step": 4158, + "time_per_iteration": 2.744764566421509 + }, + { + "auxiliary_loss_clip": 0.01198679, + "auxiliary_loss_mlp": 0.01027423, + "balance_loss_clip": 1.05398715, + "balance_loss_mlp": 1.01832771, + "epoch": 0.5000901821679793, + "flos": 23327212533120.0, + "grad_norm": 1.8788892006732707, + "language_loss": 0.7884469, + "learning_rate": 2.0965549378822618e-06, + "loss": 0.81070787, + "num_input_tokens_seen": 89678415, + "step": 4159, + "time_per_iteration": 2.6801350116729736 + }, + { + "auxiliary_loss_clip": 0.01246315, + "auxiliary_loss_mlp": 0.01026711, + "balance_loss_clip": 1.04620957, + "balance_loss_mlp": 1.01802111, + "epoch": 0.5002104250586185, + "flos": 20339014239360.0, + "grad_norm": 1.9951874797360132, + "language_loss": 0.83935165, + "learning_rate": 2.095776867240599e-06, + "loss": 0.86208194, + "num_input_tokens_seen": 89695405, + "step": 4160, + "time_per_iteration": 3.092473030090332 + }, + { + "auxiliary_loss_clip": 0.01203586, + "auxiliary_loss_mlp": 0.01031443, + "balance_loss_clip": 1.04914165, + "balance_loss_mlp": 1.02331901, + "epoch": 0.5003306679492575, + "flos": 13991372634240.0, + "grad_norm": 2.1669098138689993, + "language_loss": 0.82978964, + "learning_rate": 2.094998782069661e-06, + "loss": 0.85213995, + "num_input_tokens_seen": 89713110, + "step": 4161, + "time_per_iteration": 2.9164085388183594 + }, + { + "auxiliary_loss_clip": 0.01184517, + "auxiliary_loss_mlp": 0.01031715, + "balance_loss_clip": 1.05521703, + "balance_loss_mlp": 1.02326345, + "epoch": 0.5004509108398966, + "flos": 27672762896640.0, + "grad_norm": 1.6730589091141241, + "language_loss": 0.75457203, + "learning_rate": 2.0942206824874845e-06, + "loss": 0.77673429, + "num_input_tokens_seen": 89735885, + "step": 4162, + "time_per_iteration": 2.6426703929901123 + }, + { + "auxiliary_loss_clip": 0.01189498, + "auxiliary_loss_mlp": 0.01028186, + "balance_loss_clip": 1.05141079, + "balance_loss_mlp": 1.01951373, + "epoch": 0.5005711537305357, + "flos": 14976186796800.0, + "grad_norm": 2.271569678709389, + "language_loss": 0.79489672, + "learning_rate": 2.093442568612105e-06, + "loss": 0.81707358, + "num_input_tokens_seen": 89753690, + "step": 4163, + "time_per_iteration": 2.720627546310425 + }, + { + "auxiliary_loss_clip": 0.01183597, + "auxiliary_loss_mlp": 0.01025582, + "balance_loss_clip": 1.05204463, + "balance_loss_mlp": 1.01748741, + "epoch": 0.5006913966211748, + "flos": 26503259978880.0, + "grad_norm": 2.1530586056419203, + "language_loss": 0.8540687, + "learning_rate": 2.0926644405615613e-06, + "loss": 0.8761605, + "num_input_tokens_seen": 89774590, + "step": 4164, + "time_per_iteration": 2.6350209712982178 + }, + { + "auxiliary_loss_clip": 0.01208113, + "auxiliary_loss_mlp": 0.01028027, + "balance_loss_clip": 1.05168915, + "balance_loss_mlp": 1.02005756, + "epoch": 0.5008116395118138, + "flos": 20449295971200.0, + "grad_norm": 1.6802859809808044, + "language_loss": 0.81371111, + "learning_rate": 2.091886298453897e-06, + "loss": 0.8360725, + "num_input_tokens_seen": 89792775, + "step": 4165, + "time_per_iteration": 2.728499412536621 + }, + { + "auxiliary_loss_clip": 0.01188711, + "auxiliary_loss_mlp": 0.01031253, + "balance_loss_clip": 1.05061209, + "balance_loss_mlp": 1.02318215, + "epoch": 0.500931882402453, + "flos": 21579871524480.0, + "grad_norm": 1.7534048759308924, + "language_loss": 0.7276119, + "learning_rate": 2.091108142407153e-06, + "loss": 0.74981153, + "num_input_tokens_seen": 89811515, + "step": 4166, + "time_per_iteration": 2.6581478118896484 + }, + { + "auxiliary_loss_clip": 0.01101446, + "auxiliary_loss_mlp": 0.0100204, + "balance_loss_clip": 1.0208776, + "balance_loss_mlp": 1.00077057, + "epoch": 0.5010521252930921, + "flos": 57785011925760.0, + "grad_norm": 0.8470511922342652, + "language_loss": 0.62414753, + "learning_rate": 2.090329972539377e-06, + "loss": 0.64518237, + "num_input_tokens_seen": 89870080, + "step": 4167, + "time_per_iteration": 3.3752589225769043 + }, + { + "auxiliary_loss_clip": 0.01236458, + "auxiliary_loss_mlp": 0.01028327, + "balance_loss_clip": 1.04392862, + "balance_loss_mlp": 1.02000666, + "epoch": 0.5011723681837311, + "flos": 18625500864000.0, + "grad_norm": 1.8439223356787446, + "language_loss": 0.68728232, + "learning_rate": 2.089551788968616e-06, + "loss": 0.70993012, + "num_input_tokens_seen": 89888045, + "step": 4168, + "time_per_iteration": 3.0597009658813477 + }, + { + "auxiliary_loss_clip": 0.01072595, + "auxiliary_loss_mlp": 0.01001473, + "balance_loss_clip": 1.01423764, + "balance_loss_mlp": 1.00013757, + "epoch": 0.5012926110743702, + "flos": 55883146608000.0, + "grad_norm": 0.8369582570301688, + "language_loss": 0.60759568, + "learning_rate": 2.08877359181292e-06, + "loss": 0.62833631, + "num_input_tokens_seen": 89944610, + "step": 4169, + "time_per_iteration": 3.863163948059082 + }, + { + "auxiliary_loss_clip": 0.01212087, + "auxiliary_loss_mlp": 0.01028191, + "balance_loss_clip": 1.04635227, + "balance_loss_mlp": 1.02011466, + "epoch": 0.5014128539650093, + "flos": 24238266117120.0, + "grad_norm": 2.876812507086548, + "language_loss": 0.85618323, + "learning_rate": 2.0879953811903396e-06, + "loss": 0.87858605, + "num_input_tokens_seen": 89959495, + "step": 4170, + "time_per_iteration": 3.8428854942321777 + }, + { + "auxiliary_loss_clip": 0.01190193, + "auxiliary_loss_mlp": 0.01031088, + "balance_loss_clip": 1.05220342, + "balance_loss_mlp": 1.0223496, + "epoch": 0.5015330968556484, + "flos": 27527468382720.0, + "grad_norm": 1.8302992096489146, + "language_loss": 0.78687465, + "learning_rate": 2.08721715721893e-06, + "loss": 0.80908746, + "num_input_tokens_seen": 89978820, + "step": 4171, + "time_per_iteration": 3.6044836044311523 + }, + { + "auxiliary_loss_clip": 0.0118938, + "auxiliary_loss_mlp": 0.0102634, + "balance_loss_clip": 1.05129516, + "balance_loss_mlp": 1.01823974, + "epoch": 0.5016533397462875, + "flos": 23800802376960.0, + "grad_norm": 1.866549148834846, + "language_loss": 0.77148724, + "learning_rate": 2.0864389200167477e-06, + "loss": 0.79364443, + "num_input_tokens_seen": 89997075, + "step": 4172, + "time_per_iteration": 2.612589120864868 + }, + { + "auxiliary_loss_clip": 0.01193999, + "auxiliary_loss_mlp": 0.01378725, + "balance_loss_clip": 1.05308974, + "balance_loss_mlp": 1.00011444, + "epoch": 0.5017735826369266, + "flos": 25295009264640.0, + "grad_norm": 1.7327922480836042, + "language_loss": 0.79210263, + "learning_rate": 2.0856606697018504e-06, + "loss": 0.81782985, + "num_input_tokens_seen": 90015085, + "step": 4173, + "time_per_iteration": 2.715695858001709 + }, + { + "auxiliary_loss_clip": 0.01195161, + "auxiliary_loss_mlp": 0.01029705, + "balance_loss_clip": 1.04986835, + "balance_loss_mlp": 1.02117586, + "epoch": 0.5018938255275657, + "flos": 16873203778560.0, + "grad_norm": 2.199373621693502, + "language_loss": 0.72977757, + "learning_rate": 2.084882406392297e-06, + "loss": 0.7520262, + "num_input_tokens_seen": 90033045, + "step": 4174, + "time_per_iteration": 2.625715970993042 + }, + { + "auxiliary_loss_clip": 0.01194193, + "auxiliary_loss_mlp": 0.01034622, + "balance_loss_clip": 1.05356669, + "balance_loss_mlp": 1.02644444, + "epoch": 0.5020140684182047, + "flos": 25515429073920.0, + "grad_norm": 2.511431450406713, + "language_loss": 0.70769036, + "learning_rate": 2.0841041302061496e-06, + "loss": 0.7299785, + "num_input_tokens_seen": 90052505, + "step": 4175, + "time_per_iteration": 2.681112289428711 + }, + { + "auxiliary_loss_clip": 0.01184665, + "auxiliary_loss_mlp": 0.01029973, + "balance_loss_clip": 1.04596341, + "balance_loss_mlp": 1.02167034, + "epoch": 0.5021343113088439, + "flos": 23659278791040.0, + "grad_norm": 1.7430344418466175, + "language_loss": 0.75240082, + "learning_rate": 2.083325841261473e-06, + "loss": 0.77454716, + "num_input_tokens_seen": 90071565, + "step": 4176, + "time_per_iteration": 3.59285831451416 + }, + { + "auxiliary_loss_clip": 0.01190251, + "auxiliary_loss_mlp": 0.01026269, + "balance_loss_clip": 1.04620862, + "balance_loss_mlp": 1.01819825, + "epoch": 0.502254554199483, + "flos": 24534673148160.0, + "grad_norm": 1.803483913253401, + "language_loss": 0.65694654, + "learning_rate": 2.0825475396763322e-06, + "loss": 0.67911172, + "num_input_tokens_seen": 90092215, + "step": 4177, + "time_per_iteration": 2.6792399883270264 + }, + { + "auxiliary_loss_clip": 0.01228444, + "auxiliary_loss_mlp": 0.01027852, + "balance_loss_clip": 1.03829443, + "balance_loss_mlp": 1.01984167, + "epoch": 0.502374797090122, + "flos": 34240285607040.0, + "grad_norm": 1.3604828878155926, + "language_loss": 0.65599179, + "learning_rate": 2.081769225568796e-06, + "loss": 0.67855477, + "num_input_tokens_seen": 90114665, + "step": 4178, + "time_per_iteration": 3.3385097980499268 + }, + { + "auxiliary_loss_clip": 0.01189447, + "auxiliary_loss_mlp": 0.01030811, + "balance_loss_clip": 1.04792619, + "balance_loss_mlp": 1.0229311, + "epoch": 0.5024950399807612, + "flos": 26031106679040.0, + "grad_norm": 1.4723376157169787, + "language_loss": 0.7599169, + "learning_rate": 2.0809908990569327e-06, + "loss": 0.78211939, + "num_input_tokens_seen": 90136445, + "step": 4179, + "time_per_iteration": 3.2573533058166504 + }, + { + "auxiliary_loss_clip": 0.01198294, + "auxiliary_loss_mlp": 0.01028872, + "balance_loss_clip": 1.0512042, + "balance_loss_mlp": 1.02046156, + "epoch": 0.5026152828714002, + "flos": 21252438120960.0, + "grad_norm": 2.207292024674372, + "language_loss": 0.78967702, + "learning_rate": 2.0802125602588146e-06, + "loss": 0.8119486, + "num_input_tokens_seen": 90155710, + "step": 4180, + "time_per_iteration": 2.7208211421966553 + }, + { + "auxiliary_loss_clip": 0.01183157, + "auxiliary_loss_mlp": 0.01028815, + "balance_loss_clip": 1.05297828, + "balance_loss_mlp": 1.02099538, + "epoch": 0.5027355257620393, + "flos": 30956111245440.0, + "grad_norm": 1.8192926814681456, + "language_loss": 0.66846609, + "learning_rate": 2.0794342092925146e-06, + "loss": 0.69058585, + "num_input_tokens_seen": 90176845, + "step": 4181, + "time_per_iteration": 2.733980417251587 + }, + { + "auxiliary_loss_clip": 0.01194046, + "auxiliary_loss_mlp": 0.01029141, + "balance_loss_clip": 1.05410123, + "balance_loss_mlp": 1.02104688, + "epoch": 0.5028557686526784, + "flos": 24791147233920.0, + "grad_norm": 2.9031398201760896, + "language_loss": 0.67778802, + "learning_rate": 2.078655846276108e-06, + "loss": 0.7000199, + "num_input_tokens_seen": 90197175, + "step": 4182, + "time_per_iteration": 2.6474199295043945 + }, + { + "auxiliary_loss_clip": 0.0119504, + "auxiliary_loss_mlp": 0.01022862, + "balance_loss_clip": 1.04913163, + "balance_loss_mlp": 1.01433861, + "epoch": 0.5029760115433175, + "flos": 22966992990720.0, + "grad_norm": 1.9859089512149288, + "language_loss": 0.68570924, + "learning_rate": 2.0778774713276727e-06, + "loss": 0.70788825, + "num_input_tokens_seen": 90216650, + "step": 4183, + "time_per_iteration": 2.7469139099121094 + }, + { + "auxiliary_loss_clip": 0.01185055, + "auxiliary_loss_mlp": 0.01035109, + "balance_loss_clip": 1.04785228, + "balance_loss_mlp": 1.0266571, + "epoch": 0.5030962544339566, + "flos": 15305164485120.0, + "grad_norm": 2.2036266318288913, + "language_loss": 0.68169582, + "learning_rate": 2.077099084565287e-06, + "loss": 0.70389748, + "num_input_tokens_seen": 90234055, + "step": 4184, + "time_per_iteration": 2.581916332244873 + }, + { + "auxiliary_loss_clip": 0.01190243, + "auxiliary_loss_mlp": 0.01025071, + "balance_loss_clip": 1.04644084, + "balance_loss_mlp": 1.01658309, + "epoch": 0.5032164973245957, + "flos": 24494847943680.0, + "grad_norm": 2.160041498070599, + "language_loss": 0.65321255, + "learning_rate": 2.0763206861070313e-06, + "loss": 0.67536563, + "num_input_tokens_seen": 90253115, + "step": 4185, + "time_per_iteration": 2.7438180446624756 + }, + { + "auxiliary_loss_clip": 0.01183793, + "auxiliary_loss_mlp": 0.01029589, + "balance_loss_clip": 1.05392122, + "balance_loss_mlp": 1.02147698, + "epoch": 0.5033367402152348, + "flos": 16213452721920.0, + "grad_norm": 1.9028012000849146, + "language_loss": 0.74994385, + "learning_rate": 2.0755422760709876e-06, + "loss": 0.77207762, + "num_input_tokens_seen": 90270515, + "step": 4186, + "time_per_iteration": 2.5779714584350586 + }, + { + "auxiliary_loss_clip": 0.01208207, + "auxiliary_loss_mlp": 0.01032022, + "balance_loss_clip": 1.0454216, + "balance_loss_mlp": 1.02402294, + "epoch": 0.5034569831058738, + "flos": 21391375927680.0, + "grad_norm": 1.9272742785702048, + "language_loss": 0.77094871, + "learning_rate": 2.0747638545752417e-06, + "loss": 0.79335105, + "num_input_tokens_seen": 90289075, + "step": 4187, + "time_per_iteration": 2.8046255111694336 + }, + { + "auxiliary_loss_clip": 0.01198297, + "auxiliary_loss_mlp": 0.01028958, + "balance_loss_clip": 1.05094385, + "balance_loss_mlp": 1.02082515, + "epoch": 0.503577225996513, + "flos": 20558751690240.0, + "grad_norm": 1.9499905273283713, + "language_loss": 0.83387184, + "learning_rate": 2.073985421737878e-06, + "loss": 0.85614443, + "num_input_tokens_seen": 90306385, + "step": 4188, + "time_per_iteration": 2.6772778034210205 + }, + { + "auxiliary_loss_clip": 0.01194282, + "auxiliary_loss_mlp": 0.01028242, + "balance_loss_clip": 1.05294919, + "balance_loss_mlp": 1.02012396, + "epoch": 0.5036974688871521, + "flos": 27229157930880.0, + "grad_norm": 2.3099353066453685, + "language_loss": 0.74252772, + "learning_rate": 2.0732069776769844e-06, + "loss": 0.76475298, + "num_input_tokens_seen": 90323795, + "step": 4189, + "time_per_iteration": 2.7222442626953125 + }, + { + "auxiliary_loss_clip": 0.01183676, + "auxiliary_loss_mlp": 0.01031397, + "balance_loss_clip": 1.05484951, + "balance_loss_mlp": 1.0235889, + "epoch": 0.5038177117777911, + "flos": 20412164286720.0, + "grad_norm": 2.2046957662008593, + "language_loss": 0.73019892, + "learning_rate": 2.072428522510651e-06, + "loss": 0.75234967, + "num_input_tokens_seen": 90340360, + "step": 4190, + "time_per_iteration": 2.5828685760498047 + }, + { + "auxiliary_loss_clip": 0.0120428, + "auxiliary_loss_mlp": 0.0103324, + "balance_loss_clip": 1.05003464, + "balance_loss_mlp": 1.02449596, + "epoch": 0.5039379546684303, + "flos": 21907987286400.0, + "grad_norm": 2.21822791567805, + "language_loss": 0.76308161, + "learning_rate": 2.071650056356968e-06, + "loss": 0.78545678, + "num_input_tokens_seen": 90357900, + "step": 4191, + "time_per_iteration": 2.7816150188446045 + }, + { + "auxiliary_loss_clip": 0.01185952, + "auxiliary_loss_mlp": 0.01028232, + "balance_loss_clip": 1.05533767, + "balance_loss_mlp": 1.02068019, + "epoch": 0.5040581975590693, + "flos": 20010718909440.0, + "grad_norm": 1.8677542116316441, + "language_loss": 0.80058891, + "learning_rate": 2.070871579334028e-06, + "loss": 0.82273078, + "num_input_tokens_seen": 90377010, + "step": 4192, + "time_per_iteration": 2.6574864387512207 + }, + { + "auxiliary_loss_clip": 0.01181821, + "auxiliary_loss_mlp": 0.01028945, + "balance_loss_clip": 1.05258906, + "balance_loss_mlp": 1.02070725, + "epoch": 0.5041784404497084, + "flos": 20959837931520.0, + "grad_norm": 1.5890414929765648, + "language_loss": 0.71957028, + "learning_rate": 2.0700930915599264e-06, + "loss": 0.74167794, + "num_input_tokens_seen": 90396740, + "step": 4193, + "time_per_iteration": 2.606353759765625 + }, + { + "auxiliary_loss_clip": 0.01181082, + "auxiliary_loss_mlp": 0.01027879, + "balance_loss_clip": 1.05233431, + "balance_loss_mlp": 1.01954627, + "epoch": 0.5042986833403476, + "flos": 12495082757760.0, + "grad_norm": 2.415063964299962, + "language_loss": 0.78366131, + "learning_rate": 2.0693145931527583e-06, + "loss": 0.80575085, + "num_input_tokens_seen": 90413220, + "step": 4194, + "time_per_iteration": 2.5975489616394043 + }, + { + "auxiliary_loss_clip": 0.01195696, + "auxiliary_loss_mlp": 0.01023885, + "balance_loss_clip": 1.04954588, + "balance_loss_mlp": 1.01584458, + "epoch": 0.5044189262309866, + "flos": 29202305788800.0, + "grad_norm": 1.548855649335128, + "language_loss": 0.77898717, + "learning_rate": 2.068536084230622e-06, + "loss": 0.80118299, + "num_input_tokens_seen": 90435085, + "step": 4195, + "time_per_iteration": 2.7075090408325195 + }, + { + "auxiliary_loss_clip": 0.01191305, + "auxiliary_loss_mlp": 0.01027814, + "balance_loss_clip": 1.05206537, + "balance_loss_mlp": 1.01953518, + "epoch": 0.5045391691216257, + "flos": 23873198238720.0, + "grad_norm": 2.041508815146441, + "language_loss": 0.88616765, + "learning_rate": 2.067757564911616e-06, + "loss": 0.90835887, + "num_input_tokens_seen": 90453660, + "step": 4196, + "time_per_iteration": 3.5742573738098145 + }, + { + "auxiliary_loss_clip": 0.0120743, + "auxiliary_loss_mlp": 0.0137894, + "balance_loss_clip": 1.05072427, + "balance_loss_mlp": 1.00020337, + "epoch": 0.5046594120122648, + "flos": 24644990793600.0, + "grad_norm": 2.22933762188639, + "language_loss": 0.9279412, + "learning_rate": 2.0669790353138407e-06, + "loss": 0.95380485, + "num_input_tokens_seen": 90472625, + "step": 4197, + "time_per_iteration": 3.6386640071868896 + }, + { + "auxiliary_loss_clip": 0.01202771, + "auxiliary_loss_mlp": 0.01379079, + "balance_loss_clip": 1.04970586, + "balance_loss_mlp": 1.00014937, + "epoch": 0.5047796549029039, + "flos": 23362835846400.0, + "grad_norm": 5.203059265483145, + "language_loss": 0.73089796, + "learning_rate": 2.0662004955553995e-06, + "loss": 0.75671649, + "num_input_tokens_seen": 90492325, + "step": 4198, + "time_per_iteration": 2.6833908557891846 + }, + { + "auxiliary_loss_clip": 0.01195279, + "auxiliary_loss_mlp": 0.01023116, + "balance_loss_clip": 1.04916108, + "balance_loss_mlp": 1.01563549, + "epoch": 0.5048998977935429, + "flos": 17304095329920.0, + "grad_norm": 1.9717738953388206, + "language_loss": 0.76846743, + "learning_rate": 2.065421945754395e-06, + "loss": 0.79065132, + "num_input_tokens_seen": 90510055, + "step": 4199, + "time_per_iteration": 2.6621265411376953 + }, + { + "auxiliary_loss_clip": 0.01221961, + "auxiliary_loss_mlp": 0.0103059, + "balance_loss_clip": 1.05097985, + "balance_loss_mlp": 1.02322304, + "epoch": 0.505020140684182, + "flos": 34856979235200.0, + "grad_norm": 1.6200746120054, + "language_loss": 0.77825332, + "learning_rate": 2.0646433860289344e-06, + "loss": 0.80077881, + "num_input_tokens_seen": 90528980, + "step": 4200, + "time_per_iteration": 2.8580482006073 + }, + { + "auxiliary_loss_clip": 0.01195915, + "auxiliary_loss_mlp": 0.01378908, + "balance_loss_clip": 1.0517478, + "balance_loss_mlp": 1.00020063, + "epoch": 0.5051403835748212, + "flos": 24863974058880.0, + "grad_norm": 1.8605652709869227, + "language_loss": 0.82237375, + "learning_rate": 2.0638648164971233e-06, + "loss": 0.84812188, + "num_input_tokens_seen": 90547445, + "step": 4201, + "time_per_iteration": 2.731656074523926 + }, + { + "auxiliary_loss_clip": 0.01197796, + "auxiliary_loss_mlp": 0.01025889, + "balance_loss_clip": 1.05063784, + "balance_loss_mlp": 1.01781249, + "epoch": 0.5052606264654602, + "flos": 20959694277120.0, + "grad_norm": 1.8659191317496833, + "language_loss": 0.88594878, + "learning_rate": 2.06308623727707e-06, + "loss": 0.9081856, + "num_input_tokens_seen": 90567545, + "step": 4202, + "time_per_iteration": 3.578005313873291 + }, + { + "auxiliary_loss_clip": 0.01188057, + "auxiliary_loss_mlp": 0.01032353, + "balance_loss_clip": 1.05104983, + "balance_loss_mlp": 1.02411556, + "epoch": 0.5053808693560993, + "flos": 19642382893440.0, + "grad_norm": 2.624907402181854, + "language_loss": 0.76461703, + "learning_rate": 2.0623076484868846e-06, + "loss": 0.78682113, + "num_input_tokens_seen": 90585000, + "step": 4203, + "time_per_iteration": 2.6581549644470215 + }, + { + "auxiliary_loss_clip": 0.01102948, + "auxiliary_loss_mlp": 0.01002341, + "balance_loss_clip": 1.02408814, + "balance_loss_mlp": 1.00112462, + "epoch": 0.5055011122467384, + "flos": 67504915019520.0, + "grad_norm": 0.8312215852847408, + "language_loss": 0.60719287, + "learning_rate": 2.061529050244679e-06, + "loss": 0.62824583, + "num_input_tokens_seen": 90644745, + "step": 4204, + "time_per_iteration": 3.186854362487793 + }, + { + "auxiliary_loss_clip": 0.01214261, + "auxiliary_loss_mlp": 0.01022755, + "balance_loss_clip": 1.04931366, + "balance_loss_mlp": 1.0154233, + "epoch": 0.5056213551373775, + "flos": 16872952383360.0, + "grad_norm": 1.8844299833910358, + "language_loss": 0.73947453, + "learning_rate": 2.060750442668565e-06, + "loss": 0.76184464, + "num_input_tokens_seen": 90662500, + "step": 4205, + "time_per_iteration": 2.639244794845581 + }, + { + "auxiliary_loss_clip": 0.01192466, + "auxiliary_loss_mlp": 0.01024678, + "balance_loss_clip": 1.0533917, + "balance_loss_mlp": 1.01688766, + "epoch": 0.5057415980280165, + "flos": 15334179696000.0, + "grad_norm": 2.0860630702150593, + "language_loss": 0.63940305, + "learning_rate": 2.059971825876657e-06, + "loss": 0.66157454, + "num_input_tokens_seen": 90677010, + "step": 4206, + "time_per_iteration": 2.5846738815307617 + }, + { + "auxiliary_loss_clip": 0.01195646, + "auxiliary_loss_mlp": 0.01025722, + "balance_loss_clip": 1.05389953, + "balance_loss_mlp": 1.01837015, + "epoch": 0.5058618409186557, + "flos": 19025976574080.0, + "grad_norm": 1.8944856613905996, + "language_loss": 0.7622571, + "learning_rate": 2.0591931999870713e-06, + "loss": 0.7844708, + "num_input_tokens_seen": 90695935, + "step": 4207, + "time_per_iteration": 2.625774383544922 + }, + { + "auxiliary_loss_clip": 0.0108932, + "auxiliary_loss_mlp": 0.01002194, + "balance_loss_clip": 1.02373183, + "balance_loss_mlp": 1.00098443, + "epoch": 0.5059820838092948, + "flos": 63453114080640.0, + "grad_norm": 0.8189266150537097, + "language_loss": 0.57602072, + "learning_rate": 2.0584145651179234e-06, + "loss": 0.59693587, + "num_input_tokens_seen": 90751645, + "step": 4208, + "time_per_iteration": 3.26898193359375 + }, + { + "auxiliary_loss_clip": 0.01199808, + "auxiliary_loss_mlp": 0.01379208, + "balance_loss_clip": 1.05189741, + "balance_loss_mlp": 1.00021577, + "epoch": 0.5061023266999338, + "flos": 15441803821440.0, + "grad_norm": 2.1773639363124455, + "language_loss": 0.79969376, + "learning_rate": 2.0576359213873327e-06, + "loss": 0.82548392, + "num_input_tokens_seen": 90766795, + "step": 4209, + "time_per_iteration": 2.596806764602661 + }, + { + "auxiliary_loss_clip": 0.0120603, + "auxiliary_loss_mlp": 0.01023577, + "balance_loss_clip": 1.04847383, + "balance_loss_mlp": 1.01486254, + "epoch": 0.506222569590573, + "flos": 22451063990400.0, + "grad_norm": 3.513519291064929, + "language_loss": 0.70292211, + "learning_rate": 2.056857268913419e-06, + "loss": 0.72521818, + "num_input_tokens_seen": 90786845, + "step": 4210, + "time_per_iteration": 2.7465686798095703 + }, + { + "auxiliary_loss_clip": 0.01193391, + "auxiliary_loss_mlp": 0.01026059, + "balance_loss_clip": 1.0533551, + "balance_loss_mlp": 1.01810789, + "epoch": 0.506342812481212, + "flos": 17558665994880.0, + "grad_norm": 2.051502297501611, + "language_loss": 0.84287244, + "learning_rate": 2.056078607814303e-06, + "loss": 0.86506689, + "num_input_tokens_seen": 90802630, + "step": 4211, + "time_per_iteration": 2.563680410385132 + }, + { + "auxiliary_loss_clip": 0.01188693, + "auxiliary_loss_mlp": 0.01025321, + "balance_loss_clip": 1.05235839, + "balance_loss_mlp": 1.01714945, + "epoch": 0.5064630553718511, + "flos": 23402050519680.0, + "grad_norm": 1.7655655525784133, + "language_loss": 0.78387988, + "learning_rate": 2.055299938208106e-06, + "loss": 0.80602002, + "num_input_tokens_seen": 90823620, + "step": 4212, + "time_per_iteration": 2.6358232498168945 + }, + { + "auxiliary_loss_clip": 0.01197664, + "auxiliary_loss_mlp": 0.01031336, + "balance_loss_clip": 1.05446839, + "balance_loss_mlp": 1.02299094, + "epoch": 0.5065832982624903, + "flos": 23987035416960.0, + "grad_norm": 1.814544053183134, + "language_loss": 0.86333054, + "learning_rate": 2.0545212602129526e-06, + "loss": 0.88562053, + "num_input_tokens_seen": 90843475, + "step": 4213, + "time_per_iteration": 2.6932077407836914 + }, + { + "auxiliary_loss_clip": 0.0119176, + "auxiliary_loss_mlp": 0.01026011, + "balance_loss_clip": 1.04816175, + "balance_loss_mlp": 1.01801753, + "epoch": 0.5067035411531293, + "flos": 21503058289920.0, + "grad_norm": 2.5467333480862915, + "language_loss": 0.66420746, + "learning_rate": 2.0537425739469673e-06, + "loss": 0.68638521, + "num_input_tokens_seen": 90862410, + "step": 4214, + "time_per_iteration": 2.648442029953003 + }, + { + "auxiliary_loss_clip": 0.01087173, + "auxiliary_loss_mlp": 0.01004974, + "balance_loss_clip": 1.01508737, + "balance_loss_mlp": 1.00365722, + "epoch": 0.5068237840437684, + "flos": 65934397687680.0, + "grad_norm": 0.8369795297594648, + "language_loss": 0.59478855, + "learning_rate": 2.052963879528276e-06, + "loss": 0.61571002, + "num_input_tokens_seen": 90922280, + "step": 4215, + "time_per_iteration": 3.14144229888916 + }, + { + "auxiliary_loss_clip": 0.01193579, + "auxiliary_loss_mlp": 0.01027158, + "balance_loss_clip": 1.05340612, + "balance_loss_mlp": 1.01902819, + "epoch": 0.5069440269344075, + "flos": 27264206626560.0, + "grad_norm": 2.0444915881870265, + "language_loss": 0.76787794, + "learning_rate": 2.052185177075007e-06, + "loss": 0.79008532, + "num_input_tokens_seen": 90941850, + "step": 4216, + "time_per_iteration": 2.653721809387207 + }, + { + "auxiliary_loss_clip": 0.01191144, + "auxiliary_loss_mlp": 0.01030543, + "balance_loss_clip": 1.04982519, + "balance_loss_mlp": 1.0224669, + "epoch": 0.5070642698250466, + "flos": 23366319465600.0, + "grad_norm": 2.7671847664660287, + "language_loss": 0.83070576, + "learning_rate": 2.051406466705288e-06, + "loss": 0.85292262, + "num_input_tokens_seen": 90961390, + "step": 4217, + "time_per_iteration": 2.659158229827881 + }, + { + "auxiliary_loss_clip": 0.01181443, + "auxiliary_loss_mlp": 0.01029751, + "balance_loss_clip": 1.05142593, + "balance_loss_mlp": 1.02204728, + "epoch": 0.5071845127156857, + "flos": 20340127560960.0, + "grad_norm": 1.8862319385345128, + "language_loss": 0.80824143, + "learning_rate": 2.0506277485372486e-06, + "loss": 0.83035338, + "num_input_tokens_seen": 90980215, + "step": 4218, + "time_per_iteration": 2.549957752227783 + }, + { + "auxiliary_loss_clip": 0.01183666, + "auxiliary_loss_mlp": 0.01027825, + "balance_loss_clip": 1.05037713, + "balance_loss_mlp": 1.01958752, + "epoch": 0.5073047556063248, + "flos": 12092955022080.0, + "grad_norm": 1.6846218843206282, + "language_loss": 0.66703296, + "learning_rate": 2.04984902268902e-06, + "loss": 0.68914789, + "num_input_tokens_seen": 90997415, + "step": 4219, + "time_per_iteration": 2.690603256225586 + }, + { + "auxiliary_loss_clip": 0.01198169, + "auxiliary_loss_mlp": 0.01029843, + "balance_loss_clip": 1.04981887, + "balance_loss_mlp": 1.02127206, + "epoch": 0.5074249984969639, + "flos": 19682854542720.0, + "grad_norm": 2.108960464643654, + "language_loss": 0.75453663, + "learning_rate": 2.0490702892787345e-06, + "loss": 0.77681673, + "num_input_tokens_seen": 91016475, + "step": 4220, + "time_per_iteration": 2.5813379287719727 + }, + { + "auxiliary_loss_clip": 0.01180678, + "auxiliary_loss_mlp": 0.01026051, + "balance_loss_clip": 1.04776907, + "balance_loss_mlp": 1.01809931, + "epoch": 0.5075452413876029, + "flos": 28765703975040.0, + "grad_norm": 1.6920425620266961, + "language_loss": 0.62302315, + "learning_rate": 2.0482915484245246e-06, + "loss": 0.64509046, + "num_input_tokens_seen": 91038095, + "step": 4221, + "time_per_iteration": 2.6778907775878906 + }, + { + "auxiliary_loss_clip": 0.01208067, + "auxiliary_loss_mlp": 0.01033971, + "balance_loss_clip": 1.04715228, + "balance_loss_mlp": 1.02503073, + "epoch": 0.5076654842782421, + "flos": 20339445202560.0, + "grad_norm": 2.187823650943413, + "language_loss": 0.8447926, + "learning_rate": 2.047512800244526e-06, + "loss": 0.86721301, + "num_input_tokens_seen": 91053360, + "step": 4222, + "time_per_iteration": 4.478796005249023 + }, + { + "auxiliary_loss_clip": 0.01193617, + "auxiliary_loss_mlp": 0.01031012, + "balance_loss_clip": 1.05403328, + "balance_loss_mlp": 1.02211261, + "epoch": 0.5077857271688812, + "flos": 26359653404160.0, + "grad_norm": 2.065036462573562, + "language_loss": 0.78999931, + "learning_rate": 2.046734044856873e-06, + "loss": 0.81224561, + "num_input_tokens_seen": 91072770, + "step": 4223, + "time_per_iteration": 3.598926305770874 + }, + { + "auxiliary_loss_clip": 0.01191134, + "auxiliary_loss_mlp": 0.01029034, + "balance_loss_clip": 1.0532546, + "balance_loss_mlp": 1.02143478, + "epoch": 0.5079059700595202, + "flos": 21798962530560.0, + "grad_norm": 2.2657351198672204, + "language_loss": 0.81351215, + "learning_rate": 2.045955282379702e-06, + "loss": 0.83571386, + "num_input_tokens_seen": 91091430, + "step": 4224, + "time_per_iteration": 2.6534054279327393 + }, + { + "auxiliary_loss_clip": 0.01184435, + "auxiliary_loss_mlp": 0.01026233, + "balance_loss_clip": 1.04696965, + "balance_loss_mlp": 1.01811433, + "epoch": 0.5080262129501594, + "flos": 13187943175680.0, + "grad_norm": 3.283106013495537, + "language_loss": 0.75033176, + "learning_rate": 2.045176512931152e-06, + "loss": 0.77243835, + "num_input_tokens_seen": 91106060, + "step": 4225, + "time_per_iteration": 2.5693135261535645 + }, + { + "auxiliary_loss_clip": 0.01209548, + "auxiliary_loss_mlp": 0.01029604, + "balance_loss_clip": 1.04785323, + "balance_loss_mlp": 1.02205169, + "epoch": 0.5081464558407984, + "flos": 25301473712640.0, + "grad_norm": 1.837881498020555, + "language_loss": 0.75779283, + "learning_rate": 2.0443977366293604e-06, + "loss": 0.78018433, + "num_input_tokens_seen": 91124100, + "step": 4226, + "time_per_iteration": 2.724564790725708 + }, + { + "auxiliary_loss_clip": 0.01226668, + "auxiliary_loss_mlp": 0.0102965, + "balance_loss_clip": 1.04517484, + "balance_loss_mlp": 1.0205667, + "epoch": 0.5082666987314375, + "flos": 30951226995840.0, + "grad_norm": 1.8530646650637663, + "language_loss": 0.76890373, + "learning_rate": 2.043618953592468e-06, + "loss": 0.79146695, + "num_input_tokens_seen": 91146555, + "step": 4227, + "time_per_iteration": 2.8553354740142822 + }, + { + "auxiliary_loss_clip": 0.01202464, + "auxiliary_loss_mlp": 0.01031742, + "balance_loss_clip": 1.0522604, + "balance_loss_mlp": 1.0228014, + "epoch": 0.5083869416220766, + "flos": 19682495406720.0, + "grad_norm": 1.983143002443886, + "language_loss": 0.80969268, + "learning_rate": 2.0428401639386144e-06, + "loss": 0.83203477, + "num_input_tokens_seen": 91167120, + "step": 4228, + "time_per_iteration": 3.6082475185394287 + }, + { + "auxiliary_loss_clip": 0.0109676, + "auxiliary_loss_mlp": 0.01001998, + "balance_loss_clip": 1.01730883, + "balance_loss_mlp": 1.00081813, + "epoch": 0.5085071845127157, + "flos": 71817535589760.0, + "grad_norm": 0.8247902242443426, + "language_loss": 0.58106309, + "learning_rate": 2.042061367785943e-06, + "loss": 0.60205072, + "num_input_tokens_seen": 91220260, + "step": 4229, + "time_per_iteration": 3.232391119003296 + }, + { + "auxiliary_loss_clip": 0.01207406, + "auxiliary_loss_mlp": 0.01023935, + "balance_loss_clip": 1.04551864, + "balance_loss_mlp": 1.01592422, + "epoch": 0.5086274274033548, + "flos": 35951608252800.0, + "grad_norm": 2.2549373937418506, + "language_loss": 0.75168002, + "learning_rate": 2.041282565252594e-06, + "loss": 0.77399343, + "num_input_tokens_seen": 91240425, + "step": 4230, + "time_per_iteration": 2.812490463256836 + }, + { + "auxiliary_loss_clip": 0.01203123, + "auxiliary_loss_mlp": 0.01027024, + "balance_loss_clip": 1.04461563, + "balance_loss_mlp": 1.01903749, + "epoch": 0.5087476702939938, + "flos": 23513732881920.0, + "grad_norm": 1.5877623582984164, + "language_loss": 0.77035087, + "learning_rate": 2.040503756456714e-06, + "loss": 0.79265237, + "num_input_tokens_seen": 91259635, + "step": 4231, + "time_per_iteration": 2.7537803649902344 + }, + { + "auxiliary_loss_clip": 0.01184514, + "auxiliary_loss_mlp": 0.01023176, + "balance_loss_clip": 1.04966831, + "balance_loss_mlp": 1.01523066, + "epoch": 0.508867913184633, + "flos": 15122091841920.0, + "grad_norm": 2.126478439146268, + "language_loss": 0.79118925, + "learning_rate": 2.0397249415164456e-06, + "loss": 0.8132661, + "num_input_tokens_seen": 91276990, + "step": 4232, + "time_per_iteration": 2.703097105026245 + }, + { + "auxiliary_loss_clip": 0.01186842, + "auxiliary_loss_mlp": 0.01025332, + "balance_loss_clip": 1.04441857, + "balance_loss_mlp": 1.01732063, + "epoch": 0.508988156075272, + "flos": 25885309374720.0, + "grad_norm": 1.5756996220930362, + "language_loss": 0.80015373, + "learning_rate": 2.0389461205499354e-06, + "loss": 0.82227546, + "num_input_tokens_seen": 91296125, + "step": 4233, + "time_per_iteration": 2.7373218536376953 + }, + { + "auxiliary_loss_clip": 0.01207644, + "auxiliary_loss_mlp": 0.01025593, + "balance_loss_clip": 1.04657102, + "balance_loss_mlp": 1.01801181, + "epoch": 0.5091083989659111, + "flos": 13844857057920.0, + "grad_norm": 2.1839219653213826, + "language_loss": 0.73481596, + "learning_rate": 2.03816729367533e-06, + "loss": 0.75714827, + "num_input_tokens_seen": 91314280, + "step": 4234, + "time_per_iteration": 2.7088165283203125 + }, + { + "auxiliary_loss_clip": 0.01200439, + "auxiliary_loss_mlp": 0.01028126, + "balance_loss_clip": 1.05000961, + "balance_loss_mlp": 1.01981688, + "epoch": 0.5092286418565503, + "flos": 21104881050240.0, + "grad_norm": 1.9534041122886516, + "language_loss": 0.71627581, + "learning_rate": 2.0373884610107765e-06, + "loss": 0.73856151, + "num_input_tokens_seen": 91334595, + "step": 4235, + "time_per_iteration": 2.6765947341918945 + }, + { + "auxiliary_loss_clip": 0.01192675, + "auxiliary_loss_mlp": 0.01025706, + "balance_loss_clip": 1.04895163, + "balance_loss_mlp": 1.01805878, + "epoch": 0.5093488847471893, + "flos": 18621298972800.0, + "grad_norm": 2.4287168117702134, + "language_loss": 0.69661742, + "learning_rate": 2.0366096226744225e-06, + "loss": 0.71880126, + "num_input_tokens_seen": 91349790, + "step": 4236, + "time_per_iteration": 2.6265926361083984 + }, + { + "auxiliary_loss_clip": 0.01178847, + "auxiliary_loss_mlp": 0.01028869, + "balance_loss_clip": 1.0496428, + "balance_loss_mlp": 1.02092028, + "epoch": 0.5094691276378284, + "flos": 23803783205760.0, + "grad_norm": 1.678808675196861, + "language_loss": 0.76641446, + "learning_rate": 2.035830778784418e-06, + "loss": 0.78849161, + "num_input_tokens_seen": 91370465, + "step": 4237, + "time_per_iteration": 2.654825210571289 + }, + { + "auxiliary_loss_clip": 0.01205204, + "auxiliary_loss_mlp": 0.01034153, + "balance_loss_clip": 1.05470848, + "balance_loss_mlp": 1.02576637, + "epoch": 0.5095893705284675, + "flos": 17420410546560.0, + "grad_norm": 1.9402762844121708, + "language_loss": 0.79962939, + "learning_rate": 2.0350519294589134e-06, + "loss": 0.82202291, + "num_input_tokens_seen": 91388505, + "step": 4238, + "time_per_iteration": 2.7223565578460693 + }, + { + "auxiliary_loss_clip": 0.01213282, + "auxiliary_loss_mlp": 0.01033223, + "balance_loss_clip": 1.04502606, + "balance_loss_mlp": 1.02478611, + "epoch": 0.5097096134191066, + "flos": 25849362839040.0, + "grad_norm": 1.8953426453986115, + "language_loss": 0.82776439, + "learning_rate": 2.0342730748160588e-06, + "loss": 0.85022944, + "num_input_tokens_seen": 91408970, + "step": 4239, + "time_per_iteration": 2.718949794769287 + }, + { + "auxiliary_loss_clip": 0.0120113, + "auxiliary_loss_mlp": 0.01028361, + "balance_loss_clip": 1.04993141, + "balance_loss_mlp": 1.01980138, + "epoch": 0.5098298563097456, + "flos": 27745122844800.0, + "grad_norm": 2.2908271610364634, + "language_loss": 0.70233715, + "learning_rate": 2.033494214974006e-06, + "loss": 0.72463202, + "num_input_tokens_seen": 91430115, + "step": 4240, + "time_per_iteration": 2.7284200191497803 + }, + { + "auxiliary_loss_clip": 0.01186792, + "auxiliary_loss_mlp": 0.0102853, + "balance_loss_clip": 1.04852247, + "balance_loss_mlp": 1.020576, + "epoch": 0.5099500992003848, + "flos": 21358913011200.0, + "grad_norm": 1.704327754525701, + "language_loss": 0.83870757, + "learning_rate": 2.0327153500509067e-06, + "loss": 0.86086071, + "num_input_tokens_seen": 91449140, + "step": 4241, + "time_per_iteration": 2.6527035236358643 + }, + { + "auxiliary_loss_clip": 0.01204792, + "auxiliary_loss_mlp": 0.01028407, + "balance_loss_clip": 1.05361593, + "balance_loss_mlp": 1.02027059, + "epoch": 0.5100703420910239, + "flos": 19865999013120.0, + "grad_norm": 1.9380654205302148, + "language_loss": 0.84522396, + "learning_rate": 2.031936480164916e-06, + "loss": 0.86755598, + "num_input_tokens_seen": 91466880, + "step": 4242, + "time_per_iteration": 2.662372589111328 + }, + { + "auxiliary_loss_clip": 0.01195129, + "auxiliary_loss_mlp": 0.0102578, + "balance_loss_clip": 1.05088484, + "balance_loss_mlp": 1.01744103, + "epoch": 0.5101905849816629, + "flos": 24648797635200.0, + "grad_norm": 3.4197626685218325, + "language_loss": 0.7976017, + "learning_rate": 2.0311576054341857e-06, + "loss": 0.81981075, + "num_input_tokens_seen": 91487495, + "step": 4243, + "time_per_iteration": 2.7211644649505615 + }, + { + "auxiliary_loss_clip": 0.01184661, + "auxiliary_loss_mlp": 0.01022833, + "balance_loss_clip": 1.05541909, + "balance_loss_mlp": 1.01429749, + "epoch": 0.5103108278723021, + "flos": 22930076787840.0, + "grad_norm": 1.6657361934664099, + "language_loss": 0.62520099, + "learning_rate": 2.0303787259768715e-06, + "loss": 0.64727592, + "num_input_tokens_seen": 91508395, + "step": 4244, + "time_per_iteration": 2.578350305557251 + }, + { + "auxiliary_loss_clip": 0.01202351, + "auxiliary_loss_mlp": 0.01030038, + "balance_loss_clip": 1.0520426, + "balance_loss_mlp": 1.02080476, + "epoch": 0.5104310707629411, + "flos": 21506613736320.0, + "grad_norm": 2.4030752443452053, + "language_loss": 0.68932045, + "learning_rate": 2.0295998419111294e-06, + "loss": 0.71164441, + "num_input_tokens_seen": 91525685, + "step": 4245, + "time_per_iteration": 2.696711540222168 + }, + { + "auxiliary_loss_clip": 0.0122529, + "auxiliary_loss_mlp": 0.01030086, + "balance_loss_clip": 1.04317594, + "balance_loss_mlp": 1.02164555, + "epoch": 0.5105513136535802, + "flos": 14903180403840.0, + "grad_norm": 4.97617845830936, + "language_loss": 0.73747247, + "learning_rate": 2.028820953355115e-06, + "loss": 0.76002622, + "num_input_tokens_seen": 91543785, + "step": 4246, + "time_per_iteration": 2.8281266689300537 + }, + { + "auxiliary_loss_clip": 0.01203967, + "auxiliary_loss_mlp": 0.01025376, + "balance_loss_clip": 1.04784894, + "balance_loss_mlp": 1.01709151, + "epoch": 0.5106715565442194, + "flos": 22602212421120.0, + "grad_norm": 1.8940298421875448, + "language_loss": 0.78261977, + "learning_rate": 2.0280420604269834e-06, + "loss": 0.80491322, + "num_input_tokens_seen": 91563325, + "step": 4247, + "time_per_iteration": 3.257953643798828 + }, + { + "auxiliary_loss_clip": 0.01087215, + "auxiliary_loss_mlp": 0.01003211, + "balance_loss_clip": 1.0164845, + "balance_loss_mlp": 1.00200713, + "epoch": 0.5107917994348584, + "flos": 71027645558400.0, + "grad_norm": 0.7182377277571154, + "language_loss": 0.58872664, + "learning_rate": 2.027263163244895e-06, + "loss": 0.60963094, + "num_input_tokens_seen": 91632450, + "step": 4248, + "time_per_iteration": 4.2831714153289795 + }, + { + "auxiliary_loss_clip": 0.01189916, + "auxiliary_loss_mlp": 0.01029632, + "balance_loss_clip": 1.05301535, + "balance_loss_mlp": 1.02213645, + "epoch": 0.5109120423254975, + "flos": 24827416992000.0, + "grad_norm": 1.6020699383215982, + "language_loss": 0.74413538, + "learning_rate": 2.026484261927005e-06, + "loss": 0.76633084, + "num_input_tokens_seen": 91651945, + "step": 4249, + "time_per_iteration": 3.552326202392578 + }, + { + "auxiliary_loss_clip": 0.01200022, + "auxiliary_loss_mlp": 0.01038025, + "balance_loss_clip": 1.05572283, + "balance_loss_mlp": 1.02915025, + "epoch": 0.5110322852161366, + "flos": 21247661612160.0, + "grad_norm": 2.2360065597908636, + "language_loss": 0.73929453, + "learning_rate": 2.025705356591475e-06, + "loss": 0.761675, + "num_input_tokens_seen": 91669635, + "step": 4250, + "time_per_iteration": 2.68507981300354 + }, + { + "auxiliary_loss_clip": 0.01109673, + "auxiliary_loss_mlp": 0.0137414, + "balance_loss_clip": 1.01658726, + "balance_loss_mlp": 0.99972099, + "epoch": 0.5111525281067757, + "flos": 66457114358400.0, + "grad_norm": 0.7625412980226846, + "language_loss": 0.57933217, + "learning_rate": 2.024926447356462e-06, + "loss": 0.60417026, + "num_input_tokens_seen": 91731920, + "step": 4251, + "time_per_iteration": 3.24774432182312 + }, + { + "auxiliary_loss_clip": 0.01190074, + "auxiliary_loss_mlp": 0.01025594, + "balance_loss_clip": 1.0506295, + "balance_loss_mlp": 1.01734507, + "epoch": 0.5112727709974147, + "flos": 14866731077760.0, + "grad_norm": 2.0339026151702604, + "language_loss": 0.79045558, + "learning_rate": 2.024147534340127e-06, + "loss": 0.81261224, + "num_input_tokens_seen": 91749780, + "step": 4252, + "time_per_iteration": 2.6461668014526367 + }, + { + "auxiliary_loss_clip": 0.01194599, + "auxiliary_loss_mlp": 0.01029466, + "balance_loss_clip": 1.04641485, + "balance_loss_mlp": 1.02124095, + "epoch": 0.5113930138880539, + "flos": 21177600134400.0, + "grad_norm": 1.5860251053559642, + "language_loss": 0.80097044, + "learning_rate": 2.02336861766063e-06, + "loss": 0.82321107, + "num_input_tokens_seen": 91768840, + "step": 4253, + "time_per_iteration": 2.733358144760132 + }, + { + "auxiliary_loss_clip": 0.01198233, + "auxiliary_loss_mlp": 0.01031529, + "balance_loss_clip": 1.0528419, + "balance_loss_mlp": 1.02272558, + "epoch": 0.511513256778693, + "flos": 20409111630720.0, + "grad_norm": 1.6175407761654863, + "language_loss": 0.78880787, + "learning_rate": 2.0225896974361327e-06, + "loss": 0.81110555, + "num_input_tokens_seen": 91788945, + "step": 4254, + "time_per_iteration": 3.5460398197174072 + }, + { + "auxiliary_loss_clip": 0.01112813, + "auxiliary_loss_mlp": 0.01002756, + "balance_loss_clip": 1.01756692, + "balance_loss_mlp": 1.00146818, + "epoch": 0.511633499669332, + "flos": 69879975131520.0, + "grad_norm": 0.8554837048553069, + "language_loss": 0.59941441, + "learning_rate": 2.0218107737847962e-06, + "loss": 0.62057006, + "num_input_tokens_seen": 91850990, + "step": 4255, + "time_per_iteration": 3.298104763031006 + }, + { + "auxiliary_loss_clip": 0.01181787, + "auxiliary_loss_mlp": 0.01024639, + "balance_loss_clip": 1.05234408, + "balance_loss_mlp": 1.01688445, + "epoch": 0.5117537425599712, + "flos": 24097855852800.0, + "grad_norm": 1.8146125027497997, + "language_loss": 0.7469511, + "learning_rate": 2.0210318468247826e-06, + "loss": 0.76901537, + "num_input_tokens_seen": 91869960, + "step": 4256, + "time_per_iteration": 2.6337594985961914 + }, + { + "auxiliary_loss_clip": 0.01197445, + "auxiliary_loss_mlp": 0.01026802, + "balance_loss_clip": 1.04828954, + "balance_loss_mlp": 1.01911616, + "epoch": 0.5118739854506102, + "flos": 20959550622720.0, + "grad_norm": 1.8450960353561097, + "language_loss": 0.8152591, + "learning_rate": 2.020252916674255e-06, + "loss": 0.83750165, + "num_input_tokens_seen": 91889075, + "step": 4257, + "time_per_iteration": 2.6639790534973145 + }, + { + "auxiliary_loss_clip": 0.01191535, + "auxiliary_loss_mlp": 0.01029636, + "balance_loss_clip": 1.05045307, + "balance_loss_mlp": 1.02147055, + "epoch": 0.5119942283412493, + "flos": 17457326749440.0, + "grad_norm": 1.733048039960415, + "language_loss": 0.81406891, + "learning_rate": 2.019473983451375e-06, + "loss": 0.83628058, + "num_input_tokens_seen": 91907495, + "step": 4258, + "time_per_iteration": 2.721486806869507 + }, + { + "auxiliary_loss_clip": 0.01216801, + "auxiliary_loss_mlp": 0.01030941, + "balance_loss_clip": 1.04887938, + "balance_loss_mlp": 1.02256656, + "epoch": 0.5121144712318885, + "flos": 21066743784960.0, + "grad_norm": 1.7361112358377944, + "language_loss": 0.71338832, + "learning_rate": 2.0186950472743076e-06, + "loss": 0.73586571, + "num_input_tokens_seen": 91927400, + "step": 4259, + "time_per_iteration": 2.7176501750946045 + }, + { + "auxiliary_loss_clip": 0.0118261, + "auxiliary_loss_mlp": 0.01027605, + "balance_loss_clip": 1.05274606, + "balance_loss_mlp": 1.02010965, + "epoch": 0.5122347141225275, + "flos": 19860791541120.0, + "grad_norm": 1.6679383123609997, + "language_loss": 0.73727894, + "learning_rate": 2.0179161082612162e-06, + "loss": 0.75938106, + "num_input_tokens_seen": 91946790, + "step": 4260, + "time_per_iteration": 2.617565631866455 + }, + { + "auxiliary_loss_clip": 0.0119169, + "auxiliary_loss_mlp": 0.01028585, + "balance_loss_clip": 1.04609406, + "balance_loss_mlp": 1.02043748, + "epoch": 0.5123549570131666, + "flos": 22528487756160.0, + "grad_norm": 1.8264499097371105, + "language_loss": 0.72916496, + "learning_rate": 2.017137166530266e-06, + "loss": 0.75136769, + "num_input_tokens_seen": 91966325, + "step": 4261, + "time_per_iteration": 2.6825015544891357 + }, + { + "auxiliary_loss_clip": 0.01204677, + "auxiliary_loss_mlp": 0.0103161, + "balance_loss_clip": 1.05128837, + "balance_loss_mlp": 1.02340531, + "epoch": 0.5124751999038056, + "flos": 20333375804160.0, + "grad_norm": 3.723204724368576, + "language_loss": 0.80336761, + "learning_rate": 2.0163582221996213e-06, + "loss": 0.8257305, + "num_input_tokens_seen": 91984700, + "step": 4262, + "time_per_iteration": 2.7075703144073486 + }, + { + "auxiliary_loss_clip": 0.01204332, + "auxiliary_loss_mlp": 0.01036057, + "balance_loss_clip": 1.05271792, + "balance_loss_mlp": 1.02768862, + "epoch": 0.5125954427944448, + "flos": 39785970211200.0, + "grad_norm": 1.9828129234612581, + "language_loss": 0.6804899, + "learning_rate": 2.015579275387446e-06, + "loss": 0.70289373, + "num_input_tokens_seen": 92010020, + "step": 4263, + "time_per_iteration": 2.8691349029541016 + }, + { + "auxiliary_loss_clip": 0.01190939, + "auxiliary_loss_mlp": 0.01026932, + "balance_loss_clip": 1.05013561, + "balance_loss_mlp": 1.01852214, + "epoch": 0.5127156856850839, + "flos": 29205394358400.0, + "grad_norm": 2.189678209596872, + "language_loss": 0.68469739, + "learning_rate": 2.0148003262119085e-06, + "loss": 0.7068761, + "num_input_tokens_seen": 92030990, + "step": 4264, + "time_per_iteration": 2.718309164047241 + }, + { + "auxiliary_loss_clip": 0.01209544, + "auxiliary_loss_mlp": 0.01031127, + "balance_loss_clip": 1.05006838, + "balance_loss_mlp": 1.02262783, + "epoch": 0.5128359285757229, + "flos": 13553693412480.0, + "grad_norm": 1.9108576427467758, + "language_loss": 0.77154493, + "learning_rate": 2.0140213747911728e-06, + "loss": 0.79395163, + "num_input_tokens_seen": 92049525, + "step": 4265, + "time_per_iteration": 2.685335636138916 + }, + { + "auxiliary_loss_clip": 0.01210282, + "auxiliary_loss_mlp": 0.01027997, + "balance_loss_clip": 1.05172849, + "balance_loss_mlp": 1.01944959, + "epoch": 0.5129561714663621, + "flos": 25192089820800.0, + "grad_norm": 2.2040964426832015, + "language_loss": 0.80843115, + "learning_rate": 2.013242421243406e-06, + "loss": 0.83081394, + "num_input_tokens_seen": 92068430, + "step": 4266, + "time_per_iteration": 2.86452317237854 + }, + { + "auxiliary_loss_clip": 0.01221874, + "auxiliary_loss_mlp": 0.01031184, + "balance_loss_clip": 1.04972124, + "balance_loss_mlp": 1.02327466, + "epoch": 0.5130764143570011, + "flos": 18150223080960.0, + "grad_norm": 1.58564693430018, + "language_loss": 0.79038215, + "learning_rate": 2.012463465686774e-06, + "loss": 0.8129127, + "num_input_tokens_seen": 92088180, + "step": 4267, + "time_per_iteration": 2.798032760620117 + }, + { + "auxiliary_loss_clip": 0.01132416, + "auxiliary_loss_mlp": 0.01001652, + "balance_loss_clip": 1.0248816, + "balance_loss_mlp": 1.00047803, + "epoch": 0.5131966572476402, + "flos": 59794896418560.0, + "grad_norm": 0.8059625880018605, + "language_loss": 0.54784667, + "learning_rate": 2.0116845082394446e-06, + "loss": 0.56918734, + "num_input_tokens_seen": 92153015, + "step": 4268, + "time_per_iteration": 3.2932794094085693 + }, + { + "auxiliary_loss_clip": 0.01192941, + "auxiliary_loss_mlp": 0.01028476, + "balance_loss_clip": 1.04897177, + "balance_loss_mlp": 1.02024508, + "epoch": 0.5133169001382794, + "flos": 18515219132160.0, + "grad_norm": 1.8321251319235827, + "language_loss": 0.78737372, + "learning_rate": 2.0109055490195836e-06, + "loss": 0.80958784, + "num_input_tokens_seen": 92171470, + "step": 4269, + "time_per_iteration": 2.6755690574645996 + }, + { + "auxiliary_loss_clip": 0.01229435, + "auxiliary_loss_mlp": 0.01026483, + "balance_loss_clip": 1.04127312, + "balance_loss_mlp": 1.01823938, + "epoch": 0.5134371430289184, + "flos": 15523537219200.0, + "grad_norm": 1.775010778660789, + "language_loss": 0.6405834, + "learning_rate": 2.0101265881453605e-06, + "loss": 0.66314262, + "num_input_tokens_seen": 92189945, + "step": 4270, + "time_per_iteration": 2.7349252700805664 + }, + { + "auxiliary_loss_clip": 0.01195872, + "auxiliary_loss_mlp": 0.01027468, + "balance_loss_clip": 1.05225801, + "balance_loss_mlp": 1.01978493, + "epoch": 0.5135573859195575, + "flos": 21433786911360.0, + "grad_norm": 1.99090513606824, + "language_loss": 0.78335178, + "learning_rate": 2.009347625734941e-06, + "loss": 0.80558515, + "num_input_tokens_seen": 92209855, + "step": 4271, + "time_per_iteration": 2.7085490226745605 + }, + { + "auxiliary_loss_clip": 0.01187922, + "auxiliary_loss_mlp": 0.01033946, + "balance_loss_clip": 1.0561378, + "balance_loss_mlp": 1.02554774, + "epoch": 0.5136776288101966, + "flos": 17712651600000.0, + "grad_norm": 2.326424931285016, + "language_loss": 0.74965882, + "learning_rate": 2.0085686619064954e-06, + "loss": 0.77187753, + "num_input_tokens_seen": 92226295, + "step": 4272, + "time_per_iteration": 2.6980419158935547 + }, + { + "auxiliary_loss_clip": 0.01199109, + "auxiliary_loss_mlp": 0.01029095, + "balance_loss_clip": 1.05487478, + "balance_loss_mlp": 1.0204165, + "epoch": 0.5137978717008357, + "flos": 16581680997120.0, + "grad_norm": 2.5672657226168107, + "language_loss": 0.82755423, + "learning_rate": 2.00778969677819e-06, + "loss": 0.84983629, + "num_input_tokens_seen": 92243330, + "step": 4273, + "time_per_iteration": 2.67531156539917 + }, + { + "auxiliary_loss_clip": 0.01199752, + "auxiliary_loss_mlp": 0.01029393, + "balance_loss_clip": 1.05054522, + "balance_loss_mlp": 1.02073264, + "epoch": 0.5139181145914747, + "flos": 20668243322880.0, + "grad_norm": 1.941033206805366, + "language_loss": 0.64206553, + "learning_rate": 2.0070107304681934e-06, + "loss": 0.66435701, + "num_input_tokens_seen": 92262285, + "step": 4274, + "time_per_iteration": 4.554033041000366 + }, + { + "auxiliary_loss_clip": 0.01211847, + "auxiliary_loss_mlp": 0.01031798, + "balance_loss_clip": 1.05142665, + "balance_loss_mlp": 1.02344716, + "epoch": 0.5140383574821139, + "flos": 32926996546560.0, + "grad_norm": 1.772214370494794, + "language_loss": 0.78416544, + "learning_rate": 2.006231763094675e-06, + "loss": 0.80660188, + "num_input_tokens_seen": 92283305, + "step": 4275, + "time_per_iteration": 3.7094359397888184 + }, + { + "auxiliary_loss_clip": 0.01200795, + "auxiliary_loss_mlp": 0.01027318, + "balance_loss_clip": 1.05593622, + "balance_loss_mlp": 1.01862144, + "epoch": 0.514158600372753, + "flos": 19537093152000.0, + "grad_norm": 1.8665000956380458, + "language_loss": 0.87623155, + "learning_rate": 2.0054527947758027e-06, + "loss": 0.89851272, + "num_input_tokens_seen": 92302105, + "step": 4276, + "time_per_iteration": 2.6500961780548096 + }, + { + "auxiliary_loss_clip": 0.01084596, + "auxiliary_loss_mlp": 0.01002838, + "balance_loss_clip": 1.01413345, + "balance_loss_mlp": 1.0015744, + "epoch": 0.514278843263392, + "flos": 62523855279360.0, + "grad_norm": 0.7217479365856613, + "language_loss": 0.55932343, + "learning_rate": 2.004673825629746e-06, + "loss": 0.58019775, + "num_input_tokens_seen": 92362885, + "step": 4277, + "time_per_iteration": 3.176731824874878 + }, + { + "auxiliary_loss_clip": 0.01192732, + "auxiliary_loss_mlp": 0.01024528, + "balance_loss_clip": 1.04903984, + "balance_loss_mlp": 1.01616526, + "epoch": 0.5143990861540312, + "flos": 25882328545920.0, + "grad_norm": 1.7647482408744621, + "language_loss": 0.72506988, + "learning_rate": 2.0038948557746744e-06, + "loss": 0.74724245, + "num_input_tokens_seen": 92384740, + "step": 4278, + "time_per_iteration": 2.6899330615997314 + }, + { + "auxiliary_loss_clip": 0.011895, + "auxiliary_loss_mlp": 0.01030127, + "balance_loss_clip": 1.05190778, + "balance_loss_mlp": 1.02166891, + "epoch": 0.5145193290446702, + "flos": 23330660238720.0, + "grad_norm": 2.8039472601653332, + "language_loss": 0.75061083, + "learning_rate": 2.0031158853287558e-06, + "loss": 0.77280712, + "num_input_tokens_seen": 92405175, + "step": 4279, + "time_per_iteration": 2.665022134780884 + }, + { + "auxiliary_loss_clip": 0.01198864, + "auxiliary_loss_mlp": 0.01026308, + "balance_loss_clip": 1.0507443, + "balance_loss_mlp": 1.01809418, + "epoch": 0.5146395719353093, + "flos": 22856603518080.0, + "grad_norm": 2.540624832250022, + "language_loss": 0.70609075, + "learning_rate": 2.0023369144101593e-06, + "loss": 0.72834247, + "num_input_tokens_seen": 92423345, + "step": 4280, + "time_per_iteration": 3.5757317543029785 + }, + { + "auxiliary_loss_clip": 0.01189383, + "auxiliary_loss_mlp": 0.01024984, + "balance_loss_clip": 1.04748785, + "balance_loss_mlp": 1.01739037, + "epoch": 0.5147598148259485, + "flos": 26391577616640.0, + "grad_norm": 1.7906013693314988, + "language_loss": 0.76733994, + "learning_rate": 2.0015579431370555e-06, + "loss": 0.78948367, + "num_input_tokens_seen": 92445025, + "step": 4281, + "time_per_iteration": 2.7023661136627197 + }, + { + "auxiliary_loss_clip": 0.01190977, + "auxiliary_loss_mlp": 0.01030606, + "balance_loss_clip": 1.05337429, + "balance_loss_mlp": 1.02294064, + "epoch": 0.5148800577165875, + "flos": 29965694561280.0, + "grad_norm": 2.0897383819107596, + "language_loss": 0.6964767, + "learning_rate": 2.000778971627612e-06, + "loss": 0.71869248, + "num_input_tokens_seen": 92464490, + "step": 4282, + "time_per_iteration": 2.715644121170044 + }, + { + "auxiliary_loss_clip": 0.01189578, + "auxiliary_loss_mlp": 0.01032283, + "balance_loss_clip": 1.04792905, + "balance_loss_mlp": 1.02389646, + "epoch": 0.5150003006072266, + "flos": 17931383470080.0, + "grad_norm": 2.0201711943116227, + "language_loss": 0.89963877, + "learning_rate": 2e-06, + "loss": 0.92185742, + "num_input_tokens_seen": 92482085, + "step": 4283, + "time_per_iteration": 2.6211423873901367 + }, + { + "auxiliary_loss_clip": 0.01183137, + "auxiliary_loss_mlp": 0.01022847, + "balance_loss_clip": 1.05409312, + "balance_loss_mlp": 1.0149312, + "epoch": 0.5151205434978657, + "flos": 18478733892480.0, + "grad_norm": 1.8397448714316347, + "language_loss": 0.85424304, + "learning_rate": 1.9992210283723878e-06, + "loss": 0.87630284, + "num_input_tokens_seen": 92499325, + "step": 4284, + "time_per_iteration": 2.644788980484009 + }, + { + "auxiliary_loss_clip": 0.01179901, + "auxiliary_loss_mlp": 0.01033426, + "balance_loss_clip": 1.05310166, + "balance_loss_mlp": 1.02601695, + "epoch": 0.5152407863885048, + "flos": 25341263003520.0, + "grad_norm": 1.589607054437867, + "language_loss": 0.79467469, + "learning_rate": 1.9984420568629448e-06, + "loss": 0.81680799, + "num_input_tokens_seen": 92522090, + "step": 4285, + "time_per_iteration": 2.6095128059387207 + }, + { + "auxiliary_loss_clip": 0.01192013, + "auxiliary_loss_mlp": 0.01023344, + "balance_loss_clip": 1.05044866, + "balance_loss_mlp": 1.01455259, + "epoch": 0.5153610292791438, + "flos": 18329740277760.0, + "grad_norm": 2.1534942881994374, + "language_loss": 0.78091216, + "learning_rate": 1.9976630855898405e-06, + "loss": 0.80306572, + "num_input_tokens_seen": 92539845, + "step": 4286, + "time_per_iteration": 2.6656336784362793 + }, + { + "auxiliary_loss_clip": 0.01193123, + "auxiliary_loss_mlp": 0.01032427, + "balance_loss_clip": 1.04611588, + "balance_loss_mlp": 1.02469063, + "epoch": 0.515481272169783, + "flos": 30409945971840.0, + "grad_norm": 2.0913847864497956, + "language_loss": 0.74889541, + "learning_rate": 1.9968841146712445e-06, + "loss": 0.77115089, + "num_input_tokens_seen": 92559460, + "step": 4287, + "time_per_iteration": 2.7978436946868896 + }, + { + "auxiliary_loss_clip": 0.01228047, + "auxiliary_loss_mlp": 0.01379502, + "balance_loss_clip": 1.04883492, + "balance_loss_mlp": 1.00020504, + "epoch": 0.5156015150604221, + "flos": 23037305863680.0, + "grad_norm": 1.5817612451838596, + "language_loss": 0.71196866, + "learning_rate": 1.996105144225326e-06, + "loss": 0.73804408, + "num_input_tokens_seen": 92579695, + "step": 4288, + "time_per_iteration": 2.8785789012908936 + }, + { + "auxiliary_loss_clip": 0.011917, + "auxiliary_loss_mlp": 0.01027268, + "balance_loss_clip": 1.05354691, + "balance_loss_mlp": 1.01961446, + "epoch": 0.5157217579510611, + "flos": 17858556645120.0, + "grad_norm": 1.9952748110772003, + "language_loss": 0.78688782, + "learning_rate": 1.995326174370254e-06, + "loss": 0.8090775, + "num_input_tokens_seen": 92598795, + "step": 4289, + "time_per_iteration": 2.638683557510376 + }, + { + "auxiliary_loss_clip": 0.01191959, + "auxiliary_loss_mlp": 0.01378339, + "balance_loss_clip": 1.05351436, + "balance_loss_mlp": 1.00010514, + "epoch": 0.5158420008417003, + "flos": 19171486569600.0, + "grad_norm": 1.510247531974646, + "language_loss": 0.72869873, + "learning_rate": 1.994547205224197e-06, + "loss": 0.75440174, + "num_input_tokens_seen": 92617700, + "step": 4290, + "time_per_iteration": 2.678252696990967 + }, + { + "auxiliary_loss_clip": 0.01198398, + "auxiliary_loss_mlp": 0.01021294, + "balance_loss_clip": 1.05200493, + "balance_loss_mlp": 1.0134027, + "epoch": 0.5159622437323393, + "flos": 22419534827520.0, + "grad_norm": 1.8378209126736835, + "language_loss": 0.67418879, + "learning_rate": 1.993768236905325e-06, + "loss": 0.69638574, + "num_input_tokens_seen": 92638370, + "step": 4291, + "time_per_iteration": 2.664149761199951 + }, + { + "auxiliary_loss_clip": 0.01195851, + "auxiliary_loss_mlp": 0.01029964, + "balance_loss_clip": 1.04923058, + "balance_loss_mlp": 1.02120209, + "epoch": 0.5160824866229784, + "flos": 24603010773120.0, + "grad_norm": 4.448344070774574, + "language_loss": 0.66200769, + "learning_rate": 1.992989269531807e-06, + "loss": 0.68426585, + "num_input_tokens_seen": 92657180, + "step": 4292, + "time_per_iteration": 2.7871482372283936 + }, + { + "auxiliary_loss_clip": 0.01197621, + "auxiliary_loss_mlp": 0.01029541, + "balance_loss_clip": 1.04863787, + "balance_loss_mlp": 1.02131534, + "epoch": 0.5162027295136175, + "flos": 18002737837440.0, + "grad_norm": 2.286570247286406, + "language_loss": 0.68097115, + "learning_rate": 1.99221030322181e-06, + "loss": 0.70324278, + "num_input_tokens_seen": 92673985, + "step": 4293, + "time_per_iteration": 2.6694414615631104 + }, + { + "auxiliary_loss_clip": 0.01204661, + "auxiliary_loss_mlp": 0.01032194, + "balance_loss_clip": 1.05138862, + "balance_loss_mlp": 1.02426648, + "epoch": 0.5163229724042566, + "flos": 27344611221120.0, + "grad_norm": 1.6223809659508437, + "language_loss": 0.80794966, + "learning_rate": 1.991431338093505e-06, + "loss": 0.83031815, + "num_input_tokens_seen": 92696340, + "step": 4294, + "time_per_iteration": 2.730187177658081 + }, + { + "auxiliary_loss_clip": 0.01202097, + "auxiliary_loss_mlp": 0.01027197, + "balance_loss_clip": 1.05384588, + "balance_loss_mlp": 1.02040517, + "epoch": 0.5164432152948957, + "flos": 21762764599680.0, + "grad_norm": 1.6593012449769, + "language_loss": 0.79708982, + "learning_rate": 1.9906523742650587e-06, + "loss": 0.81938279, + "num_input_tokens_seen": 92715200, + "step": 4295, + "time_per_iteration": 2.810819149017334 + }, + { + "auxiliary_loss_clip": 0.01181527, + "auxiliary_loss_mlp": 0.01030718, + "balance_loss_clip": 1.05071855, + "balance_loss_mlp": 1.02233744, + "epoch": 0.5165634581855347, + "flos": 25550334115200.0, + "grad_norm": 2.0602955112077854, + "language_loss": 0.77388704, + "learning_rate": 1.9898734118546397e-06, + "loss": 0.79600948, + "num_input_tokens_seen": 92735150, + "step": 4296, + "time_per_iteration": 2.7022006511688232 + }, + { + "auxiliary_loss_clip": 0.01240902, + "auxiliary_loss_mlp": 0.01030937, + "balance_loss_clip": 1.0463599, + "balance_loss_mlp": 1.02209139, + "epoch": 0.5166837010761739, + "flos": 19901191363200.0, + "grad_norm": 1.4825255754448232, + "language_loss": 0.80630744, + "learning_rate": 1.989094450980416e-06, + "loss": 0.8290258, + "num_input_tokens_seen": 92755250, + "step": 4297, + "time_per_iteration": 2.971190929412842 + }, + { + "auxiliary_loss_clip": 0.01187699, + "auxiliary_loss_mlp": 0.01025553, + "balance_loss_clip": 1.05113626, + "balance_loss_mlp": 1.01751864, + "epoch": 0.516803943966813, + "flos": 26646076454400.0, + "grad_norm": 1.965432249982437, + "language_loss": 0.7652548, + "learning_rate": 1.9883154917605556e-06, + "loss": 0.78738737, + "num_input_tokens_seen": 92774460, + "step": 4298, + "time_per_iteration": 3.1732699871063232 + }, + { + "auxiliary_loss_clip": 0.01181534, + "auxiliary_loss_mlp": 0.01023109, + "balance_loss_clip": 1.05314946, + "balance_loss_mlp": 1.01534867, + "epoch": 0.516924186857452, + "flos": 19682854542720.0, + "grad_norm": 1.7168210619883484, + "language_loss": 0.83327591, + "learning_rate": 1.9875365343132262e-06, + "loss": 0.85532236, + "num_input_tokens_seen": 92791580, + "step": 4299, + "time_per_iteration": 2.6553409099578857 + }, + { + "auxiliary_loss_clip": 0.01189573, + "auxiliary_loss_mlp": 0.01378734, + "balance_loss_clip": 1.05194855, + "balance_loss_mlp": 1.00013995, + "epoch": 0.5170444297480912, + "flos": 15956583586560.0, + "grad_norm": 2.117435765714074, + "language_loss": 0.85146099, + "learning_rate": 1.9867575787565946e-06, + "loss": 0.8771441, + "num_input_tokens_seen": 92806240, + "step": 4300, + "time_per_iteration": 3.562592029571533 + }, + { + "auxiliary_loss_clip": 0.01193418, + "auxiliary_loss_mlp": 0.010252, + "balance_loss_clip": 1.05329597, + "balance_loss_mlp": 1.01745749, + "epoch": 0.5171646726387302, + "flos": 14174157968640.0, + "grad_norm": 1.872063125186999, + "language_loss": 0.85938954, + "learning_rate": 1.9859786252088275e-06, + "loss": 0.8815757, + "num_input_tokens_seen": 92823420, + "step": 4301, + "time_per_iteration": 3.5056488513946533 + }, + { + "auxiliary_loss_clip": 0.01210859, + "auxiliary_loss_mlp": 0.01029319, + "balance_loss_clip": 1.0510385, + "balance_loss_mlp": 1.02071166, + "epoch": 0.5172849155293693, + "flos": 23578550974080.0, + "grad_norm": 2.259316388104983, + "language_loss": 0.6678226, + "learning_rate": 1.9851996737880914e-06, + "loss": 0.69022441, + "num_input_tokens_seen": 92838605, + "step": 4302, + "time_per_iteration": 2.772366523742676 + }, + { + "auxiliary_loss_clip": 0.0119538, + "auxiliary_loss_mlp": 0.01031079, + "balance_loss_clip": 1.05130339, + "balance_loss_mlp": 1.02215624, + "epoch": 0.5174051584200084, + "flos": 14283541860480.0, + "grad_norm": 1.9215279229089768, + "language_loss": 0.74501479, + "learning_rate": 1.9844207246125537e-06, + "loss": 0.76727939, + "num_input_tokens_seen": 92855185, + "step": 4303, + "time_per_iteration": 2.600714683532715 + }, + { + "auxiliary_loss_clip": 0.01196246, + "auxiliary_loss_mlp": 0.01022867, + "balance_loss_clip": 1.050246, + "balance_loss_mlp": 1.01559234, + "epoch": 0.5175254013106475, + "flos": 37889384192640.0, + "grad_norm": 1.8631522247807444, + "language_loss": 0.68745595, + "learning_rate": 1.983641777800379e-06, + "loss": 0.70964706, + "num_input_tokens_seen": 92877830, + "step": 4304, + "time_per_iteration": 2.8562424182891846 + }, + { + "auxiliary_loss_clip": 0.01100576, + "auxiliary_loss_mlp": 0.01001794, + "balance_loss_clip": 1.01606894, + "balance_loss_mlp": 1.00057817, + "epoch": 0.5176456442012866, + "flos": 68549737829760.0, + "grad_norm": 0.746284884237278, + "language_loss": 0.58769381, + "learning_rate": 1.9828628334697343e-06, + "loss": 0.60871756, + "num_input_tokens_seen": 92945040, + "step": 4305, + "time_per_iteration": 3.410823345184326 + }, + { + "auxiliary_loss_clip": 0.01100617, + "auxiliary_loss_mlp": 0.01001942, + "balance_loss_clip": 1.01457262, + "balance_loss_mlp": 1.00066018, + "epoch": 0.5177658870919257, + "flos": 64084137235200.0, + "grad_norm": 0.7683907705434665, + "language_loss": 0.54714936, + "learning_rate": 1.982083891738784e-06, + "loss": 0.56817496, + "num_input_tokens_seen": 93005910, + "step": 4306, + "time_per_iteration": 4.320364952087402 + }, + { + "auxiliary_loss_clip": 0.01197557, + "auxiliary_loss_mlp": 0.01029798, + "balance_loss_clip": 1.05461383, + "balance_loss_mlp": 1.02237725, + "epoch": 0.5178861299825648, + "flos": 26651248012800.0, + "grad_norm": 1.5548716876261888, + "language_loss": 0.82836044, + "learning_rate": 1.9813049527256923e-06, + "loss": 0.85063398, + "num_input_tokens_seen": 93026305, + "step": 4307, + "time_per_iteration": 2.7884199619293213 + }, + { + "auxiliary_loss_clip": 0.01203685, + "auxiliary_loss_mlp": 0.01028534, + "balance_loss_clip": 1.04536211, + "balance_loss_mlp": 1.02041566, + "epoch": 0.5180063728732038, + "flos": 17931886260480.0, + "grad_norm": 2.315132889811791, + "language_loss": 0.82268018, + "learning_rate": 1.9805260165486252e-06, + "loss": 0.84500241, + "num_input_tokens_seen": 93045675, + "step": 4308, + "time_per_iteration": 2.737682819366455 + }, + { + "auxiliary_loss_clip": 0.01189956, + "auxiliary_loss_mlp": 0.01024436, + "balance_loss_clip": 1.05170369, + "balance_loss_mlp": 1.01646113, + "epoch": 0.518126615763843, + "flos": 19500895221120.0, + "grad_norm": 4.737068429183804, + "language_loss": 0.86637866, + "learning_rate": 1.9797470833257457e-06, + "loss": 0.88852251, + "num_input_tokens_seen": 93065375, + "step": 4309, + "time_per_iteration": 2.6348326206207275 + }, + { + "auxiliary_loss_clip": 0.01193409, + "auxiliary_loss_mlp": 0.01028716, + "balance_loss_clip": 1.05511689, + "balance_loss_mlp": 1.02017498, + "epoch": 0.5182468586544821, + "flos": 20704082117760.0, + "grad_norm": 2.1407413336592693, + "language_loss": 0.77430362, + "learning_rate": 1.9789681531752177e-06, + "loss": 0.79652488, + "num_input_tokens_seen": 93085595, + "step": 4310, + "time_per_iteration": 2.9379937648773193 + }, + { + "auxiliary_loss_clip": 0.01214698, + "auxiliary_loss_mlp": 0.01024852, + "balance_loss_clip": 1.0473876, + "balance_loss_mlp": 1.01719904, + "epoch": 0.5183671015451211, + "flos": 23112107936640.0, + "grad_norm": 1.4710381415634026, + "language_loss": 0.72590196, + "learning_rate": 1.978189226215204e-06, + "loss": 0.74829751, + "num_input_tokens_seen": 93106140, + "step": 4311, + "time_per_iteration": 2.7119178771972656 + }, + { + "auxiliary_loss_clip": 0.01182376, + "auxiliary_loss_mlp": 0.01025474, + "balance_loss_clip": 1.05356371, + "balance_loss_mlp": 1.01778495, + "epoch": 0.5184873444357603, + "flos": 17597090568960.0, + "grad_norm": 1.743792689348917, + "language_loss": 0.77255034, + "learning_rate": 1.9774103025638675e-06, + "loss": 0.79462886, + "num_input_tokens_seen": 93124265, + "step": 4312, + "time_per_iteration": 2.6239662170410156 + }, + { + "auxiliary_loss_clip": 0.01227941, + "auxiliary_loss_mlp": 0.01035929, + "balance_loss_clip": 1.05711365, + "balance_loss_mlp": 1.02752519, + "epoch": 0.5186075873263993, + "flos": 24936800883840.0, + "grad_norm": 1.6906777708642684, + "language_loss": 0.7646212, + "learning_rate": 1.9766313823393696e-06, + "loss": 0.78725994, + "num_input_tokens_seen": 93145130, + "step": 4313, + "time_per_iteration": 2.7542200088500977 + }, + { + "auxiliary_loss_clip": 0.01205709, + "auxiliary_loss_mlp": 0.0103012, + "balance_loss_clip": 1.04374528, + "balance_loss_mlp": 1.02212715, + "epoch": 0.5187278302170384, + "flos": 15190106244480.0, + "grad_norm": 2.0295375814249366, + "language_loss": 0.69450259, + "learning_rate": 1.975852465659873e-06, + "loss": 0.71686089, + "num_input_tokens_seen": 93161110, + "step": 4314, + "time_per_iteration": 2.8775477409362793 + }, + { + "auxiliary_loss_clip": 0.01195949, + "auxiliary_loss_mlp": 0.01031114, + "balance_loss_clip": 1.0554533, + "balance_loss_mlp": 1.02234042, + "epoch": 0.5188480731076776, + "flos": 25009412227200.0, + "grad_norm": 2.3173121315382392, + "language_loss": 0.69975805, + "learning_rate": 1.9750735526435377e-06, + "loss": 0.72202873, + "num_input_tokens_seen": 93178055, + "step": 4315, + "time_per_iteration": 2.650116443634033 + }, + { + "auxiliary_loss_clip": 0.01197186, + "auxiliary_loss_mlp": 0.0103235, + "balance_loss_clip": 1.04958498, + "balance_loss_mlp": 1.02388024, + "epoch": 0.5189683159983166, + "flos": 24790141653120.0, + "grad_norm": 3.056469458789006, + "language_loss": 0.79330254, + "learning_rate": 1.974294643408525e-06, + "loss": 0.81559795, + "num_input_tokens_seen": 93195850, + "step": 4316, + "time_per_iteration": 2.7023253440856934 + }, + { + "auxiliary_loss_clip": 0.01195505, + "auxiliary_loss_mlp": 0.01031842, + "balance_loss_clip": 1.05189431, + "balance_loss_mlp": 1.02364647, + "epoch": 0.5190885588889557, + "flos": 24754266944640.0, + "grad_norm": 2.2000888338505935, + "language_loss": 0.66709113, + "learning_rate": 1.9735157380729947e-06, + "loss": 0.68936455, + "num_input_tokens_seen": 93216260, + "step": 4317, + "time_per_iteration": 2.674159288406372 + }, + { + "auxiliary_loss_clip": 0.01201071, + "auxiliary_loss_mlp": 0.01026725, + "balance_loss_clip": 1.04946709, + "balance_loss_mlp": 1.01958466, + "epoch": 0.5192088017795948, + "flos": 24712646060160.0, + "grad_norm": 2.774468723991424, + "language_loss": 0.84131974, + "learning_rate": 1.9727368367551053e-06, + "loss": 0.86359763, + "num_input_tokens_seen": 93234810, + "step": 4318, + "time_per_iteration": 2.7231884002685547 + }, + { + "auxiliary_loss_clip": 0.0118327, + "auxiliary_loss_mlp": 0.01032031, + "balance_loss_clip": 1.04639339, + "balance_loss_mlp": 1.02391267, + "epoch": 0.5193290446702339, + "flos": 27229588894080.0, + "grad_norm": 1.924536508514286, + "language_loss": 0.68554378, + "learning_rate": 1.9719579395730164e-06, + "loss": 0.7076968, + "num_input_tokens_seen": 93254185, + "step": 4319, + "time_per_iteration": 2.7502522468566895 + }, + { + "auxiliary_loss_clip": 0.01187484, + "auxiliary_loss_mlp": 0.01032616, + "balance_loss_clip": 1.05713511, + "balance_loss_mlp": 1.02455187, + "epoch": 0.5194492875608729, + "flos": 11473352392320.0, + "grad_norm": 2.383943361687713, + "language_loss": 0.93663788, + "learning_rate": 1.9711790466448854e-06, + "loss": 0.95883888, + "num_input_tokens_seen": 93268205, + "step": 4320, + "time_per_iteration": 2.5842535495758057 + }, + { + "auxiliary_loss_clip": 0.01219051, + "auxiliary_loss_mlp": 0.01031905, + "balance_loss_clip": 1.04744613, + "balance_loss_mlp": 1.02297592, + "epoch": 0.5195695304515121, + "flos": 20338906498560.0, + "grad_norm": 3.153869998655887, + "language_loss": 0.71332318, + "learning_rate": 1.9704001580888704e-06, + "loss": 0.73583269, + "num_input_tokens_seen": 93286945, + "step": 4321, + "time_per_iteration": 2.7469112873077393 + }, + { + "auxiliary_loss_clip": 0.0119313, + "auxiliary_loss_mlp": 0.01378793, + "balance_loss_clip": 1.0480969, + "balance_loss_mlp": 1.000067, + "epoch": 0.5196897733421512, + "flos": 20048317470720.0, + "grad_norm": 1.9500215705572246, + "language_loss": 0.8685565, + "learning_rate": 1.9696212740231283e-06, + "loss": 0.89427578, + "num_input_tokens_seen": 93305595, + "step": 4322, + "time_per_iteration": 2.691466808319092 + }, + { + "auxiliary_loss_clip": 0.01199906, + "auxiliary_loss_mlp": 0.01029934, + "balance_loss_clip": 1.05325699, + "balance_loss_mlp": 1.02159548, + "epoch": 0.5198100162327902, + "flos": 23805507058560.0, + "grad_norm": 2.227544507137627, + "language_loss": 0.82053423, + "learning_rate": 1.9688423945658146e-06, + "loss": 0.84283262, + "num_input_tokens_seen": 93326460, + "step": 4323, + "time_per_iteration": 2.6921041011810303 + }, + { + "auxiliary_loss_clip": 0.01203032, + "auxiliary_loss_mlp": 0.01034351, + "balance_loss_clip": 1.04117954, + "balance_loss_mlp": 1.02567208, + "epoch": 0.5199302591234293, + "flos": 24023951619840.0, + "grad_norm": 3.821460458683394, + "language_loss": 0.71893984, + "learning_rate": 1.9680635198350845e-06, + "loss": 0.7413137, + "num_input_tokens_seen": 93346170, + "step": 4324, + "time_per_iteration": 2.791393280029297 + }, + { + "auxiliary_loss_clip": 0.0119059, + "auxiliary_loss_mlp": 0.01030611, + "balance_loss_clip": 1.05055499, + "balance_loss_mlp": 1.02201569, + "epoch": 0.5200505020140684, + "flos": 26359366095360.0, + "grad_norm": 1.8148603423299647, + "language_loss": 0.72221565, + "learning_rate": 1.967284649949093e-06, + "loss": 0.74442768, + "num_input_tokens_seen": 93365380, + "step": 4325, + "time_per_iteration": 2.7256412506103516 + }, + { + "auxiliary_loss_clip": 0.01205356, + "auxiliary_loss_mlp": 0.01030809, + "balance_loss_clip": 1.04673529, + "balance_loss_mlp": 1.02205324, + "epoch": 0.5201707449047075, + "flos": 39604262284800.0, + "grad_norm": 2.0045561474689744, + "language_loss": 0.72593379, + "learning_rate": 1.966505785025994e-06, + "loss": 0.74829543, + "num_input_tokens_seen": 93387285, + "step": 4326, + "time_per_iteration": 3.8169000148773193 + }, + { + "auxiliary_loss_clip": 0.01210084, + "auxiliary_loss_mlp": 0.01025887, + "balance_loss_clip": 1.05062461, + "balance_loss_mlp": 1.01789999, + "epoch": 0.5202909877953465, + "flos": 53682788292480.0, + "grad_norm": 1.7332763874681065, + "language_loss": 0.76233673, + "learning_rate": 1.965726925183941e-06, + "loss": 0.7846964, + "num_input_tokens_seen": 93410390, + "step": 4327, + "time_per_iteration": 4.85212254524231 + }, + { + "auxiliary_loss_clip": 0.01182615, + "auxiliary_loss_mlp": 0.01034308, + "balance_loss_clip": 1.05377197, + "balance_loss_mlp": 1.02600503, + "epoch": 0.5204112306859857, + "flos": 19537021324800.0, + "grad_norm": 2.430385330871557, + "language_loss": 0.84819549, + "learning_rate": 1.964948070541087e-06, + "loss": 0.87036473, + "num_input_tokens_seen": 93429050, + "step": 4328, + "time_per_iteration": 2.6054635047912598 + }, + { + "auxiliary_loss_clip": 0.0117785, + "auxiliary_loss_mlp": 0.01027408, + "balance_loss_clip": 1.04796505, + "balance_loss_mlp": 1.01895595, + "epoch": 0.5205314735766248, + "flos": 15304697608320.0, + "grad_norm": 2.3595384610244445, + "language_loss": 0.69779563, + "learning_rate": 1.9641692212155816e-06, + "loss": 0.71984822, + "num_input_tokens_seen": 93446815, + "step": 4329, + "time_per_iteration": 2.584810733795166 + }, + { + "auxiliary_loss_clip": 0.01213371, + "auxiliary_loss_mlp": 0.010247, + "balance_loss_clip": 1.04963112, + "balance_loss_mlp": 1.01677823, + "epoch": 0.5206517164672638, + "flos": 59263701160320.0, + "grad_norm": 2.5923703394384505, + "language_loss": 0.72813272, + "learning_rate": 1.9633903773255777e-06, + "loss": 0.75051343, + "num_input_tokens_seen": 93469130, + "step": 4330, + "time_per_iteration": 2.998823404312134 + }, + { + "auxiliary_loss_clip": 0.01181941, + "auxiliary_loss_mlp": 0.01028713, + "balance_loss_clip": 1.05294871, + "balance_loss_mlp": 1.02099395, + "epoch": 0.520771959357903, + "flos": 26871129118080.0, + "grad_norm": 1.6572187630057038, + "language_loss": 0.74740899, + "learning_rate": 1.9626115389892237e-06, + "loss": 0.76951551, + "num_input_tokens_seen": 93489920, + "step": 4331, + "time_per_iteration": 2.6800925731658936 + }, + { + "auxiliary_loss_clip": 0.01216023, + "auxiliary_loss_mlp": 0.01030795, + "balance_loss_clip": 1.05112314, + "balance_loss_mlp": 1.02256989, + "epoch": 0.520892202248542, + "flos": 26907075653760.0, + "grad_norm": 2.1043302497291583, + "language_loss": 0.85221457, + "learning_rate": 1.96183270632467e-06, + "loss": 0.87468266, + "num_input_tokens_seen": 93509770, + "step": 4332, + "time_per_iteration": 3.679096221923828 + }, + { + "auxiliary_loss_clip": 0.01200496, + "auxiliary_loss_mlp": 0.0137965, + "balance_loss_clip": 1.04675245, + "balance_loss_mlp": 1.00007486, + "epoch": 0.5210124451391811, + "flos": 25849434666240.0, + "grad_norm": 1.5963261335262267, + "language_loss": 0.79089797, + "learning_rate": 1.9610538794500644e-06, + "loss": 0.81669945, + "num_input_tokens_seen": 93529320, + "step": 4333, + "time_per_iteration": 2.7087597846984863 + }, + { + "auxiliary_loss_clip": 0.01114388, + "auxiliary_loss_mlp": 0.01002109, + "balance_loss_clip": 1.01733589, + "balance_loss_mlp": 1.0008986, + "epoch": 0.5211326880298203, + "flos": 70553804319360.0, + "grad_norm": 0.7721693952852219, + "language_loss": 0.59468198, + "learning_rate": 1.9602750584835542e-06, + "loss": 0.61584699, + "num_input_tokens_seen": 93595255, + "step": 4334, + "time_per_iteration": 3.377195358276367 + }, + { + "auxiliary_loss_clip": 0.011938, + "auxiliary_loss_mlp": 0.01028196, + "balance_loss_clip": 1.04755235, + "balance_loss_mlp": 1.02038193, + "epoch": 0.5212529309204593, + "flos": 15628898787840.0, + "grad_norm": 2.4788958643889245, + "language_loss": 0.82613206, + "learning_rate": 1.959496243543286e-06, + "loss": 0.84835196, + "num_input_tokens_seen": 93613135, + "step": 4335, + "time_per_iteration": 2.7142210006713867 + }, + { + "auxiliary_loss_clip": 0.01195087, + "auxiliary_loss_mlp": 0.01028444, + "balance_loss_clip": 1.05568218, + "balance_loss_mlp": 1.02027249, + "epoch": 0.5213731738110984, + "flos": 26242655829120.0, + "grad_norm": 2.0255153054918056, + "language_loss": 0.79338461, + "learning_rate": 1.9587174347474057e-06, + "loss": 0.81561995, + "num_input_tokens_seen": 93629645, + "step": 4336, + "time_per_iteration": 2.686155080795288 + }, + { + "auxiliary_loss_clip": 0.01197966, + "auxiliary_loss_mlp": 0.01026871, + "balance_loss_clip": 1.04415941, + "balance_loss_mlp": 1.01844311, + "epoch": 0.5214934167017375, + "flos": 19418407637760.0, + "grad_norm": 2.164112858339135, + "language_loss": 0.81954944, + "learning_rate": 1.9579386322140574e-06, + "loss": 0.84179771, + "num_input_tokens_seen": 93645325, + "step": 4337, + "time_per_iteration": 2.685662031173706 + }, + { + "auxiliary_loss_clip": 0.01185989, + "auxiliary_loss_mlp": 0.01379133, + "balance_loss_clip": 1.05494416, + "balance_loss_mlp": 1.00017357, + "epoch": 0.5216136595923766, + "flos": 30955788023040.0, + "grad_norm": 1.679191041689763, + "language_loss": 0.80875671, + "learning_rate": 1.9571598360613854e-06, + "loss": 0.83440793, + "num_input_tokens_seen": 93668200, + "step": 4338, + "time_per_iteration": 2.6689114570617676 + }, + { + "auxiliary_loss_clip": 0.01181182, + "auxiliary_loss_mlp": 0.01023128, + "balance_loss_clip": 1.04596829, + "balance_loss_mlp": 1.01454544, + "epoch": 0.5217339024830157, + "flos": 21945047143680.0, + "grad_norm": 3.3768143166115236, + "language_loss": 0.69669449, + "learning_rate": 1.956381046407532e-06, + "loss": 0.7187376, + "num_input_tokens_seen": 93688495, + "step": 4339, + "time_per_iteration": 2.7317206859588623 + }, + { + "auxiliary_loss_clip": 0.01208308, + "auxiliary_loss_mlp": 0.01029082, + "balance_loss_clip": 1.0507412, + "balance_loss_mlp": 1.02121425, + "epoch": 0.5218541453736548, + "flos": 20923209037440.0, + "grad_norm": 1.745179772236563, + "language_loss": 0.86471814, + "learning_rate": 1.9556022633706394e-06, + "loss": 0.88709199, + "num_input_tokens_seen": 93707285, + "step": 4340, + "time_per_iteration": 2.7079007625579834 + }, + { + "auxiliary_loss_clip": 0.01191308, + "auxiliary_loss_mlp": 0.01029499, + "balance_loss_clip": 1.0494175, + "balance_loss_mlp": 1.02104068, + "epoch": 0.5219743882642939, + "flos": 23951663498880.0, + "grad_norm": 2.301515176757384, + "language_loss": 0.80061567, + "learning_rate": 1.954823487068848e-06, + "loss": 0.82282364, + "num_input_tokens_seen": 93727495, + "step": 4341, + "time_per_iteration": 2.6967830657958984 + }, + { + "auxiliary_loss_clip": 0.01193186, + "auxiliary_loss_mlp": 0.01024576, + "balance_loss_clip": 1.05591536, + "balance_loss_mlp": 1.01682186, + "epoch": 0.5220946311549329, + "flos": 28799280213120.0, + "grad_norm": 1.8495154883329026, + "language_loss": 0.811122, + "learning_rate": 1.9540447176202976e-06, + "loss": 0.83329958, + "num_input_tokens_seen": 93748740, + "step": 4342, + "time_per_iteration": 2.6991987228393555 + }, + { + "auxiliary_loss_clip": 0.01089016, + "auxiliary_loss_mlp": 0.01002935, + "balance_loss_clip": 1.02042162, + "balance_loss_mlp": 1.0017786, + "epoch": 0.5222148740455721, + "flos": 67189369017600.0, + "grad_norm": 0.8690887268818329, + "language_loss": 0.60796529, + "learning_rate": 1.9532659551431272e-06, + "loss": 0.62888485, + "num_input_tokens_seen": 93815770, + "step": 4343, + "time_per_iteration": 3.3723278045654297 + }, + { + "auxiliary_loss_clip": 0.01194732, + "auxiliary_loss_mlp": 0.01026842, + "balance_loss_clip": 1.05218291, + "balance_loss_mlp": 1.01825929, + "epoch": 0.5223351169362112, + "flos": 61856164339200.0, + "grad_norm": 1.4900672448998578, + "language_loss": 0.67440915, + "learning_rate": 1.9524871997554744e-06, + "loss": 0.69662488, + "num_input_tokens_seen": 93843530, + "step": 4344, + "time_per_iteration": 2.990633964538574 + }, + { + "auxiliary_loss_clip": 0.01192306, + "auxiliary_loss_mlp": 0.01026433, + "balance_loss_clip": 1.05274045, + "balance_loss_mlp": 1.01821947, + "epoch": 0.5224553598268502, + "flos": 14647388676480.0, + "grad_norm": 2.1776230687102345, + "language_loss": 0.80536795, + "learning_rate": 1.951708451575475e-06, + "loss": 0.8275553, + "num_input_tokens_seen": 93860595, + "step": 4345, + "time_per_iteration": 2.608811855316162 + }, + { + "auxiliary_loss_clip": 0.0121718, + "auxiliary_loss_mlp": 0.01024157, + "balance_loss_clip": 1.04911971, + "balance_loss_mlp": 1.01629496, + "epoch": 0.5225756027174894, + "flos": 14826043946880.0, + "grad_norm": 1.8451586030693665, + "language_loss": 0.82226217, + "learning_rate": 1.9509297107212657e-06, + "loss": 0.84467554, + "num_input_tokens_seen": 93877365, + "step": 4346, + "time_per_iteration": 2.670227527618408 + }, + { + "auxiliary_loss_clip": 0.0118353, + "auxiliary_loss_mlp": 0.01024246, + "balance_loss_clip": 1.05437422, + "balance_loss_mlp": 1.01633072, + "epoch": 0.5226958456081284, + "flos": 23512009029120.0, + "grad_norm": 3.5913229873557997, + "language_loss": 0.79083049, + "learning_rate": 1.95015097731098e-06, + "loss": 0.81290817, + "num_input_tokens_seen": 93896855, + "step": 4347, + "time_per_iteration": 2.5990965366363525 + }, + { + "auxiliary_loss_clip": 0.01185728, + "auxiliary_loss_mlp": 0.01029081, + "balance_loss_clip": 1.05567396, + "balance_loss_mlp": 1.02074206, + "epoch": 0.5228160884987675, + "flos": 19062928690560.0, + "grad_norm": 2.1966536032458124, + "language_loss": 0.81939685, + "learning_rate": 1.949372251462751e-06, + "loss": 0.84154499, + "num_input_tokens_seen": 93914270, + "step": 4348, + "time_per_iteration": 2.578572988510132 + }, + { + "auxiliary_loss_clip": 0.0120986, + "auxiliary_loss_mlp": 0.01378753, + "balance_loss_clip": 1.05116379, + "balance_loss_mlp": 1.0000906, + "epoch": 0.5229363313894067, + "flos": 21063224252160.0, + "grad_norm": 1.8252719670731112, + "language_loss": 0.82952237, + "learning_rate": 1.9485935332947124e-06, + "loss": 0.85540843, + "num_input_tokens_seen": 93932180, + "step": 4349, + "time_per_iteration": 2.7708382606506348 + }, + { + "auxiliary_loss_clip": 0.01197373, + "auxiliary_loss_mlp": 0.0102341, + "balance_loss_clip": 1.05162895, + "balance_loss_mlp": 1.01569974, + "epoch": 0.5230565742800457, + "flos": 14830389492480.0, + "grad_norm": 2.253418129442952, + "language_loss": 0.83594501, + "learning_rate": 1.947814822924993e-06, + "loss": 0.85815287, + "num_input_tokens_seen": 93949690, + "step": 4350, + "time_per_iteration": 2.6853995323181152 + }, + { + "auxiliary_loss_clip": 0.01186145, + "auxiliary_loss_mlp": 0.0102688, + "balance_loss_clip": 1.05775476, + "balance_loss_mlp": 1.01889861, + "epoch": 0.5231768171706848, + "flos": 25813021253760.0, + "grad_norm": 1.84852926898192, + "language_loss": 0.83082092, + "learning_rate": 1.9470361204717236e-06, + "loss": 0.85295111, + "num_input_tokens_seen": 93968830, + "step": 4351, + "time_per_iteration": 2.7150914669036865 + }, + { + "auxiliary_loss_clip": 0.0120751, + "auxiliary_loss_mlp": 0.01379425, + "balance_loss_clip": 1.04742885, + "balance_loss_mlp": 1.00011194, + "epoch": 0.5232970600613239, + "flos": 22743807834240.0, + "grad_norm": 1.5511126466742966, + "language_loss": 0.80878907, + "learning_rate": 1.9462574260530326e-06, + "loss": 0.83465844, + "num_input_tokens_seen": 93989110, + "step": 4352, + "time_per_iteration": 3.6050148010253906 + }, + { + "auxiliary_loss_clip": 0.01179109, + "auxiliary_loss_mlp": 0.01032871, + "balance_loss_clip": 1.04911518, + "balance_loss_mlp": 1.0245086, + "epoch": 0.523417302951963, + "flos": 17310703432320.0, + "grad_norm": 1.7197393030333812, + "language_loss": 0.81111813, + "learning_rate": 1.9454787397870472e-06, + "loss": 0.83323801, + "num_input_tokens_seen": 94006430, + "step": 4353, + "time_per_iteration": 4.4409565925598145 + }, + { + "auxiliary_loss_clip": 0.01214968, + "auxiliary_loss_mlp": 0.01026172, + "balance_loss_clip": 1.04704225, + "balance_loss_mlp": 1.0181073, + "epoch": 0.523537545842602, + "flos": 18551740285440.0, + "grad_norm": 1.8603209493962947, + "language_loss": 0.7151435, + "learning_rate": 1.944700061791894e-06, + "loss": 0.73755491, + "num_input_tokens_seen": 94024825, + "step": 4354, + "time_per_iteration": 2.7213211059570312 + }, + { + "auxiliary_loss_clip": 0.0119125, + "auxiliary_loss_mlp": 0.01027363, + "balance_loss_clip": 1.05283797, + "balance_loss_mlp": 1.01916146, + "epoch": 0.5236577887332411, + "flos": 19719267955200.0, + "grad_norm": 2.597493778950774, + "language_loss": 0.6513375, + "learning_rate": 1.943921392185698e-06, + "loss": 0.67352366, + "num_input_tokens_seen": 94043450, + "step": 4355, + "time_per_iteration": 2.6397671699523926 + }, + { + "auxiliary_loss_clip": 0.01203467, + "auxiliary_loss_mlp": 0.0102923, + "balance_loss_clip": 1.05045831, + "balance_loss_mlp": 1.02160096, + "epoch": 0.5237780316238803, + "flos": 23550218121600.0, + "grad_norm": 2.2551916931935967, + "language_loss": 0.77574116, + "learning_rate": 1.9431427310865814e-06, + "loss": 0.79806817, + "num_input_tokens_seen": 94063055, + "step": 4356, + "time_per_iteration": 2.6805107593536377 + }, + { + "auxiliary_loss_clip": 0.01186758, + "auxiliary_loss_mlp": 0.01033806, + "balance_loss_clip": 1.04448032, + "balance_loss_mlp": 1.02637959, + "epoch": 0.5238982745145193, + "flos": 22491894775680.0, + "grad_norm": 1.584195194587471, + "language_loss": 0.78574902, + "learning_rate": 1.942364078612667e-06, + "loss": 0.80795467, + "num_input_tokens_seen": 94081785, + "step": 4357, + "time_per_iteration": 3.663130044937134 + }, + { + "auxiliary_loss_clip": 0.01216328, + "auxiliary_loss_mlp": 0.01029597, + "balance_loss_clip": 1.04973745, + "balance_loss_mlp": 1.02096009, + "epoch": 0.5240185174051584, + "flos": 27088927234560.0, + "grad_norm": 1.7873769494546674, + "language_loss": 0.75394726, + "learning_rate": 1.9415854348820765e-06, + "loss": 0.77640653, + "num_input_tokens_seen": 94101635, + "step": 4358, + "time_per_iteration": 2.7331273555755615 + }, + { + "auxiliary_loss_clip": 0.01200087, + "auxiliary_loss_mlp": 0.01025489, + "balance_loss_clip": 1.05521715, + "balance_loss_mlp": 1.01698911, + "epoch": 0.5241387602957975, + "flos": 22674680110080.0, + "grad_norm": 2.93675545107719, + "language_loss": 0.68558669, + "learning_rate": 1.940806800012929e-06, + "loss": 0.70784241, + "num_input_tokens_seen": 94121705, + "step": 4359, + "time_per_iteration": 2.592783212661743 + }, + { + "auxiliary_loss_clip": 0.01214197, + "auxiliary_loss_mlp": 0.01379291, + "balance_loss_clip": 1.04912186, + "balance_loss_mlp": 1.00011206, + "epoch": 0.5242590031864366, + "flos": 40553453134080.0, + "grad_norm": 1.4775664399250596, + "language_loss": 0.63405746, + "learning_rate": 1.9400281741233432e-06, + "loss": 0.65999234, + "num_input_tokens_seen": 94146595, + "step": 4360, + "time_per_iteration": 2.8901946544647217 + }, + { + "auxiliary_loss_clip": 0.01110835, + "auxiliary_loss_mlp": 0.00999707, + "balance_loss_clip": 1.02237225, + "balance_loss_mlp": 0.99834824, + "epoch": 0.5243792460770756, + "flos": 66676313105280.0, + "grad_norm": 0.8218269721257189, + "language_loss": 0.52556849, + "learning_rate": 1.939249557331435e-06, + "loss": 0.54667389, + "num_input_tokens_seen": 94212410, + "step": 4361, + "time_per_iteration": 3.299093723297119 + }, + { + "auxiliary_loss_clip": 0.012171, + "auxiliary_loss_mlp": 0.01037547, + "balance_loss_clip": 1.04974377, + "balance_loss_mlp": 1.02922642, + "epoch": 0.5244994889677148, + "flos": 28183663992960.0, + "grad_norm": 1.867186365998483, + "language_loss": 0.73053551, + "learning_rate": 1.938470949755321e-06, + "loss": 0.75308198, + "num_input_tokens_seen": 94232290, + "step": 4362, + "time_per_iteration": 2.7708752155303955 + }, + { + "auxiliary_loss_clip": 0.01111955, + "auxiliary_loss_mlp": 0.01001399, + "balance_loss_clip": 1.01727498, + "balance_loss_mlp": 1.00008774, + "epoch": 0.5246197318583539, + "flos": 65950379239680.0, + "grad_norm": 0.809182170752748, + "language_loss": 0.55660462, + "learning_rate": 1.937692351513115e-06, + "loss": 0.57773817, + "num_input_tokens_seen": 94291285, + "step": 4363, + "time_per_iteration": 3.2425150871276855 + }, + { + "auxiliary_loss_clip": 0.01195635, + "auxiliary_loss_mlp": 0.0103231, + "balance_loss_clip": 1.05135989, + "balance_loss_mlp": 1.0232923, + "epoch": 0.5247399747489929, + "flos": 21033490769280.0, + "grad_norm": 3.9884140872251606, + "language_loss": 0.80091751, + "learning_rate": 1.9369137627229297e-06, + "loss": 0.82319689, + "num_input_tokens_seen": 94309685, + "step": 4364, + "time_per_iteration": 2.6404106616973877 + }, + { + "auxiliary_loss_clip": 0.01192448, + "auxiliary_loss_mlp": 0.0102635, + "balance_loss_clip": 1.05453432, + "balance_loss_mlp": 1.01810694, + "epoch": 0.5248602176396321, + "flos": 19025940660480.0, + "grad_norm": 1.8835507305933723, + "language_loss": 0.88091964, + "learning_rate": 1.936135183502877e-06, + "loss": 0.90310758, + "num_input_tokens_seen": 94326985, + "step": 4365, + "time_per_iteration": 2.626495599746704 + }, + { + "auxiliary_loss_clip": 0.01212883, + "auxiliary_loss_mlp": 0.01032436, + "balance_loss_clip": 1.04939938, + "balance_loss_mlp": 1.02402925, + "epoch": 0.5249804605302711, + "flos": 22200084685440.0, + "grad_norm": 2.055479654960087, + "language_loss": 0.80353677, + "learning_rate": 1.935356613971066e-06, + "loss": 0.82598996, + "num_input_tokens_seen": 94347645, + "step": 4366, + "time_per_iteration": 2.6732654571533203 + }, + { + "auxiliary_loss_clip": 0.01199724, + "auxiliary_loss_mlp": 0.01379147, + "balance_loss_clip": 1.05138791, + "balance_loss_mlp": 1.00002694, + "epoch": 0.5251007034209102, + "flos": 23805686626560.0, + "grad_norm": 1.8172662367086396, + "language_loss": 0.76522851, + "learning_rate": 1.9345780542456047e-06, + "loss": 0.79101717, + "num_input_tokens_seen": 94367020, + "step": 4367, + "time_per_iteration": 2.7669146060943604 + }, + { + "auxiliary_loss_clip": 0.01179492, + "auxiliary_loss_mlp": 0.01029263, + "balance_loss_clip": 1.05043674, + "balance_loss_mlp": 1.02191365, + "epoch": 0.5252209463115494, + "flos": 23294605962240.0, + "grad_norm": 1.856760432877937, + "language_loss": 0.71569324, + "learning_rate": 1.9337995044446007e-06, + "loss": 0.73778081, + "num_input_tokens_seen": 94385860, + "step": 4368, + "time_per_iteration": 2.6309237480163574 + }, + { + "auxiliary_loss_clip": 0.01195455, + "auxiliary_loss_mlp": 0.01029483, + "balance_loss_clip": 1.05211949, + "balance_loss_mlp": 1.02120411, + "epoch": 0.5253411892021884, + "flos": 19828687760640.0, + "grad_norm": 1.9796747264535326, + "language_loss": 0.80011147, + "learning_rate": 1.9330209646861596e-06, + "loss": 0.82236087, + "num_input_tokens_seen": 94405010, + "step": 4369, + "time_per_iteration": 2.587864875793457 + }, + { + "auxiliary_loss_clip": 0.01200893, + "auxiliary_loss_mlp": 0.01028808, + "balance_loss_clip": 1.05193901, + "balance_loss_mlp": 1.02101147, + "epoch": 0.5254614320928275, + "flos": 24133730561280.0, + "grad_norm": 1.887565011534886, + "language_loss": 0.7759378, + "learning_rate": 1.9322424350883843e-06, + "loss": 0.79823476, + "num_input_tokens_seen": 94426845, + "step": 4370, + "time_per_iteration": 2.6553351879119873 + }, + { + "auxiliary_loss_clip": 0.0120058, + "auxiliary_loss_mlp": 0.01027562, + "balance_loss_clip": 1.04907489, + "balance_loss_mlp": 1.01946211, + "epoch": 0.5255816749834666, + "flos": 24644954880000.0, + "grad_norm": 1.5925678053936454, + "language_loss": 0.78540212, + "learning_rate": 1.931463915769379e-06, + "loss": 0.80768359, + "num_input_tokens_seen": 94446960, + "step": 4371, + "time_per_iteration": 2.7145349979400635 + }, + { + "auxiliary_loss_clip": 0.01213711, + "auxiliary_loss_mlp": 0.01031219, + "balance_loss_clip": 1.04692054, + "balance_loss_mlp": 1.02173042, + "epoch": 0.5257019178741057, + "flos": 14136595320960.0, + "grad_norm": 2.331644203319181, + "language_loss": 0.7374295, + "learning_rate": 1.930685406847242e-06, + "loss": 0.75987881, + "num_input_tokens_seen": 94461535, + "step": 4372, + "time_per_iteration": 2.7316737174987793 + }, + { + "auxiliary_loss_clip": 0.01200503, + "auxiliary_loss_mlp": 0.01025159, + "balance_loss_clip": 1.05140066, + "balance_loss_mlp": 1.01732659, + "epoch": 0.5258221607647448, + "flos": 23548961145600.0, + "grad_norm": 1.42736594639593, + "language_loss": 0.81725097, + "learning_rate": 1.9299069084400734e-06, + "loss": 0.83950758, + "num_input_tokens_seen": 94482395, + "step": 4373, + "time_per_iteration": 2.700714111328125 + }, + { + "auxiliary_loss_clip": 0.01202867, + "auxiliary_loss_mlp": 0.01026745, + "balance_loss_clip": 1.04900587, + "balance_loss_mlp": 1.01746416, + "epoch": 0.5259424036553839, + "flos": 24966103403520.0, + "grad_norm": 1.858021540305601, + "language_loss": 0.69353062, + "learning_rate": 1.9291284206659717e-06, + "loss": 0.71582675, + "num_input_tokens_seen": 94500580, + "step": 4374, + "time_per_iteration": 2.714278221130371 + }, + { + "auxiliary_loss_clip": 0.01188472, + "auxiliary_loss_mlp": 0.01030199, + "balance_loss_clip": 1.05661321, + "balance_loss_mlp": 1.02172315, + "epoch": 0.526062646546023, + "flos": 28763908295040.0, + "grad_norm": 1.6960876306266093, + "language_loss": 0.71619594, + "learning_rate": 1.928349943643032e-06, + "loss": 0.73838258, + "num_input_tokens_seen": 94519680, + "step": 4375, + "time_per_iteration": 2.624508857727051 + }, + { + "auxiliary_loss_clip": 0.01191325, + "auxiliary_loss_mlp": 0.0102549, + "balance_loss_clip": 1.05554318, + "balance_loss_mlp": 1.01722884, + "epoch": 0.526182889436662, + "flos": 22821375254400.0, + "grad_norm": 1.6635332066086774, + "language_loss": 0.81853306, + "learning_rate": 1.9275714774893493e-06, + "loss": 0.84070122, + "num_input_tokens_seen": 94539135, + "step": 4376, + "time_per_iteration": 2.6119441986083984 + }, + { + "auxiliary_loss_clip": 0.01195167, + "auxiliary_loss_mlp": 0.01027009, + "balance_loss_clip": 1.04716825, + "balance_loss_mlp": 1.01827669, + "epoch": 0.5263031323273012, + "flos": 22929466256640.0, + "grad_norm": 2.2702724670473287, + "language_loss": 0.73082107, + "learning_rate": 1.9267930223230154e-06, + "loss": 0.75304282, + "num_input_tokens_seen": 94557610, + "step": 4377, + "time_per_iteration": 2.8027265071868896 + }, + { + "auxiliary_loss_clip": 0.01201482, + "auxiliary_loss_mlp": 0.01026554, + "balance_loss_clip": 1.05054069, + "balance_loss_mlp": 1.01897228, + "epoch": 0.5264233752179402, + "flos": 17748634049280.0, + "grad_norm": 2.00275743249663, + "language_loss": 0.78311419, + "learning_rate": 1.9260145782621224e-06, + "loss": 0.80539453, + "num_input_tokens_seen": 94575390, + "step": 4378, + "time_per_iteration": 3.598954916000366 + }, + { + "auxiliary_loss_clip": 0.01200148, + "auxiliary_loss_mlp": 0.01031411, + "balance_loss_clip": 1.05436802, + "balance_loss_mlp": 1.02346623, + "epoch": 0.5265436181085793, + "flos": 24421626069120.0, + "grad_norm": 1.8272815628040076, + "language_loss": 0.88028097, + "learning_rate": 1.925236145424758e-06, + "loss": 0.90259659, + "num_input_tokens_seen": 94594210, + "step": 4379, + "time_per_iteration": 3.620918035507202 + }, + { + "auxiliary_loss_clip": 0.01090673, + "auxiliary_loss_mlp": 0.01001614, + "balance_loss_clip": 1.01895928, + "balance_loss_mlp": 1.0003798, + "epoch": 0.5266638609992185, + "flos": 69207298156800.0, + "grad_norm": 0.6919615333354734, + "language_loss": 0.576087, + "learning_rate": 1.924457723929012e-06, + "loss": 0.5970099, + "num_input_tokens_seen": 94665020, + "step": 4380, + "time_per_iteration": 4.21877646446228 + }, + { + "auxiliary_loss_clip": 0.01195536, + "auxiliary_loss_mlp": 0.01029642, + "balance_loss_clip": 1.0544343, + "balance_loss_mlp": 1.02123785, + "epoch": 0.5267841038898575, + "flos": 20738699850240.0, + "grad_norm": 1.6027498445122854, + "language_loss": 0.82932711, + "learning_rate": 1.9236793138929685e-06, + "loss": 0.85157883, + "num_input_tokens_seen": 94684290, + "step": 4381, + "time_per_iteration": 2.878899574279785 + }, + { + "auxiliary_loss_clip": 0.01195669, + "auxiliary_loss_mlp": 0.01027242, + "balance_loss_clip": 1.05224812, + "balance_loss_mlp": 1.01938629, + "epoch": 0.5269043467804966, + "flos": 17234392988160.0, + "grad_norm": 2.015808840819632, + "language_loss": 0.81118053, + "learning_rate": 1.9229009154347133e-06, + "loss": 0.83340967, + "num_input_tokens_seen": 94701880, + "step": 4382, + "time_per_iteration": 2.7050621509552 + }, + { + "auxiliary_loss_clip": 0.01198019, + "auxiliary_loss_mlp": 0.01378855, + "balance_loss_clip": 1.0444181, + "balance_loss_mlp": 1.00004196, + "epoch": 0.5270245896711357, + "flos": 18223157646720.0, + "grad_norm": 2.3284726796529394, + "language_loss": 0.80483234, + "learning_rate": 1.922122528672327e-06, + "loss": 0.83060104, + "num_input_tokens_seen": 94720545, + "step": 4383, + "time_per_iteration": 3.6462953090667725 + }, + { + "auxiliary_loss_clip": 0.01180908, + "auxiliary_loss_mlp": 0.01022657, + "balance_loss_clip": 1.05278683, + "balance_loss_mlp": 1.01533127, + "epoch": 0.5271448325617748, + "flos": 21287558643840.0, + "grad_norm": 2.3732941302508714, + "language_loss": 0.78411484, + "learning_rate": 1.9213441537238914e-06, + "loss": 0.8061505, + "num_input_tokens_seen": 94737420, + "step": 4384, + "time_per_iteration": 2.5904481410980225 + }, + { + "auxiliary_loss_clip": 0.01129251, + "auxiliary_loss_mlp": 0.01002112, + "balance_loss_clip": 1.02784944, + "balance_loss_mlp": 1.0004667, + "epoch": 0.5272650754524139, + "flos": 65495497403520.0, + "grad_norm": 0.8408811534584526, + "language_loss": 0.5736984, + "learning_rate": 1.920565790707485e-06, + "loss": 0.59501207, + "num_input_tokens_seen": 94802810, + "step": 4385, + "time_per_iteration": 3.4679954051971436 + }, + { + "auxiliary_loss_clip": 0.01225276, + "auxiliary_loss_mlp": 0.01029038, + "balance_loss_clip": 1.04898357, + "balance_loss_mlp": 1.02012098, + "epoch": 0.527385318343053, + "flos": 19676426008320.0, + "grad_norm": 1.9782234606688347, + "language_loss": 0.65893567, + "learning_rate": 1.9197874397411853e-06, + "loss": 0.68147886, + "num_input_tokens_seen": 94819440, + "step": 4386, + "time_per_iteration": 2.8566672801971436 + }, + { + "auxiliary_loss_clip": 0.01206062, + "auxiliary_loss_mlp": 0.01028578, + "balance_loss_clip": 1.04450512, + "balance_loss_mlp": 1.02047133, + "epoch": 0.5275055612336921, + "flos": 12712018947840.0, + "grad_norm": 3.310358605440645, + "language_loss": 0.6678347, + "learning_rate": 1.919009100943067e-06, + "loss": 0.69018108, + "num_input_tokens_seen": 94835130, + "step": 4387, + "time_per_iteration": 2.6320366859436035 + }, + { + "auxiliary_loss_clip": 0.01229385, + "auxiliary_loss_mlp": 0.01027966, + "balance_loss_clip": 1.05039275, + "balance_loss_mlp": 1.01946044, + "epoch": 0.5276258041243311, + "flos": 17749029098880.0, + "grad_norm": 2.1756052974772944, + "language_loss": 0.66051036, + "learning_rate": 1.9182307744312043e-06, + "loss": 0.68308383, + "num_input_tokens_seen": 94852235, + "step": 4388, + "time_per_iteration": 2.7508299350738525 + }, + { + "auxiliary_loss_clip": 0.01204506, + "auxiliary_loss_mlp": 0.01027998, + "balance_loss_clip": 1.0509311, + "balance_loss_mlp": 1.01992178, + "epoch": 0.5277460470149702, + "flos": 22710447077760.0, + "grad_norm": 2.7917571953813667, + "language_loss": 0.7637285, + "learning_rate": 1.9174524603236676e-06, + "loss": 0.7860536, + "num_input_tokens_seen": 94871185, + "step": 4389, + "time_per_iteration": 2.70493221282959 + }, + { + "auxiliary_loss_clip": 0.01200897, + "auxiliary_loss_mlp": 0.01026326, + "balance_loss_clip": 1.05086827, + "balance_loss_mlp": 1.01769495, + "epoch": 0.5278662899056094, + "flos": 19902699734400.0, + "grad_norm": 1.7184105873710935, + "language_loss": 0.76377201, + "learning_rate": 1.916674158738527e-06, + "loss": 0.78604424, + "num_input_tokens_seen": 94890090, + "step": 4390, + "time_per_iteration": 2.686227321624756 + }, + { + "auxiliary_loss_clip": 0.01206567, + "auxiliary_loss_mlp": 0.01379252, + "balance_loss_clip": 1.05227637, + "balance_loss_mlp": 1.00012112, + "epoch": 0.5279865327962484, + "flos": 18005215875840.0, + "grad_norm": 2.215455190517794, + "language_loss": 0.60268044, + "learning_rate": 1.9158958697938506e-06, + "loss": 0.62853867, + "num_input_tokens_seen": 94908470, + "step": 4391, + "time_per_iteration": 2.7348413467407227 + }, + { + "auxiliary_loss_clip": 0.01194981, + "auxiliary_loss_mlp": 0.01029378, + "balance_loss_clip": 1.05041063, + "balance_loss_mlp": 1.02109325, + "epoch": 0.5281067756868875, + "flos": 15924443892480.0, + "grad_norm": 2.467329768799419, + "language_loss": 0.85792667, + "learning_rate": 1.9151175936077032e-06, + "loss": 0.88017023, + "num_input_tokens_seen": 94923440, + "step": 4392, + "time_per_iteration": 2.613577365875244 + }, + { + "auxiliary_loss_clip": 0.01191628, + "auxiliary_loss_mlp": 0.01029608, + "balance_loss_clip": 1.05560815, + "balance_loss_mlp": 1.02134037, + "epoch": 0.5282270185775266, + "flos": 19426488197760.0, + "grad_norm": 1.534587237742481, + "language_loss": 0.79399037, + "learning_rate": 1.9143393302981507e-06, + "loss": 0.81620276, + "num_input_tokens_seen": 94941125, + "step": 4393, + "time_per_iteration": 2.597357749938965 + }, + { + "auxiliary_loss_clip": 0.01202868, + "auxiliary_loss_mlp": 0.01028082, + "balance_loss_clip": 1.05015123, + "balance_loss_mlp": 1.02000523, + "epoch": 0.5283472614681657, + "flos": 16399613934720.0, + "grad_norm": 1.856666958945822, + "language_loss": 0.83398414, + "learning_rate": 1.913561079983252e-06, + "loss": 0.85629362, + "num_input_tokens_seen": 94959950, + "step": 4394, + "time_per_iteration": 2.6588239669799805 + }, + { + "auxiliary_loss_clip": 0.01205496, + "auxiliary_loss_mlp": 0.01028907, + "balance_loss_clip": 1.05119848, + "balance_loss_mlp": 1.02049112, + "epoch": 0.5284675043588047, + "flos": 26760524163840.0, + "grad_norm": 2.0551637310120827, + "language_loss": 0.74448186, + "learning_rate": 1.9127828427810693e-06, + "loss": 0.76682591, + "num_input_tokens_seen": 94980515, + "step": 4395, + "time_per_iteration": 2.619074583053589 + }, + { + "auxiliary_loss_clip": 0.0121835, + "auxiliary_loss_mlp": 0.01031154, + "balance_loss_clip": 1.05017352, + "balance_loss_mlp": 1.02226067, + "epoch": 0.5285877472494439, + "flos": 19899898473600.0, + "grad_norm": 2.458168490070746, + "language_loss": 0.80875599, + "learning_rate": 1.9120046188096607e-06, + "loss": 0.83125103, + "num_input_tokens_seen": 94998560, + "step": 4396, + "time_per_iteration": 2.7237279415130615 + }, + { + "auxiliary_loss_clip": 0.01198908, + "auxiliary_loss_mlp": 0.01030285, + "balance_loss_clip": 1.05102336, + "balance_loss_mlp": 1.0226078, + "epoch": 0.528707990140083, + "flos": 20011257613440.0, + "grad_norm": 1.933529691280535, + "language_loss": 0.7406311, + "learning_rate": 1.9112264081870804e-06, + "loss": 0.762923, + "num_input_tokens_seen": 95016950, + "step": 4397, + "time_per_iteration": 2.6825621128082275 + }, + { + "auxiliary_loss_clip": 0.01205963, + "auxiliary_loss_mlp": 0.01031124, + "balance_loss_clip": 1.04964554, + "balance_loss_mlp": 1.02229667, + "epoch": 0.528828233030722, + "flos": 20667956014080.0, + "grad_norm": 2.201066336803474, + "language_loss": 0.75904924, + "learning_rate": 1.9104482110313843e-06, + "loss": 0.78142011, + "num_input_tokens_seen": 95036540, + "step": 4398, + "time_per_iteration": 2.701559066772461 + }, + { + "auxiliary_loss_clip": 0.01191547, + "auxiliary_loss_mlp": 0.01036376, + "balance_loss_clip": 1.05306542, + "balance_loss_mlp": 1.02844894, + "epoch": 0.5289484759213612, + "flos": 25192448956800.0, + "grad_norm": 1.94548113808084, + "language_loss": 0.74794817, + "learning_rate": 1.909670027460623e-06, + "loss": 0.77022731, + "num_input_tokens_seen": 95053840, + "step": 4399, + "time_per_iteration": 2.690164804458618 + }, + { + "auxiliary_loss_clip": 0.01193023, + "auxiliary_loss_mlp": 0.01027215, + "balance_loss_clip": 1.05381322, + "balance_loss_mlp": 1.01908541, + "epoch": 0.5290687188120002, + "flos": 31139255715840.0, + "grad_norm": 1.921128338848143, + "language_loss": 0.71799344, + "learning_rate": 1.908891857592847e-06, + "loss": 0.74019581, + "num_input_tokens_seen": 95074910, + "step": 4400, + "time_per_iteration": 2.728548765182495 + }, + { + "auxiliary_loss_clip": 0.01202203, + "auxiliary_loss_mlp": 0.01029628, + "balance_loss_clip": 1.05035567, + "balance_loss_mlp": 1.02100372, + "epoch": 0.5291889617026393, + "flos": 20119851406080.0, + "grad_norm": 2.465616641296621, + "language_loss": 0.90511978, + "learning_rate": 1.9081137015461034e-06, + "loss": 0.92743802, + "num_input_tokens_seen": 95090985, + "step": 4401, + "time_per_iteration": 2.7096352577209473 + }, + { + "auxiliary_loss_clip": 0.0121247, + "auxiliary_loss_mlp": 0.01031748, + "balance_loss_clip": 1.05165756, + "balance_loss_mlp": 1.02351713, + "epoch": 0.5293092045932785, + "flos": 19643747610240.0, + "grad_norm": 2.0644307033647347, + "language_loss": 0.90320575, + "learning_rate": 1.9073355594384383e-06, + "loss": 0.92564785, + "num_input_tokens_seen": 95109225, + "step": 4402, + "time_per_iteration": 2.6596322059631348 + }, + { + "auxiliary_loss_clip": 0.01198496, + "auxiliary_loss_mlp": 0.01023643, + "balance_loss_clip": 1.04819238, + "balance_loss_mlp": 1.01534021, + "epoch": 0.5294294474839175, + "flos": 24317736958080.0, + "grad_norm": 1.9606370662847072, + "language_loss": 0.80492413, + "learning_rate": 1.906557431387895e-06, + "loss": 0.82714558, + "num_input_tokens_seen": 95128215, + "step": 4403, + "time_per_iteration": 2.6144745349884033 + }, + { + "auxiliary_loss_clip": 0.01209402, + "auxiliary_loss_mlp": 0.01036269, + "balance_loss_clip": 1.05465746, + "balance_loss_mlp": 1.02748322, + "epoch": 0.5295496903745566, + "flos": 18875941464960.0, + "grad_norm": 1.9875460522715003, + "language_loss": 0.78908205, + "learning_rate": 1.905779317512516e-06, + "loss": 0.8115387, + "num_input_tokens_seen": 95145760, + "step": 4404, + "time_per_iteration": 3.3761978149414062 + }, + { + "auxiliary_loss_clip": 0.01189389, + "auxiliary_loss_mlp": 0.0102227, + "balance_loss_clip": 1.05140114, + "balance_loss_mlp": 1.01471853, + "epoch": 0.5296699332651957, + "flos": 20923101296640.0, + "grad_norm": 2.167380249829395, + "language_loss": 0.80527622, + "learning_rate": 1.9050012179303385e-06, + "loss": 0.82739282, + "num_input_tokens_seen": 95164270, + "step": 4405, + "time_per_iteration": 4.645838499069214 + }, + { + "auxiliary_loss_clip": 0.01191327, + "auxiliary_loss_mlp": 0.01025318, + "balance_loss_clip": 1.04964638, + "balance_loss_mlp": 1.01696754, + "epoch": 0.5297901761558348, + "flos": 22046745525120.0, + "grad_norm": 2.391626229364063, + "language_loss": 0.69372869, + "learning_rate": 1.904223132759401e-06, + "loss": 0.71589518, + "num_input_tokens_seen": 95182870, + "step": 4406, + "time_per_iteration": 2.6265335083007812 + }, + { + "auxiliary_loss_clip": 0.01191689, + "auxiliary_loss_mlp": 0.01029245, + "balance_loss_clip": 1.05132699, + "balance_loss_mlp": 1.02175832, + "epoch": 0.5299104190464738, + "flos": 21798495653760.0, + "grad_norm": 2.8400378162548816, + "language_loss": 0.68783993, + "learning_rate": 1.9034450621177383e-06, + "loss": 0.71004927, + "num_input_tokens_seen": 95201190, + "step": 4407, + "time_per_iteration": 2.6311044692993164 + }, + { + "auxiliary_loss_clip": 0.01192318, + "auxiliary_loss_mlp": 0.01029232, + "balance_loss_clip": 1.05470216, + "balance_loss_mlp": 1.02073801, + "epoch": 0.530030661937113, + "flos": 14720790119040.0, + "grad_norm": 2.071788318620992, + "language_loss": 0.70821857, + "learning_rate": 1.9026670061233824e-06, + "loss": 0.73043406, + "num_input_tokens_seen": 95218625, + "step": 4408, + "time_per_iteration": 2.590153694152832 + }, + { + "auxiliary_loss_clip": 0.01197492, + "auxiliary_loss_mlp": 0.01034214, + "balance_loss_clip": 1.05264974, + "balance_loss_mlp": 1.02620363, + "epoch": 0.5301509048277521, + "flos": 21251504367360.0, + "grad_norm": 2.2218599541008954, + "language_loss": 0.80512369, + "learning_rate": 1.901888964894365e-06, + "loss": 0.8274408, + "num_input_tokens_seen": 95237665, + "step": 4409, + "time_per_iteration": 3.514362096786499 + }, + { + "auxiliary_loss_clip": 0.01185406, + "auxiliary_loss_mlp": 0.01026676, + "balance_loss_clip": 1.05386209, + "balance_loss_mlp": 1.01864088, + "epoch": 0.5302711477183911, + "flos": 25957058791680.0, + "grad_norm": 2.5244959596873584, + "language_loss": 0.67881179, + "learning_rate": 1.9011109385487134e-06, + "loss": 0.70093262, + "num_input_tokens_seen": 95258915, + "step": 4410, + "time_per_iteration": 2.599339485168457 + }, + { + "auxiliary_loss_clip": 0.01184992, + "auxiliary_loss_mlp": 0.01026467, + "balance_loss_clip": 1.05399919, + "balance_loss_mlp": 1.01778865, + "epoch": 0.5303913906090303, + "flos": 22273126992000.0, + "grad_norm": 2.6578613796207167, + "language_loss": 0.66535103, + "learning_rate": 1.900332927204454e-06, + "loss": 0.68746567, + "num_input_tokens_seen": 95277365, + "step": 4411, + "time_per_iteration": 2.5770716667175293 + }, + { + "auxiliary_loss_clip": 0.01205994, + "auxiliary_loss_mlp": 0.01031431, + "balance_loss_clip": 1.05025673, + "balance_loss_mlp": 1.02243984, + "epoch": 0.5305116334996693, + "flos": 24936010784640.0, + "grad_norm": 1.8752167047401473, + "language_loss": 0.76900959, + "learning_rate": 1.8995549309796097e-06, + "loss": 0.79138386, + "num_input_tokens_seen": 95296670, + "step": 4412, + "time_per_iteration": 2.648299217224121 + }, + { + "auxiliary_loss_clip": 0.01198564, + "auxiliary_loss_mlp": 0.0102805, + "balance_loss_clip": 1.05457973, + "balance_loss_mlp": 1.01988375, + "epoch": 0.5306318763903084, + "flos": 20189338266240.0, + "grad_norm": 1.6778583208150244, + "language_loss": 0.76810312, + "learning_rate": 1.8987769499922028e-06, + "loss": 0.79036927, + "num_input_tokens_seen": 95315640, + "step": 4413, + "time_per_iteration": 2.5617945194244385 + }, + { + "auxiliary_loss_clip": 0.01192087, + "auxiliary_loss_mlp": 0.01378936, + "balance_loss_clip": 1.05291748, + "balance_loss_mlp": 1.0000658, + "epoch": 0.5307521192809476, + "flos": 20266366982400.0, + "grad_norm": 2.16814166511355, + "language_loss": 0.7060855, + "learning_rate": 1.897998984360252e-06, + "loss": 0.73179579, + "num_input_tokens_seen": 95334610, + "step": 4414, + "time_per_iteration": 2.6006181240081787 + }, + { + "auxiliary_loss_clip": 0.01195834, + "auxiliary_loss_mlp": 0.01031313, + "balance_loss_clip": 1.04964149, + "balance_loss_mlp": 1.02366018, + "epoch": 0.5308723621715866, + "flos": 28844276976000.0, + "grad_norm": 1.3949343471850149, + "language_loss": 0.78439492, + "learning_rate": 1.897221034201775e-06, + "loss": 0.80666637, + "num_input_tokens_seen": 95358350, + "step": 4415, + "time_per_iteration": 2.725276231765747 + }, + { + "auxiliary_loss_clip": 0.01208049, + "auxiliary_loss_mlp": 0.01035587, + "balance_loss_clip": 1.04812932, + "balance_loss_mlp": 1.02802932, + "epoch": 0.5309926050622257, + "flos": 27457766040960.0, + "grad_norm": 1.538514435514034, + "language_loss": 0.66997182, + "learning_rate": 1.8964430996347842e-06, + "loss": 0.6924082, + "num_input_tokens_seen": 95379900, + "step": 4416, + "time_per_iteration": 2.7826080322265625 + }, + { + "auxiliary_loss_clip": 0.01198394, + "auxiliary_loss_mlp": 0.01025429, + "balance_loss_clip": 1.05044162, + "balance_loss_mlp": 1.01716781, + "epoch": 0.5311128479528648, + "flos": 20514545026560.0, + "grad_norm": 1.8847139029883664, + "language_loss": 0.82385343, + "learning_rate": 1.8956651807772931e-06, + "loss": 0.84609163, + "num_input_tokens_seen": 95397935, + "step": 4417, + "time_per_iteration": 2.695247173309326 + }, + { + "auxiliary_loss_clip": 0.01191769, + "auxiliary_loss_mlp": 0.01025506, + "balance_loss_clip": 1.05420804, + "balance_loss_mlp": 1.01839519, + "epoch": 0.5312330908435039, + "flos": 21397660807680.0, + "grad_norm": 1.5437853725375306, + "language_loss": 0.83859837, + "learning_rate": 1.8948872777473115e-06, + "loss": 0.86077112, + "num_input_tokens_seen": 95415890, + "step": 4418, + "time_per_iteration": 2.6388001441955566 + }, + { + "auxiliary_loss_clip": 0.01203209, + "auxiliary_loss_mlp": 0.01030206, + "balance_loss_clip": 1.05298924, + "balance_loss_mlp": 1.02217722, + "epoch": 0.531353333734143, + "flos": 24717350741760.0, + "grad_norm": 1.743213628981718, + "language_loss": 0.63174254, + "learning_rate": 1.8941093906628458e-06, + "loss": 0.6540767, + "num_input_tokens_seen": 95433675, + "step": 4419, + "time_per_iteration": 2.6467363834381104 + }, + { + "auxiliary_loss_clip": 0.01192334, + "auxiliary_loss_mlp": 0.01031386, + "balance_loss_clip": 1.04876542, + "balance_loss_mlp": 1.02335167, + "epoch": 0.531473576624782, + "flos": 30480689808000.0, + "grad_norm": 1.6420191238506492, + "language_loss": 0.70661873, + "learning_rate": 1.893331519641902e-06, + "loss": 0.72885597, + "num_input_tokens_seen": 95455820, + "step": 4420, + "time_per_iteration": 2.7399017810821533 + }, + { + "auxiliary_loss_clip": 0.01198611, + "auxiliary_loss_mlp": 0.01030872, + "balance_loss_clip": 1.04804754, + "balance_loss_mlp": 1.02253938, + "epoch": 0.5315938195154212, + "flos": 23002975440000.0, + "grad_norm": 2.8318713432276863, + "language_loss": 0.73825723, + "learning_rate": 1.8925536648024815e-06, + "loss": 0.76055205, + "num_input_tokens_seen": 95473240, + "step": 4421, + "time_per_iteration": 2.7173328399658203 + }, + { + "auxiliary_loss_clip": 0.01185034, + "auxiliary_loss_mlp": 0.01023659, + "balance_loss_clip": 1.05440426, + "balance_loss_mlp": 1.01532364, + "epoch": 0.5317140624060602, + "flos": 22748584343040.0, + "grad_norm": 2.758784820442333, + "language_loss": 0.75520402, + "learning_rate": 1.8917758262625849e-06, + "loss": 0.77729094, + "num_input_tokens_seen": 95493480, + "step": 4422, + "time_per_iteration": 2.614621639251709 + }, + { + "auxiliary_loss_clip": 0.01191567, + "auxiliary_loss_mlp": 0.0102417, + "balance_loss_clip": 1.05007839, + "balance_loss_mlp": 1.01667786, + "epoch": 0.5318343052966993, + "flos": 22821087945600.0, + "grad_norm": 1.8396325424561473, + "language_loss": 0.80974412, + "learning_rate": 1.8909980041402089e-06, + "loss": 0.83190155, + "num_input_tokens_seen": 95512075, + "step": 4423, + "time_per_iteration": 2.6202468872070312 + }, + { + "auxiliary_loss_clip": 0.01182328, + "auxiliary_loss_mlp": 0.01031009, + "balance_loss_clip": 1.04932308, + "balance_loss_mlp": 1.02244401, + "epoch": 0.5319545481873384, + "flos": 13626089274240.0, + "grad_norm": 2.3096780641583483, + "language_loss": 0.65670931, + "learning_rate": 1.8902201985533494e-06, + "loss": 0.67884266, + "num_input_tokens_seen": 95529340, + "step": 4424, + "time_per_iteration": 2.636601686477661 + }, + { + "auxiliary_loss_clip": 0.0120071, + "auxiliary_loss_mlp": 0.0102726, + "balance_loss_clip": 1.05242968, + "balance_loss_mlp": 1.01994681, + "epoch": 0.5320747910779775, + "flos": 22162522037760.0, + "grad_norm": 2.010391031116549, + "language_loss": 0.74983919, + "learning_rate": 1.8894424096199983e-06, + "loss": 0.77211893, + "num_input_tokens_seen": 95548545, + "step": 4425, + "time_per_iteration": 2.6015467643737793 + }, + { + "auxiliary_loss_clip": 0.01191849, + "auxiliary_loss_mlp": 0.01031564, + "balance_loss_clip": 1.05280995, + "balance_loss_mlp": 1.02325463, + "epoch": 0.5321950339686166, + "flos": 18588081870720.0, + "grad_norm": 1.836505246331873, + "language_loss": 0.85785115, + "learning_rate": 1.8886646374581463e-06, + "loss": 0.88008523, + "num_input_tokens_seen": 95567770, + "step": 4426, + "time_per_iteration": 2.6076552867889404 + }, + { + "auxiliary_loss_clip": 0.01187712, + "auxiliary_loss_mlp": 0.01030674, + "balance_loss_clip": 1.05084777, + "balance_loss_mlp": 1.0222398, + "epoch": 0.5323152768592557, + "flos": 22856818999680.0, + "grad_norm": 1.6965804300735605, + "language_loss": 0.71151119, + "learning_rate": 1.8878868821857795e-06, + "loss": 0.73369503, + "num_input_tokens_seen": 95587420, + "step": 4427, + "time_per_iteration": 2.599118232727051 + }, + { + "auxiliary_loss_clip": 0.01212247, + "auxiliary_loss_mlp": 0.0102949, + "balance_loss_clip": 1.04502249, + "balance_loss_mlp": 1.020859, + "epoch": 0.5324355197498948, + "flos": 33948690998400.0, + "grad_norm": 3.889349365429772, + "language_loss": 0.74765778, + "learning_rate": 1.8871091439208838e-06, + "loss": 0.77007514, + "num_input_tokens_seen": 95609030, + "step": 4428, + "time_per_iteration": 2.8410964012145996 + }, + { + "auxiliary_loss_clip": 0.01215206, + "auxiliary_loss_mlp": 0.01032371, + "balance_loss_clip": 1.05001009, + "balance_loss_mlp": 1.02398491, + "epoch": 0.5325557626405338, + "flos": 23256720092160.0, + "grad_norm": 2.581375526403214, + "language_loss": 0.77173924, + "learning_rate": 1.8863314227814414e-06, + "loss": 0.79421496, + "num_input_tokens_seen": 95627340, + "step": 4429, + "time_per_iteration": 2.7593085765838623 + }, + { + "auxiliary_loss_clip": 0.01198246, + "auxiliary_loss_mlp": 0.01030696, + "balance_loss_clip": 1.05420387, + "balance_loss_mlp": 1.0219996, + "epoch": 0.532676005531173, + "flos": 26718687797760.0, + "grad_norm": 2.2732966022324166, + "language_loss": 0.48810977, + "learning_rate": 1.8855537188854313e-06, + "loss": 0.51039922, + "num_input_tokens_seen": 95646315, + "step": 4430, + "time_per_iteration": 3.526488780975342 + }, + { + "auxiliary_loss_clip": 0.01191876, + "auxiliary_loss_mlp": 0.01027189, + "balance_loss_clip": 1.04901671, + "balance_loss_mlp": 1.01885653, + "epoch": 0.5327962484218121, + "flos": 17894610921600.0, + "grad_norm": 2.0418275353006967, + "language_loss": 0.78505015, + "learning_rate": 1.8847760323508315e-06, + "loss": 0.80724084, + "num_input_tokens_seen": 95665220, + "step": 4431, + "time_per_iteration": 4.3738625049591064 + }, + { + "auxiliary_loss_clip": 0.01194651, + "auxiliary_loss_mlp": 0.01026521, + "balance_loss_clip": 1.05282164, + "balance_loss_mlp": 1.01893365, + "epoch": 0.5329164913124511, + "flos": 17925385898880.0, + "grad_norm": 1.6483998419648889, + "language_loss": 0.7558828, + "learning_rate": 1.883998363295616e-06, + "loss": 0.77809453, + "num_input_tokens_seen": 95682700, + "step": 4432, + "time_per_iteration": 2.7532880306243896 + }, + { + "auxiliary_loss_clip": 0.01098324, + "auxiliary_loss_mlp": 0.01003317, + "balance_loss_clip": 1.01805902, + "balance_loss_mlp": 1.00202334, + "epoch": 0.5330367342030903, + "flos": 57254178781440.0, + "grad_norm": 0.8709625944699277, + "language_loss": 0.62634236, + "learning_rate": 1.8832207118377565e-06, + "loss": 0.64735878, + "num_input_tokens_seen": 95738070, + "step": 4433, + "time_per_iteration": 3.123332977294922 + }, + { + "auxiliary_loss_clip": 0.01181355, + "auxiliary_loss_mlp": 0.01027078, + "balance_loss_clip": 1.05286241, + "balance_loss_mlp": 1.01931119, + "epoch": 0.5331569770937293, + "flos": 17420518287360.0, + "grad_norm": 1.7615539381682381, + "language_loss": 0.69232249, + "learning_rate": 1.882443078095222e-06, + "loss": 0.71440673, + "num_input_tokens_seen": 95756950, + "step": 4434, + "time_per_iteration": 2.6179652214050293 + }, + { + "auxiliary_loss_clip": 0.01130498, + "auxiliary_loss_mlp": 0.01002905, + "balance_loss_clip": 1.01969814, + "balance_loss_mlp": 1.00170696, + "epoch": 0.5332772199843684, + "flos": 56750783627520.0, + "grad_norm": 0.8791326927062705, + "language_loss": 0.66725624, + "learning_rate": 1.8816654621859794e-06, + "loss": 0.68859029, + "num_input_tokens_seen": 95816615, + "step": 4435, + "time_per_iteration": 4.150341749191284 + }, + { + "auxiliary_loss_clip": 0.01180449, + "auxiliary_loss_mlp": 0.0102655, + "balance_loss_clip": 1.0540626, + "balance_loss_mlp": 1.01855731, + "epoch": 0.5333974628750076, + "flos": 18697753071360.0, + "grad_norm": 2.579735040852026, + "language_loss": 0.72112095, + "learning_rate": 1.8808878642279915e-06, + "loss": 0.743191, + "num_input_tokens_seen": 95832020, + "step": 4436, + "time_per_iteration": 2.7506442070007324 + }, + { + "auxiliary_loss_clip": 0.01206329, + "auxiliary_loss_mlp": 0.01039516, + "balance_loss_clip": 1.04465938, + "balance_loss_mlp": 1.03115988, + "epoch": 0.5335177057656466, + "flos": 23805507058560.0, + "grad_norm": 3.071412670265171, + "language_loss": 0.65156466, + "learning_rate": 1.8801102843392209e-06, + "loss": 0.67402309, + "num_input_tokens_seen": 95851425, + "step": 4437, + "time_per_iteration": 2.77162504196167 + }, + { + "auxiliary_loss_clip": 0.01207347, + "auxiliary_loss_mlp": 0.01027671, + "balance_loss_clip": 1.04917216, + "balance_loss_mlp": 1.01970828, + "epoch": 0.5336379486562857, + "flos": 25078683605760.0, + "grad_norm": 1.5251389157656918, + "language_loss": 0.85088491, + "learning_rate": 1.8793327226376238e-06, + "loss": 0.87323511, + "num_input_tokens_seen": 95870745, + "step": 4438, + "time_per_iteration": 2.7711679935455322 + }, + { + "auxiliary_loss_clip": 0.01205372, + "auxiliary_loss_mlp": 0.01022123, + "balance_loss_clip": 1.05017245, + "balance_loss_mlp": 1.01426148, + "epoch": 0.5337581915469248, + "flos": 21396691140480.0, + "grad_norm": 1.6895889619884128, + "language_loss": 0.80352616, + "learning_rate": 1.8785551792411569e-06, + "loss": 0.82580107, + "num_input_tokens_seen": 95889755, + "step": 4439, + "time_per_iteration": 2.6611568927764893 + }, + { + "auxiliary_loss_clip": 0.01199728, + "auxiliary_loss_mlp": 0.01027578, + "balance_loss_clip": 1.05058753, + "balance_loss_mlp": 1.01989532, + "epoch": 0.5338784344375639, + "flos": 14865905064960.0, + "grad_norm": 3.555519948934727, + "language_loss": 0.82497042, + "learning_rate": 1.8777776542677733e-06, + "loss": 0.84724349, + "num_input_tokens_seen": 95907805, + "step": 4440, + "time_per_iteration": 2.706894874572754 + }, + { + "auxiliary_loss_clip": 0.01203191, + "auxiliary_loss_mlp": 0.01029278, + "balance_loss_clip": 1.04499698, + "balance_loss_mlp": 1.02100527, + "epoch": 0.5339986773282029, + "flos": 20813501923200.0, + "grad_norm": 2.0290714147369338, + "language_loss": 0.73185694, + "learning_rate": 1.8770001478354216e-06, + "loss": 0.75418162, + "num_input_tokens_seen": 95927480, + "step": 4441, + "time_per_iteration": 2.672315835952759 + }, + { + "auxiliary_loss_clip": 0.01185212, + "auxiliary_loss_mlp": 0.01031762, + "balance_loss_clip": 1.05060613, + "balance_loss_mlp": 1.02317905, + "epoch": 0.5341189202188421, + "flos": 17969089772160.0, + "grad_norm": 2.102956161841419, + "language_loss": 0.83723629, + "learning_rate": 1.8762226600620504e-06, + "loss": 0.85940599, + "num_input_tokens_seen": 95946095, + "step": 4442, + "time_per_iteration": 2.6290087699890137 + }, + { + "auxiliary_loss_clip": 0.01204242, + "auxiliary_loss_mlp": 0.01034381, + "balance_loss_clip": 1.04907203, + "balance_loss_mlp": 1.02585161, + "epoch": 0.5342391631094812, + "flos": 11031866328960.0, + "grad_norm": 2.7467480144215477, + "language_loss": 0.58538848, + "learning_rate": 1.8754451910656031e-06, + "loss": 0.60777467, + "num_input_tokens_seen": 95959995, + "step": 4443, + "time_per_iteration": 2.7000114917755127 + }, + { + "auxiliary_loss_clip": 0.01220556, + "auxiliary_loss_mlp": 0.01029844, + "balance_loss_clip": 1.04786503, + "balance_loss_mlp": 1.02192271, + "epoch": 0.5343594060001202, + "flos": 15339135772800.0, + "grad_norm": 1.7678530658979004, + "language_loss": 0.82843798, + "learning_rate": 1.8746677409640212e-06, + "loss": 0.85094202, + "num_input_tokens_seen": 95977095, + "step": 4444, + "time_per_iteration": 2.7383556365966797 + }, + { + "auxiliary_loss_clip": 0.01195664, + "auxiliary_loss_mlp": 0.01027443, + "balance_loss_clip": 1.0535686, + "balance_loss_mlp": 1.01924706, + "epoch": 0.5344796488907594, + "flos": 26900898514560.0, + "grad_norm": 1.6698827128701361, + "language_loss": 0.84322643, + "learning_rate": 1.8738903098752432e-06, + "loss": 0.86545753, + "num_input_tokens_seen": 95996225, + "step": 4445, + "time_per_iteration": 2.7147631645202637 + }, + { + "auxiliary_loss_clip": 0.01200142, + "auxiliary_loss_mlp": 0.01028292, + "balance_loss_clip": 1.05073929, + "balance_loss_mlp": 1.02056158, + "epoch": 0.5345998917813984, + "flos": 25411216740480.0, + "grad_norm": 2.0992925317664484, + "language_loss": 0.73414743, + "learning_rate": 1.8731128979172052e-06, + "loss": 0.75643182, + "num_input_tokens_seen": 96015425, + "step": 4446, + "time_per_iteration": 2.7176051139831543 + }, + { + "auxiliary_loss_clip": 0.01198457, + "auxiliary_loss_mlp": 0.01026812, + "balance_loss_clip": 1.05074394, + "balance_loss_mlp": 1.01924849, + "epoch": 0.5347201346720375, + "flos": 32853379622400.0, + "grad_norm": 2.0047458656996406, + "language_loss": 0.67027444, + "learning_rate": 1.8723355052078394e-06, + "loss": 0.69252717, + "num_input_tokens_seen": 96035460, + "step": 4447, + "time_per_iteration": 2.785987615585327 + }, + { + "auxiliary_loss_clip": 0.01184221, + "auxiliary_loss_mlp": 0.010293, + "balance_loss_clip": 1.04807794, + "balance_loss_mlp": 1.02099133, + "epoch": 0.5348403775626767, + "flos": 17967940536960.0, + "grad_norm": 2.371813798870318, + "language_loss": 0.77363062, + "learning_rate": 1.8715581318650765e-06, + "loss": 0.79576582, + "num_input_tokens_seen": 96054515, + "step": 4448, + "time_per_iteration": 2.652480125427246 + }, + { + "auxiliary_loss_clip": 0.01219177, + "auxiliary_loss_mlp": 0.01029072, + "balance_loss_clip": 1.05179191, + "balance_loss_mlp": 1.02051306, + "epoch": 0.5349606204533157, + "flos": 17603339535360.0, + "grad_norm": 2.059933276237733, + "language_loss": 0.8131786, + "learning_rate": 1.8707807780068422e-06, + "loss": 0.83566111, + "num_input_tokens_seen": 96072330, + "step": 4449, + "time_per_iteration": 2.6686317920684814 + }, + { + "auxiliary_loss_clip": 0.01197577, + "auxiliary_loss_mlp": 0.01027429, + "balance_loss_clip": 1.0497458, + "balance_loss_mlp": 1.01963246, + "epoch": 0.5350808633439548, + "flos": 29167831710720.0, + "grad_norm": 1.9672133133147347, + "language_loss": 0.66485488, + "learning_rate": 1.8700034437510611e-06, + "loss": 0.68710494, + "num_input_tokens_seen": 96092425, + "step": 4450, + "time_per_iteration": 2.810141086578369 + }, + { + "auxiliary_loss_clip": 0.01197314, + "auxiliary_loss_mlp": 0.01030776, + "balance_loss_clip": 1.04968786, + "balance_loss_mlp": 1.02263391, + "epoch": 0.5352011062345938, + "flos": 19499997381120.0, + "grad_norm": 2.1695076262669133, + "language_loss": 0.81658459, + "learning_rate": 1.8692261292156549e-06, + "loss": 0.83886552, + "num_input_tokens_seen": 96111660, + "step": 4451, + "time_per_iteration": 2.740436315536499 + }, + { + "auxiliary_loss_clip": 0.01183086, + "auxiliary_loss_mlp": 0.01028825, + "balance_loss_clip": 1.05635655, + "balance_loss_mlp": 1.02107012, + "epoch": 0.535321349125233, + "flos": 23477642691840.0, + "grad_norm": 1.9404898776123074, + "language_loss": 0.80951118, + "learning_rate": 1.8684488345185401e-06, + "loss": 0.83163029, + "num_input_tokens_seen": 96131835, + "step": 4452, + "time_per_iteration": 2.640002965927124 + }, + { + "auxiliary_loss_clip": 0.01186735, + "auxiliary_loss_mlp": 0.0102632, + "balance_loss_clip": 1.05702734, + "balance_loss_mlp": 1.01849985, + "epoch": 0.535441592015872, + "flos": 20478059786880.0, + "grad_norm": 2.2706654571768676, + "language_loss": 0.78850114, + "learning_rate": 1.8676715597776332e-06, + "loss": 0.81063175, + "num_input_tokens_seen": 96150180, + "step": 4453, + "time_per_iteration": 2.591890573501587 + }, + { + "auxiliary_loss_clip": 0.01203474, + "auxiliary_loss_mlp": 0.01023674, + "balance_loss_clip": 1.04530871, + "balance_loss_mlp": 1.01631868, + "epoch": 0.5355618349065111, + "flos": 19573147428480.0, + "grad_norm": 1.6728126595061674, + "language_loss": 0.76205981, + "learning_rate": 1.8668943051108455e-06, + "loss": 0.78433126, + "num_input_tokens_seen": 96167485, + "step": 4454, + "time_per_iteration": 2.7209699153900146 + }, + { + "auxiliary_loss_clip": 0.01195705, + "auxiliary_loss_mlp": 0.01029981, + "balance_loss_clip": 1.04911923, + "balance_loss_mlp": 1.022071, + "epoch": 0.5356820777971503, + "flos": 24024633978240.0, + "grad_norm": 1.768044163487415, + "language_loss": 0.75797838, + "learning_rate": 1.8661170706360856e-06, + "loss": 0.78023523, + "num_input_tokens_seen": 96186650, + "step": 4455, + "time_per_iteration": 2.659365177154541 + }, + { + "auxiliary_loss_clip": 0.0118863, + "auxiliary_loss_mlp": 0.01024777, + "balance_loss_clip": 1.05312836, + "balance_loss_mlp": 1.01738644, + "epoch": 0.5358023206877893, + "flos": 20884676722560.0, + "grad_norm": 1.5859713233239008, + "language_loss": 0.81406689, + "learning_rate": 1.8653398564712594e-06, + "loss": 0.83620095, + "num_input_tokens_seen": 96205595, + "step": 4456, + "time_per_iteration": 2.634153366088867 + }, + { + "auxiliary_loss_clip": 0.01187253, + "auxiliary_loss_mlp": 0.01024562, + "balance_loss_clip": 1.05225015, + "balance_loss_mlp": 1.01700377, + "epoch": 0.5359225635784284, + "flos": 22418996123520.0, + "grad_norm": 1.5491945539831766, + "language_loss": 0.82143098, + "learning_rate": 1.8645626627342704e-06, + "loss": 0.84354913, + "num_input_tokens_seen": 96226360, + "step": 4457, + "time_per_iteration": 5.360980987548828 + }, + { + "auxiliary_loss_clip": 0.0119209, + "auxiliary_loss_mlp": 0.01025888, + "balance_loss_clip": 1.05180347, + "balance_loss_mlp": 1.01774561, + "epoch": 0.5360428064690675, + "flos": 24097784025600.0, + "grad_norm": 2.4660380089492753, + "language_loss": 0.8098411, + "learning_rate": 1.8637854895430172e-06, + "loss": 0.83202088, + "num_input_tokens_seen": 96245625, + "step": 4458, + "time_per_iteration": 2.6906330585479736 + }, + { + "auxiliary_loss_clip": 0.01195451, + "auxiliary_loss_mlp": 0.01030112, + "balance_loss_clip": 1.04636085, + "balance_loss_mlp": 1.02164197, + "epoch": 0.5361630493597066, + "flos": 21434505183360.0, + "grad_norm": 2.811228319265384, + "language_loss": 0.69365239, + "learning_rate": 1.8630083370153978e-06, + "loss": 0.71590799, + "num_input_tokens_seen": 96265265, + "step": 4459, + "time_per_iteration": 2.6899805068969727 + }, + { + "auxiliary_loss_clip": 0.01140372, + "auxiliary_loss_mlp": 0.01000934, + "balance_loss_clip": 1.01954246, + "balance_loss_mlp": 0.99972385, + "epoch": 0.5362832922503457, + "flos": 68888696520960.0, + "grad_norm": 0.7443759084540121, + "language_loss": 0.55372238, + "learning_rate": 1.8622312052693041e-06, + "loss": 0.57513547, + "num_input_tokens_seen": 96326445, + "step": 4460, + "time_per_iteration": 4.592390775680542 + }, + { + "auxiliary_loss_clip": 0.0117843, + "auxiliary_loss_mlp": 0.01026043, + "balance_loss_clip": 1.04650819, + "balance_loss_mlp": 1.01812792, + "epoch": 0.5364035351409848, + "flos": 9793702563840.0, + "grad_norm": 2.9456058423394613, + "language_loss": 0.72131902, + "learning_rate": 1.8614540944226267e-06, + "loss": 0.74336374, + "num_input_tokens_seen": 96343115, + "step": 4461, + "time_per_iteration": 3.2439382076263428 + }, + { + "auxiliary_loss_clip": 0.01195377, + "auxiliary_loss_mlp": 0.01025626, + "balance_loss_clip": 1.05145097, + "balance_loss_mlp": 1.01811039, + "epoch": 0.5365237780316239, + "flos": 23290080848640.0, + "grad_norm": 3.9719001134075866, + "language_loss": 0.68082958, + "learning_rate": 1.8606770045932537e-06, + "loss": 0.70303965, + "num_input_tokens_seen": 96362230, + "step": 4462, + "time_per_iteration": 2.6650516986846924 + }, + { + "auxiliary_loss_clip": 0.01193629, + "auxiliary_loss_mlp": 0.01029672, + "balance_loss_clip": 1.04290986, + "balance_loss_mlp": 1.02166426, + "epoch": 0.5366440209222629, + "flos": 26578133879040.0, + "grad_norm": 1.7474144715484192, + "language_loss": 0.81853247, + "learning_rate": 1.859899935899068e-06, + "loss": 0.84076548, + "num_input_tokens_seen": 96382085, + "step": 4463, + "time_per_iteration": 2.843906879425049 + }, + { + "auxiliary_loss_clip": 0.01193075, + "auxiliary_loss_mlp": 0.01029544, + "balance_loss_clip": 1.04984426, + "balance_loss_mlp": 1.02158689, + "epoch": 0.5367642638129021, + "flos": 19608052469760.0, + "grad_norm": 1.7736497145420869, + "language_loss": 0.79073244, + "learning_rate": 1.8591228884579506e-06, + "loss": 0.81295866, + "num_input_tokens_seen": 96400580, + "step": 4464, + "time_per_iteration": 2.6866774559020996 + }, + { + "auxiliary_loss_clip": 0.01212764, + "auxiliary_loss_mlp": 0.01028528, + "balance_loss_clip": 1.05000782, + "balance_loss_mlp": 1.02051067, + "epoch": 0.5368845067035412, + "flos": 23915214172800.0, + "grad_norm": 2.196774251612077, + "language_loss": 0.82265079, + "learning_rate": 1.8583458623877795e-06, + "loss": 0.84506369, + "num_input_tokens_seen": 96419680, + "step": 4465, + "time_per_iteration": 2.7171764373779297 + }, + { + "auxiliary_loss_clip": 0.01193212, + "auxiliary_loss_mlp": 0.0102486, + "balance_loss_clip": 1.05349278, + "balance_loss_mlp": 1.01696539, + "epoch": 0.5370047495941802, + "flos": 16873131951360.0, + "grad_norm": 1.6490450982306974, + "language_loss": 0.73879373, + "learning_rate": 1.8575688578064281e-06, + "loss": 0.76097447, + "num_input_tokens_seen": 96437805, + "step": 4466, + "time_per_iteration": 2.6374218463897705 + }, + { + "auxiliary_loss_clip": 0.01195399, + "auxiliary_loss_mlp": 0.01027099, + "balance_loss_clip": 1.05445492, + "balance_loss_mlp": 1.01923752, + "epoch": 0.5371249924848194, + "flos": 20740926493440.0, + "grad_norm": 1.7915464924768758, + "language_loss": 0.76767409, + "learning_rate": 1.8567918748317674e-06, + "loss": 0.78989905, + "num_input_tokens_seen": 96457155, + "step": 4467, + "time_per_iteration": 2.656249523162842 + }, + { + "auxiliary_loss_clip": 0.01205058, + "auxiliary_loss_mlp": 0.01028106, + "balance_loss_clip": 1.04580081, + "balance_loss_mlp": 1.01960635, + "epoch": 0.5372452353754584, + "flos": 17968120104960.0, + "grad_norm": 2.436920980981829, + "language_loss": 0.82683367, + "learning_rate": 1.8560149135816659e-06, + "loss": 0.84916532, + "num_input_tokens_seen": 96473990, + "step": 4468, + "time_per_iteration": 2.6827192306518555 + }, + { + "auxiliary_loss_clip": 0.01184149, + "auxiliary_loss_mlp": 0.01025118, + "balance_loss_clip": 1.04938722, + "balance_loss_mlp": 1.0177331, + "epoch": 0.5373654782660975, + "flos": 15377021642880.0, + "grad_norm": 2.267979824108285, + "language_loss": 0.84334803, + "learning_rate": 1.8552379741739873e-06, + "loss": 0.86544073, + "num_input_tokens_seen": 96491335, + "step": 4469, + "time_per_iteration": 2.639059066772461 + }, + { + "auxiliary_loss_clip": 0.01114027, + "auxiliary_loss_mlp": 0.01373851, + "balance_loss_clip": 1.0207963, + "balance_loss_mlp": 0.99975461, + "epoch": 0.5374857211567367, + "flos": 69000091574400.0, + "grad_norm": 0.8909455350130557, + "language_loss": 0.55648124, + "learning_rate": 1.8544610567265935e-06, + "loss": 0.58136004, + "num_input_tokens_seen": 96545275, + "step": 4470, + "time_per_iteration": 3.2308948040008545 + }, + { + "auxiliary_loss_clip": 0.0120105, + "auxiliary_loss_mlp": 0.01379096, + "balance_loss_clip": 1.05465031, + "balance_loss_mlp": 1.00024474, + "epoch": 0.5376059640473757, + "flos": 15085355207040.0, + "grad_norm": 2.0640735605977785, + "language_loss": 0.83476657, + "learning_rate": 1.853684161357341e-06, + "loss": 0.86056805, + "num_input_tokens_seen": 96562935, + "step": 4471, + "time_per_iteration": 2.8225481510162354 + }, + { + "auxiliary_loss_clip": 0.01187664, + "auxiliary_loss_mlp": 0.01378973, + "balance_loss_clip": 1.05234373, + "balance_loss_mlp": 1.00015581, + "epoch": 0.5377262069380148, + "flos": 19792597570560.0, + "grad_norm": 1.7036681358970853, + "language_loss": 0.76883721, + "learning_rate": 1.852907288184085e-06, + "loss": 0.79450357, + "num_input_tokens_seen": 96581820, + "step": 4472, + "time_per_iteration": 2.674110174179077 + }, + { + "auxiliary_loss_clip": 0.01216198, + "auxiliary_loss_mlp": 0.0102558, + "balance_loss_clip": 1.04552531, + "balance_loss_mlp": 1.01686037, + "epoch": 0.5378464498286539, + "flos": 30003077640960.0, + "grad_norm": 1.7982178056527742, + "language_loss": 0.70360798, + "learning_rate": 1.8521304373246762e-06, + "loss": 0.72602576, + "num_input_tokens_seen": 96602865, + "step": 4473, + "time_per_iteration": 2.7709763050079346 + }, + { + "auxiliary_loss_clip": 0.0119187, + "auxiliary_loss_mlp": 0.01028419, + "balance_loss_clip": 1.05215192, + "balance_loss_mlp": 1.01938939, + "epoch": 0.537966692719293, + "flos": 21251217058560.0, + "grad_norm": 2.8636858685301636, + "language_loss": 0.88499629, + "learning_rate": 1.8513536088969626e-06, + "loss": 0.90719926, + "num_input_tokens_seen": 96620530, + "step": 4474, + "time_per_iteration": 2.712024211883545 + }, + { + "auxiliary_loss_clip": 0.01192314, + "auxiliary_loss_mlp": 0.01026352, + "balance_loss_clip": 1.0535965, + "balance_loss_mlp": 1.01769793, + "epoch": 0.538086935609932, + "flos": 21543170803200.0, + "grad_norm": 1.7630236196773126, + "language_loss": 0.80446744, + "learning_rate": 1.8505768030187884e-06, + "loss": 0.82665408, + "num_input_tokens_seen": 96640660, + "step": 4475, + "time_per_iteration": 2.614927291870117 + }, + { + "auxiliary_loss_clip": 0.01195803, + "auxiliary_loss_mlp": 0.01028142, + "balance_loss_clip": 1.05221987, + "balance_loss_mlp": 1.02050114, + "epoch": 0.5382071785005712, + "flos": 22747219626240.0, + "grad_norm": 1.785119726017834, + "language_loss": 0.80182898, + "learning_rate": 1.849800019807995e-06, + "loss": 0.82406843, + "num_input_tokens_seen": 96661885, + "step": 4476, + "time_per_iteration": 2.674565553665161 + }, + { + "auxiliary_loss_clip": 0.01207357, + "auxiliary_loss_mlp": 0.01026247, + "balance_loss_clip": 1.04999375, + "balance_loss_mlp": 1.01827538, + "epoch": 0.5383274213912103, + "flos": 24934574240640.0, + "grad_norm": 6.879167886213409, + "language_loss": 0.70914364, + "learning_rate": 1.8490232593824186e-06, + "loss": 0.7314797, + "num_input_tokens_seen": 96678340, + "step": 4477, + "time_per_iteration": 2.6907031536102295 + }, + { + "auxiliary_loss_clip": 0.01195137, + "auxiliary_loss_mlp": 0.01027362, + "balance_loss_clip": 1.05044258, + "balance_loss_mlp": 1.01937461, + "epoch": 0.5384476642818493, + "flos": 22310186849280.0, + "grad_norm": 1.6148632559301173, + "language_loss": 0.8495757, + "learning_rate": 1.8482465218598935e-06, + "loss": 0.87180066, + "num_input_tokens_seen": 96698285, + "step": 4478, + "time_per_iteration": 2.674379587173462 + }, + { + "auxiliary_loss_clip": 0.01208442, + "auxiliary_loss_mlp": 0.01031689, + "balance_loss_clip": 1.05035615, + "balance_loss_mlp": 1.02322507, + "epoch": 0.5385679071724885, + "flos": 22711021695360.0, + "grad_norm": 1.7921494578811918, + "language_loss": 0.83084255, + "learning_rate": 1.8474698073582508e-06, + "loss": 0.85324389, + "num_input_tokens_seen": 96719655, + "step": 4479, + "time_per_iteration": 2.8307414054870605 + }, + { + "auxiliary_loss_clip": 0.01209411, + "auxiliary_loss_mlp": 0.01024358, + "balance_loss_clip": 1.04880977, + "balance_loss_mlp": 1.01562011, + "epoch": 0.5386881500631275, + "flos": 15953746412160.0, + "grad_norm": 2.3346040290855257, + "language_loss": 0.87279058, + "learning_rate": 1.8466931159953166e-06, + "loss": 0.89512825, + "num_input_tokens_seen": 96736290, + "step": 4480, + "time_per_iteration": 2.727872610092163 + }, + { + "auxiliary_loss_clip": 0.01204199, + "auxiliary_loss_mlp": 0.01027102, + "balance_loss_clip": 1.05461109, + "balance_loss_mlp": 1.01925182, + "epoch": 0.5388083929537666, + "flos": 24060041809920.0, + "grad_norm": 4.1452727623291805, + "language_loss": 0.84518826, + "learning_rate": 1.8459164478889158e-06, + "loss": 0.86750126, + "num_input_tokens_seen": 96757685, + "step": 4481, + "time_per_iteration": 2.712209939956665 + }, + { + "auxiliary_loss_clip": 0.01199525, + "auxiliary_loss_mlp": 0.01024748, + "balance_loss_clip": 1.04722309, + "balance_loss_mlp": 1.01710093, + "epoch": 0.5389286358444056, + "flos": 22236893147520.0, + "grad_norm": 1.7239666299129708, + "language_loss": 0.75983572, + "learning_rate": 1.8451398031568663e-06, + "loss": 0.7820785, + "num_input_tokens_seen": 96777310, + "step": 4482, + "time_per_iteration": 3.77093768119812 + }, + { + "auxiliary_loss_clip": 0.0120675, + "auxiliary_loss_mlp": 0.0103258, + "balance_loss_clip": 1.04901791, + "balance_loss_mlp": 1.02443767, + "epoch": 0.5390488787350448, + "flos": 24281718595200.0, + "grad_norm": 1.9290852315764617, + "language_loss": 0.74517417, + "learning_rate": 1.844363181916986e-06, + "loss": 0.76756752, + "num_input_tokens_seen": 96798035, + "step": 4483, + "time_per_iteration": 3.686713933944702 + }, + { + "auxiliary_loss_clip": 0.01186007, + "auxiliary_loss_mlp": 0.01030471, + "balance_loss_clip": 1.05084825, + "balance_loss_mlp": 1.02274013, + "epoch": 0.5391691216256839, + "flos": 16581393688320.0, + "grad_norm": 1.9162155125068199, + "language_loss": 0.83108836, + "learning_rate": 1.8435865842870868e-06, + "loss": 0.85325313, + "num_input_tokens_seen": 96815975, + "step": 4484, + "time_per_iteration": 2.6351144313812256 + }, + { + "auxiliary_loss_clip": 0.01185472, + "auxiliary_loss_mlp": 0.01379184, + "balance_loss_clip": 1.04514599, + "balance_loss_mlp": 1.00027442, + "epoch": 0.5392893645163229, + "flos": 23330049707520.0, + "grad_norm": 1.8040404750123884, + "language_loss": 0.71998572, + "learning_rate": 1.8428100103849787e-06, + "loss": 0.74563235, + "num_input_tokens_seen": 96835770, + "step": 4485, + "time_per_iteration": 2.6418049335479736 + }, + { + "auxiliary_loss_clip": 0.01197249, + "auxiliary_loss_mlp": 0.0102248, + "balance_loss_clip": 1.05270207, + "balance_loss_mlp": 1.01487494, + "epoch": 0.5394096074069621, + "flos": 15669801400320.0, + "grad_norm": 3.3925827094444205, + "language_loss": 0.73085463, + "learning_rate": 1.842033460328467e-06, + "loss": 0.75305194, + "num_input_tokens_seen": 96854490, + "step": 4486, + "time_per_iteration": 3.534125328063965 + }, + { + "auxiliary_loss_clip": 0.01201047, + "auxiliary_loss_mlp": 0.01379144, + "balance_loss_clip": 1.04883432, + "balance_loss_mlp": 1.0002321, + "epoch": 0.5395298502976011, + "flos": 22893447893760.0, + "grad_norm": 1.6088286467404185, + "language_loss": 0.74459767, + "learning_rate": 1.8412569342353541e-06, + "loss": 0.77039957, + "num_input_tokens_seen": 96874645, + "step": 4487, + "time_per_iteration": 2.687612295150757 + }, + { + "auxiliary_loss_clip": 0.01208408, + "auxiliary_loss_mlp": 0.01030518, + "balance_loss_clip": 1.0554992, + "balance_loss_mlp": 1.0222683, + "epoch": 0.5396500931882402, + "flos": 23842135952640.0, + "grad_norm": 1.8031787482661448, + "language_loss": 0.8479532, + "learning_rate": 1.840480432223438e-06, + "loss": 0.87034243, + "num_input_tokens_seen": 96893650, + "step": 4488, + "time_per_iteration": 2.6796159744262695 + }, + { + "auxiliary_loss_clip": 0.01199135, + "auxiliary_loss_mlp": 0.0103066, + "balance_loss_clip": 1.04877281, + "balance_loss_mlp": 1.02293253, + "epoch": 0.5397703360788794, + "flos": 26322988596480.0, + "grad_norm": 1.8920549719237147, + "language_loss": 0.78034568, + "learning_rate": 1.8397039544105131e-06, + "loss": 0.80264366, + "num_input_tokens_seen": 96912735, + "step": 4489, + "time_per_iteration": 2.722752571105957 + }, + { + "auxiliary_loss_clip": 0.01187794, + "auxiliary_loss_mlp": 0.01032548, + "balance_loss_clip": 1.04645848, + "balance_loss_mlp": 1.02423882, + "epoch": 0.5398905789695184, + "flos": 21214588164480.0, + "grad_norm": 1.7929639182017771, + "language_loss": 0.7007401, + "learning_rate": 1.8389275009143711e-06, + "loss": 0.72294354, + "num_input_tokens_seen": 96932475, + "step": 4490, + "time_per_iteration": 2.6507949829101562 + }, + { + "auxiliary_loss_clip": 0.01177872, + "auxiliary_loss_mlp": 0.01027411, + "balance_loss_clip": 1.051319, + "balance_loss_mlp": 1.02025282, + "epoch": 0.5400108218601575, + "flos": 25080335631360.0, + "grad_norm": 2.1728741872812076, + "language_loss": 0.73159224, + "learning_rate": 1.8381510718527988e-06, + "loss": 0.75364506, + "num_input_tokens_seen": 96952085, + "step": 4491, + "time_per_iteration": 2.6624999046325684 + }, + { + "auxiliary_loss_clip": 0.01199127, + "auxiliary_loss_mlp": 0.01030854, + "balance_loss_clip": 1.04727459, + "balance_loss_mlp": 1.02270007, + "epoch": 0.5401310647507966, + "flos": 26357498588160.0, + "grad_norm": 2.085410617181964, + "language_loss": 0.63337183, + "learning_rate": 1.8373746673435812e-06, + "loss": 0.65567172, + "num_input_tokens_seen": 96973110, + "step": 4492, + "time_per_iteration": 2.7649011611938477 + }, + { + "auxiliary_loss_clip": 0.0118345, + "auxiliary_loss_mlp": 0.01025196, + "balance_loss_clip": 1.05457497, + "balance_loss_mlp": 1.01719141, + "epoch": 0.5402513076414357, + "flos": 27855332749440.0, + "grad_norm": 1.6887304760861583, + "language_loss": 0.79012918, + "learning_rate": 1.8365982875044964e-06, + "loss": 0.81221563, + "num_input_tokens_seen": 96993420, + "step": 4493, + "time_per_iteration": 2.7467873096466064 + }, + { + "auxiliary_loss_clip": 0.01196895, + "auxiliary_loss_mlp": 0.01379474, + "balance_loss_clip": 1.05384922, + "balance_loss_mlp": 1.00022793, + "epoch": 0.5403715505320748, + "flos": 22893771116160.0, + "grad_norm": 1.9779233703594734, + "language_loss": 0.75995588, + "learning_rate": 1.8358219324533217e-06, + "loss": 0.78571963, + "num_input_tokens_seen": 97013685, + "step": 4494, + "time_per_iteration": 2.6670522689819336 + }, + { + "auxiliary_loss_clip": 0.01192586, + "auxiliary_loss_mlp": 0.01025329, + "balance_loss_clip": 1.04822636, + "balance_loss_mlp": 1.01843846, + "epoch": 0.5404917934227139, + "flos": 30224143895040.0, + "grad_norm": 1.9599637154280873, + "language_loss": 0.70596778, + "learning_rate": 1.8350456023078292e-06, + "loss": 0.72814691, + "num_input_tokens_seen": 97036060, + "step": 4495, + "time_per_iteration": 2.7409331798553467 + }, + { + "auxiliary_loss_clip": 0.01186659, + "auxiliary_loss_mlp": 0.01027705, + "balance_loss_clip": 1.05517483, + "balance_loss_mlp": 1.01915729, + "epoch": 0.540612036313353, + "flos": 19938502615680.0, + "grad_norm": 2.133972993650113, + "language_loss": 0.78138649, + "learning_rate": 1.8342692971857874e-06, + "loss": 0.80353016, + "num_input_tokens_seen": 97055260, + "step": 4496, + "time_per_iteration": 2.564743757247925 + }, + { + "auxiliary_loss_clip": 0.01196195, + "auxiliary_loss_mlp": 0.01027238, + "balance_loss_clip": 1.05066097, + "balance_loss_mlp": 1.019328, + "epoch": 0.540732279203992, + "flos": 24279599692800.0, + "grad_norm": 8.046086626089581, + "language_loss": 0.7168498, + "learning_rate": 1.833493017204962e-06, + "loss": 0.73908412, + "num_input_tokens_seen": 97075365, + "step": 4497, + "time_per_iteration": 2.7423927783966064 + }, + { + "auxiliary_loss_clip": 0.0118, + "auxiliary_loss_mlp": 0.01030159, + "balance_loss_clip": 1.05193853, + "balance_loss_mlp": 1.02242804, + "epoch": 0.5408525220946312, + "flos": 20193216935040.0, + "grad_norm": 2.1259461669481925, + "language_loss": 0.78016514, + "learning_rate": 1.8327167624831134e-06, + "loss": 0.80226678, + "num_input_tokens_seen": 97093095, + "step": 4498, + "time_per_iteration": 2.569519281387329 + }, + { + "auxiliary_loss_clip": 0.01180078, + "auxiliary_loss_mlp": 0.01029379, + "balance_loss_clip": 1.05314112, + "balance_loss_mlp": 1.02163696, + "epoch": 0.5409727649852702, + "flos": 24134448833280.0, + "grad_norm": 1.6041076342274732, + "language_loss": 0.70853388, + "learning_rate": 1.831940533137999e-06, + "loss": 0.73062849, + "num_input_tokens_seen": 97112000, + "step": 4499, + "time_per_iteration": 2.646745204925537 + }, + { + "auxiliary_loss_clip": 0.01187529, + "auxiliary_loss_mlp": 0.01029719, + "balance_loss_clip": 1.05308616, + "balance_loss_mlp": 1.0219171, + "epoch": 0.5410930078759093, + "flos": 23912700220800.0, + "grad_norm": 1.7810804136259233, + "language_loss": 0.72129554, + "learning_rate": 1.8311643292873718e-06, + "loss": 0.74346805, + "num_input_tokens_seen": 97130820, + "step": 4500, + "time_per_iteration": 2.629751443862915 + }, + { + "auxiliary_loss_clip": 0.0118599, + "auxiliary_loss_mlp": 0.01033684, + "balance_loss_clip": 1.05255103, + "balance_loss_mlp": 1.0257926, + "epoch": 0.5412132507665485, + "flos": 21105132445440.0, + "grad_norm": 1.8308242939181418, + "language_loss": 0.87871444, + "learning_rate": 1.8303881510489818e-06, + "loss": 0.90091121, + "num_input_tokens_seen": 97149210, + "step": 4501, + "time_per_iteration": 2.608839750289917 + }, + { + "auxiliary_loss_clip": 0.01201849, + "auxiliary_loss_mlp": 0.01034095, + "balance_loss_clip": 1.05222344, + "balance_loss_mlp": 1.02586973, + "epoch": 0.5413334936571875, + "flos": 30227340205440.0, + "grad_norm": 2.1772779381514153, + "language_loss": 0.69519985, + "learning_rate": 1.829611998540574e-06, + "loss": 0.71755928, + "num_input_tokens_seen": 97170415, + "step": 4502, + "time_per_iteration": 2.6900551319122314 + }, + { + "auxiliary_loss_clip": 0.01190301, + "auxiliary_loss_mlp": 0.01379156, + "balance_loss_clip": 1.05120504, + "balance_loss_mlp": 1.00024343, + "epoch": 0.5414537365478266, + "flos": 24279635606400.0, + "grad_norm": 1.7104264907943414, + "language_loss": 0.80003798, + "learning_rate": 1.8288358718798914e-06, + "loss": 0.82573247, + "num_input_tokens_seen": 97189605, + "step": 4503, + "time_per_iteration": 2.667942762374878 + }, + { + "auxiliary_loss_clip": 0.01185066, + "auxiliary_loss_mlp": 0.01378757, + "balance_loss_clip": 1.05242991, + "balance_loss_mlp": 1.00028658, + "epoch": 0.5415739794384657, + "flos": 16654543735680.0, + "grad_norm": 1.7213583586723564, + "language_loss": 0.72298396, + "learning_rate": 1.8280597711846703e-06, + "loss": 0.74862218, + "num_input_tokens_seen": 97207845, + "step": 4504, + "time_per_iteration": 2.5810675621032715 + }, + { + "auxiliary_loss_clip": 0.01185869, + "auxiliary_loss_mlp": 0.01031479, + "balance_loss_clip": 1.05243731, + "balance_loss_mlp": 1.02407587, + "epoch": 0.5416942223291048, + "flos": 23185724860800.0, + "grad_norm": 2.142426997803452, + "language_loss": 0.82998937, + "learning_rate": 1.8272836965726455e-06, + "loss": 0.85216284, + "num_input_tokens_seen": 97226780, + "step": 4505, + "time_per_iteration": 2.6795215606689453 + }, + { + "auxiliary_loss_clip": 0.01220407, + "auxiliary_loss_mlp": 0.01031677, + "balance_loss_clip": 1.04321718, + "balance_loss_mlp": 1.02324867, + "epoch": 0.5418144652197439, + "flos": 20303247271680.0, + "grad_norm": 1.699148890414731, + "language_loss": 0.77901483, + "learning_rate": 1.8265076481615461e-06, + "loss": 0.80153567, + "num_input_tokens_seen": 97246695, + "step": 4506, + "time_per_iteration": 2.930312156677246 + }, + { + "auxiliary_loss_clip": 0.01198819, + "auxiliary_loss_mlp": 0.01027222, + "balance_loss_clip": 1.05408955, + "balance_loss_mlp": 1.01919365, + "epoch": 0.541934708110383, + "flos": 12458633431680.0, + "grad_norm": 2.0761198289015486, + "language_loss": 0.86933541, + "learning_rate": 1.8257316260690987e-06, + "loss": 0.89159596, + "num_input_tokens_seen": 97264480, + "step": 4507, + "time_per_iteration": 2.8250668048858643 + }, + { + "auxiliary_loss_clip": 0.01188775, + "auxiliary_loss_mlp": 0.01025343, + "balance_loss_clip": 1.0512383, + "balance_loss_mlp": 1.01768684, + "epoch": 0.5420549510010221, + "flos": 21253802837760.0, + "grad_norm": 1.58345711971062, + "language_loss": 0.76097012, + "learning_rate": 1.8249556304130254e-06, + "loss": 0.78311133, + "num_input_tokens_seen": 97285760, + "step": 4508, + "time_per_iteration": 3.5370564460754395 + }, + { + "auxiliary_loss_clip": 0.01182811, + "auxiliary_loss_mlp": 0.01028642, + "balance_loss_clip": 1.04610372, + "balance_loss_mlp": 1.02086377, + "epoch": 0.5421751938916611, + "flos": 29490524519040.0, + "grad_norm": 2.2568068744631176, + "language_loss": 0.68818432, + "learning_rate": 1.824179661311044e-06, + "loss": 0.7102989, + "num_input_tokens_seen": 97304510, + "step": 4509, + "time_per_iteration": 3.8200478553771973 + }, + { + "auxiliary_loss_clip": 0.01212862, + "auxiliary_loss_mlp": 0.01024278, + "balance_loss_clip": 1.04295516, + "balance_loss_mlp": 1.01677346, + "epoch": 0.5422954367823003, + "flos": 18734238311040.0, + "grad_norm": 2.031278617464603, + "language_loss": 0.7990253, + "learning_rate": 1.823403718880868e-06, + "loss": 0.82139671, + "num_input_tokens_seen": 97323270, + "step": 4510, + "time_per_iteration": 3.61002254486084 + }, + { + "auxiliary_loss_clip": 0.01194225, + "auxiliary_loss_mlp": 0.01029629, + "balance_loss_clip": 1.04717731, + "balance_loss_mlp": 1.0221014, + "epoch": 0.5424156796729394, + "flos": 39969006940800.0, + "grad_norm": 1.7315666326230814, + "language_loss": 0.66478229, + "learning_rate": 1.822627803240207e-06, + "loss": 0.68702078, + "num_input_tokens_seen": 97345600, + "step": 4511, + "time_per_iteration": 2.7774767875671387 + }, + { + "auxiliary_loss_clip": 0.01208019, + "auxiliary_loss_mlp": 0.0102861, + "balance_loss_clip": 1.04746735, + "balance_loss_mlp": 1.02073002, + "epoch": 0.5425359225635784, + "flos": 11546538353280.0, + "grad_norm": 2.9420543587678, + "language_loss": 0.85166681, + "learning_rate": 1.8218519145067675e-06, + "loss": 0.87403309, + "num_input_tokens_seen": 97361220, + "step": 4512, + "time_per_iteration": 3.578504800796509 + }, + { + "auxiliary_loss_clip": 0.01198018, + "auxiliary_loss_mlp": 0.01026475, + "balance_loss_clip": 1.04576409, + "balance_loss_mlp": 1.01848221, + "epoch": 0.5426561654542175, + "flos": 20229702174720.0, + "grad_norm": 1.9995770542570077, + "language_loss": 0.89453518, + "learning_rate": 1.8210760527982508e-06, + "loss": 0.91678011, + "num_input_tokens_seen": 97381505, + "step": 4513, + "time_per_iteration": 2.723421573638916 + }, + { + "auxiliary_loss_clip": 0.0119919, + "auxiliary_loss_mlp": 0.01378674, + "balance_loss_clip": 1.05203652, + "balance_loss_mlp": 1.00028801, + "epoch": 0.5427764083448566, + "flos": 21871681614720.0, + "grad_norm": 2.315246754624182, + "language_loss": 0.75030792, + "learning_rate": 1.8203002182323552e-06, + "loss": 0.77608651, + "num_input_tokens_seen": 97399060, + "step": 4514, + "time_per_iteration": 2.746253252029419 + }, + { + "auxiliary_loss_clip": 0.01195656, + "auxiliary_loss_mlp": 0.01030178, + "balance_loss_clip": 1.0483129, + "balance_loss_mlp": 1.02223873, + "epoch": 0.5428966512354957, + "flos": 19640946349440.0, + "grad_norm": 2.1299234457695952, + "language_loss": 0.75606805, + "learning_rate": 1.819524410926773e-06, + "loss": 0.77832639, + "num_input_tokens_seen": 97416740, + "step": 4515, + "time_per_iteration": 2.6255905628204346 + }, + { + "auxiliary_loss_clip": 0.01218645, + "auxiliary_loss_mlp": 0.01029346, + "balance_loss_clip": 1.04538941, + "balance_loss_mlp": 1.0220449, + "epoch": 0.5430168941261347, + "flos": 22382187661440.0, + "grad_norm": 1.4554793660206404, + "language_loss": 0.76802188, + "learning_rate": 1.8187486309991944e-06, + "loss": 0.79050183, + "num_input_tokens_seen": 97437620, + "step": 4516, + "time_per_iteration": 2.844266176223755 + }, + { + "auxiliary_loss_clip": 0.0119315, + "auxiliary_loss_mlp": 0.01025413, + "balance_loss_clip": 1.05148125, + "balance_loss_mlp": 1.01811409, + "epoch": 0.5431371370167739, + "flos": 18764187275520.0, + "grad_norm": 2.0371392189090027, + "language_loss": 0.77870381, + "learning_rate": 1.817972878567304e-06, + "loss": 0.80088943, + "num_input_tokens_seen": 97456275, + "step": 4517, + "time_per_iteration": 2.6064960956573486 + }, + { + "auxiliary_loss_clip": 0.01201395, + "auxiliary_loss_mlp": 0.01029659, + "balance_loss_clip": 1.04820991, + "balance_loss_mlp": 1.02223253, + "epoch": 0.543257379907413, + "flos": 18806023641600.0, + "grad_norm": 1.7858948068906595, + "language_loss": 0.76525772, + "learning_rate": 1.8171971537487834e-06, + "loss": 0.78756821, + "num_input_tokens_seen": 97474925, + "step": 4518, + "time_per_iteration": 2.662393569946289 + }, + { + "auxiliary_loss_clip": 0.01180355, + "auxiliary_loss_mlp": 0.01022752, + "balance_loss_clip": 1.0513519, + "balance_loss_mlp": 1.01499772, + "epoch": 0.543377622798052, + "flos": 17493381025920.0, + "grad_norm": 1.8722663769261316, + "language_loss": 0.80710363, + "learning_rate": 1.8164214566613093e-06, + "loss": 0.82913464, + "num_input_tokens_seen": 97493550, + "step": 4519, + "time_per_iteration": 2.5291481018066406 + }, + { + "auxiliary_loss_clip": 0.01179243, + "auxiliary_loss_mlp": 0.0102866, + "balance_loss_clip": 1.05382681, + "balance_loss_mlp": 1.02101278, + "epoch": 0.5434978656886912, + "flos": 18989311766400.0, + "grad_norm": 3.232436653647562, + "language_loss": 0.65222585, + "learning_rate": 1.8156457874225547e-06, + "loss": 0.6743049, + "num_input_tokens_seen": 97512010, + "step": 4520, + "time_per_iteration": 2.7477147579193115 + }, + { + "auxiliary_loss_clip": 0.01195555, + "auxiliary_loss_mlp": 0.01031452, + "balance_loss_clip": 1.05500507, + "balance_loss_mlp": 1.02378654, + "epoch": 0.5436181085793302, + "flos": 17274936464640.0, + "grad_norm": 1.7869823515911063, + "language_loss": 0.803792, + "learning_rate": 1.814870146150187e-06, + "loss": 0.82606202, + "num_input_tokens_seen": 97530120, + "step": 4521, + "time_per_iteration": 2.6302473545074463 + }, + { + "auxiliary_loss_clip": 0.01201819, + "auxiliary_loss_mlp": 0.0102739, + "balance_loss_clip": 1.04797983, + "balance_loss_mlp": 1.01888442, + "epoch": 0.5437383514699693, + "flos": 19098587917440.0, + "grad_norm": 2.044911244429885, + "language_loss": 0.7874254, + "learning_rate": 1.814094532961871e-06, + "loss": 0.80971742, + "num_input_tokens_seen": 97548695, + "step": 4522, + "time_per_iteration": 2.669416666030884 + }, + { + "auxiliary_loss_clip": 0.01214765, + "auxiliary_loss_mlp": 0.01029762, + "balance_loss_clip": 1.04502487, + "balance_loss_mlp": 1.02202499, + "epoch": 0.5438585943606085, + "flos": 22602715211520.0, + "grad_norm": 1.88363292466038, + "language_loss": 0.83585113, + "learning_rate": 1.8133189479752666e-06, + "loss": 0.85829639, + "num_input_tokens_seen": 97567625, + "step": 4523, + "time_per_iteration": 2.8080520629882812 + }, + { + "auxiliary_loss_clip": 0.01180833, + "auxiliary_loss_mlp": 0.01022056, + "balance_loss_clip": 1.05373824, + "balance_loss_mlp": 1.01464415, + "epoch": 0.5439788372512475, + "flos": 21798495653760.0, + "grad_norm": 2.019775031965942, + "language_loss": 0.81926161, + "learning_rate": 1.8125433913080292e-06, + "loss": 0.84129047, + "num_input_tokens_seen": 97585325, + "step": 4524, + "time_per_iteration": 2.6090519428253174 + }, + { + "auxiliary_loss_clip": 0.01234277, + "auxiliary_loss_mlp": 0.01026435, + "balance_loss_clip": 1.0359081, + "balance_loss_mlp": 1.01930976, + "epoch": 0.5440990801418866, + "flos": 16399362539520.0, + "grad_norm": 2.0619197972584775, + "language_loss": 0.82649833, + "learning_rate": 1.811767863077811e-06, + "loss": 0.84910542, + "num_input_tokens_seen": 97604275, + "step": 4525, + "time_per_iteration": 3.1570982933044434 + }, + { + "auxiliary_loss_clip": 0.01217928, + "auxiliary_loss_mlp": 0.01025041, + "balance_loss_clip": 1.04904389, + "balance_loss_mlp": 1.01806092, + "epoch": 0.5442193230325257, + "flos": 21615638492160.0, + "grad_norm": 1.624420742339382, + "language_loss": 0.78131032, + "learning_rate": 1.8109923634022577e-06, + "loss": 0.80374002, + "num_input_tokens_seen": 97624300, + "step": 4526, + "time_per_iteration": 2.987658739089966 + }, + { + "auxiliary_loss_clip": 0.01180751, + "auxiliary_loss_mlp": 0.01026968, + "balance_loss_clip": 1.05189943, + "balance_loss_mlp": 1.01911259, + "epoch": 0.5443395659231648, + "flos": 15481198062720.0, + "grad_norm": 3.6393802699402653, + "language_loss": 0.86333418, + "learning_rate": 1.8102168923990128e-06, + "loss": 0.88541138, + "num_input_tokens_seen": 97637845, + "step": 4527, + "time_per_iteration": 2.6134891510009766 + }, + { + "auxiliary_loss_clip": 0.01193532, + "auxiliary_loss_mlp": 0.01378782, + "balance_loss_clip": 1.05389524, + "balance_loss_mlp": 1.00026596, + "epoch": 0.5444598088138038, + "flos": 18770436241920.0, + "grad_norm": 1.8550138535475795, + "language_loss": 0.79836476, + "learning_rate": 1.809441450185714e-06, + "loss": 0.82408786, + "num_input_tokens_seen": 97656330, + "step": 4528, + "time_per_iteration": 2.6673035621643066 + }, + { + "auxiliary_loss_clip": 0.0120094, + "auxiliary_loss_mlp": 0.01025568, + "balance_loss_clip": 1.04758716, + "balance_loss_mlp": 1.01821899, + "epoch": 0.544580051704443, + "flos": 21142335957120.0, + "grad_norm": 2.141545628462025, + "language_loss": 0.73298228, + "learning_rate": 1.8086660368799958e-06, + "loss": 0.75524735, + "num_input_tokens_seen": 97674380, + "step": 4529, + "time_per_iteration": 2.7002289295196533 + }, + { + "auxiliary_loss_clip": 0.01200983, + "auxiliary_loss_mlp": 0.01030346, + "balance_loss_clip": 1.05057621, + "balance_loss_mlp": 1.02189958, + "epoch": 0.5447002945950821, + "flos": 32491508054400.0, + "grad_norm": 1.6158988720930247, + "language_loss": 0.7753948, + "learning_rate": 1.807890652599488e-06, + "loss": 0.79770815, + "num_input_tokens_seen": 97698765, + "step": 4530, + "time_per_iteration": 2.741255044937134 + }, + { + "auxiliary_loss_clip": 0.01178602, + "auxiliary_loss_mlp": 0.0102373, + "balance_loss_clip": 1.05353665, + "balance_loss_mlp": 1.01664305, + "epoch": 0.5448205374857211, + "flos": 11798307757440.0, + "grad_norm": 2.3880064328282478, + "language_loss": 0.82816947, + "learning_rate": 1.8071152974618156e-06, + "loss": 0.85019279, + "num_input_tokens_seen": 97716565, + "step": 4531, + "time_per_iteration": 2.6076862812042236 + }, + { + "auxiliary_loss_clip": 0.01205018, + "auxiliary_loss_mlp": 0.01378679, + "balance_loss_clip": 1.04545975, + "balance_loss_mlp": 1.00022936, + "epoch": 0.5449407803763603, + "flos": 24133766474880.0, + "grad_norm": 2.078697863248137, + "language_loss": 0.78350812, + "learning_rate": 1.806339971584599e-06, + "loss": 0.80934513, + "num_input_tokens_seen": 97733225, + "step": 4532, + "time_per_iteration": 2.733767509460449 + }, + { + "auxiliary_loss_clip": 0.01180107, + "auxiliary_loss_mlp": 0.01027613, + "balance_loss_clip": 1.05298913, + "balance_loss_mlp": 1.01973367, + "epoch": 0.5450610232669993, + "flos": 23258551685760.0, + "grad_norm": 1.8486510107104843, + "language_loss": 0.85081249, + "learning_rate": 1.8055646750854546e-06, + "loss": 0.87288976, + "num_input_tokens_seen": 97752735, + "step": 4533, + "time_per_iteration": 2.63002610206604 + }, + { + "auxiliary_loss_clip": 0.01202456, + "auxiliary_loss_mlp": 0.01026468, + "balance_loss_clip": 1.05048156, + "balance_loss_mlp": 1.01860583, + "epoch": 0.5451812661576384, + "flos": 17785083375360.0, + "grad_norm": 2.449874366077887, + "language_loss": 0.81613624, + "learning_rate": 1.8047894080819945e-06, + "loss": 0.83842546, + "num_input_tokens_seen": 97769985, + "step": 4534, + "time_per_iteration": 3.923198938369751 + }, + { + "auxiliary_loss_clip": 0.01082952, + "auxiliary_loss_mlp": 0.01004415, + "balance_loss_clip": 1.02444148, + "balance_loss_mlp": 1.00325859, + "epoch": 0.5453015090482776, + "flos": 71062586513280.0, + "grad_norm": 0.7199414962900558, + "language_loss": 0.63167894, + "learning_rate": 1.8040141706918258e-06, + "loss": 0.6525526, + "num_input_tokens_seen": 97831225, + "step": 4535, + "time_per_iteration": 4.740273952484131 + }, + { + "auxiliary_loss_clip": 0.01202342, + "auxiliary_loss_mlp": 0.01024103, + "balance_loss_clip": 1.05404735, + "balance_loss_mlp": 1.01644409, + "epoch": 0.5454217519389166, + "flos": 25552201622400.0, + "grad_norm": 1.789998692068052, + "language_loss": 0.77099824, + "learning_rate": 1.8032389630325525e-06, + "loss": 0.79326272, + "num_input_tokens_seen": 97849975, + "step": 4536, + "time_per_iteration": 3.649118185043335 + }, + { + "auxiliary_loss_clip": 0.01193704, + "auxiliary_loss_mlp": 0.01034575, + "balance_loss_clip": 1.04663539, + "balance_loss_mlp": 1.02648675, + "epoch": 0.5455419948295557, + "flos": 23658345037440.0, + "grad_norm": 1.659110554025987, + "language_loss": 0.75601065, + "learning_rate": 1.8024637852217707e-06, + "loss": 0.77829349, + "num_input_tokens_seen": 97869700, + "step": 4537, + "time_per_iteration": 3.678656578063965 + }, + { + "auxiliary_loss_clip": 0.0119834, + "auxiliary_loss_mlp": 0.01033684, + "balance_loss_clip": 1.05145371, + "balance_loss_mlp": 1.02696359, + "epoch": 0.5456622377201948, + "flos": 23403989854080.0, + "grad_norm": 1.7339379674865059, + "language_loss": 0.8455708, + "learning_rate": 1.8016886373770766e-06, + "loss": 0.86789107, + "num_input_tokens_seen": 97888215, + "step": 4538, + "time_per_iteration": 2.7206292152404785 + }, + { + "auxiliary_loss_clip": 0.01194613, + "auxiliary_loss_mlp": 0.01026573, + "balance_loss_clip": 1.04845107, + "balance_loss_mlp": 1.01900923, + "epoch": 0.5457824806108339, + "flos": 23988040997760.0, + "grad_norm": 1.5246975725877916, + "language_loss": 0.7908591, + "learning_rate": 1.8009135196160579e-06, + "loss": 0.81307089, + "num_input_tokens_seen": 97907090, + "step": 4539, + "time_per_iteration": 2.643852710723877 + }, + { + "auxiliary_loss_clip": 0.0120269, + "auxiliary_loss_mlp": 0.01026524, + "balance_loss_clip": 1.04736423, + "balance_loss_mlp": 1.0189898, + "epoch": 0.545902723501473, + "flos": 22565870835840.0, + "grad_norm": 1.7043525362954905, + "language_loss": 0.84253871, + "learning_rate": 1.8001384320563e-06, + "loss": 0.86483085, + "num_input_tokens_seen": 97927345, + "step": 4540, + "time_per_iteration": 2.7179720401763916 + }, + { + "auxiliary_loss_clip": 0.01083393, + "auxiliary_loss_mlp": 0.01001045, + "balance_loss_clip": 1.02458286, + "balance_loss_mlp": 0.99988884, + "epoch": 0.5460229663921121, + "flos": 55198399685760.0, + "grad_norm": 0.7672403817521066, + "language_loss": 0.57725966, + "learning_rate": 1.7993633748153833e-06, + "loss": 0.598104, + "num_input_tokens_seen": 97981950, + "step": 4541, + "time_per_iteration": 3.0458149909973145 + }, + { + "auxiliary_loss_clip": 0.01190207, + "auxiliary_loss_mlp": 0.0102963, + "balance_loss_clip": 1.05034196, + "balance_loss_mlp": 1.02186334, + "epoch": 0.5461432092827512, + "flos": 15413866018560.0, + "grad_norm": 1.79993609330031, + "language_loss": 0.72994292, + "learning_rate": 1.7985883480108834e-06, + "loss": 0.7521413, + "num_input_tokens_seen": 97999585, + "step": 4542, + "time_per_iteration": 2.578174114227295 + }, + { + "auxiliary_loss_clip": 0.01181935, + "auxiliary_loss_mlp": 0.01029276, + "balance_loss_clip": 1.04921007, + "balance_loss_mlp": 1.02183771, + "epoch": 0.5462634521733902, + "flos": 24024921287040.0, + "grad_norm": 1.7775861017628432, + "language_loss": 0.72166294, + "learning_rate": 1.797813351760371e-06, + "loss": 0.74377501, + "num_input_tokens_seen": 98021290, + "step": 4543, + "time_per_iteration": 2.6522676944732666 + }, + { + "auxiliary_loss_clip": 0.0118017, + "auxiliary_loss_mlp": 0.01027474, + "balance_loss_clip": 1.05245376, + "balance_loss_mlp": 1.01989794, + "epoch": 0.5463836950640293, + "flos": 22820944291200.0, + "grad_norm": 2.8616457863013713, + "language_loss": 0.78027868, + "learning_rate": 1.7970383861814116e-06, + "loss": 0.80235505, + "num_input_tokens_seen": 98041060, + "step": 4544, + "time_per_iteration": 2.6209588050842285 + }, + { + "auxiliary_loss_clip": 0.01190187, + "auxiliary_loss_mlp": 0.01025141, + "balance_loss_clip": 1.05333114, + "balance_loss_mlp": 1.01699889, + "epoch": 0.5465039379546685, + "flos": 20448290390400.0, + "grad_norm": 1.8200997087721449, + "language_loss": 0.74013948, + "learning_rate": 1.7962634513915684e-06, + "loss": 0.7622928, + "num_input_tokens_seen": 98058410, + "step": 4545, + "time_per_iteration": 2.643921375274658 + }, + { + "auxiliary_loss_clip": 0.0117843, + "auxiliary_loss_mlp": 0.01025141, + "balance_loss_clip": 1.0520947, + "balance_loss_mlp": 1.01788688, + "epoch": 0.5466241808453075, + "flos": 17343310003200.0, + "grad_norm": 1.597210438124108, + "language_loss": 0.7942425, + "learning_rate": 1.7954885475083969e-06, + "loss": 0.81627822, + "num_input_tokens_seen": 98076080, + "step": 4546, + "time_per_iteration": 2.557011842727661 + }, + { + "auxiliary_loss_clip": 0.01179963, + "auxiliary_loss_mlp": 0.01023743, + "balance_loss_clip": 1.05186319, + "balance_loss_mlp": 1.01606596, + "epoch": 0.5467444237359466, + "flos": 21617039122560.0, + "grad_norm": 2.1050315408153915, + "language_loss": 0.72822642, + "learning_rate": 1.7947136746494513e-06, + "loss": 0.75026351, + "num_input_tokens_seen": 98096995, + "step": 4547, + "time_per_iteration": 2.6350936889648438 + }, + { + "auxiliary_loss_clip": 0.01187176, + "auxiliary_loss_mlp": 0.01028576, + "balance_loss_clip": 1.0512321, + "balance_loss_mlp": 1.02069032, + "epoch": 0.5468646666265857, + "flos": 24170467196160.0, + "grad_norm": 2.5752177018015376, + "language_loss": 0.877756, + "learning_rate": 1.793938832932277e-06, + "loss": 0.89991355, + "num_input_tokens_seen": 98115105, + "step": 4548, + "time_per_iteration": 2.6471197605133057 + }, + { + "auxiliary_loss_clip": 0.01178598, + "auxiliary_loss_mlp": 0.01024531, + "balance_loss_clip": 1.05202401, + "balance_loss_mlp": 1.01726496, + "epoch": 0.5469849095172248, + "flos": 27527001505920.0, + "grad_norm": 2.18845736799173, + "language_loss": 0.7026968, + "learning_rate": 1.7931640224744185e-06, + "loss": 0.72472811, + "num_input_tokens_seen": 98135655, + "step": 4549, + "time_per_iteration": 2.7034988403320312 + }, + { + "auxiliary_loss_clip": 0.01188734, + "auxiliary_loss_mlp": 0.01023567, + "balance_loss_clip": 1.04191375, + "balance_loss_mlp": 1.01593518, + "epoch": 0.5471051524078638, + "flos": 27964680727680.0, + "grad_norm": 1.5466531902508582, + "language_loss": 0.73993886, + "learning_rate": 1.7923892433934127e-06, + "loss": 0.76206189, + "num_input_tokens_seen": 98156730, + "step": 4550, + "time_per_iteration": 2.735454797744751 + }, + { + "auxiliary_loss_clip": 0.01199432, + "auxiliary_loss_mlp": 0.01378919, + "balance_loss_clip": 1.05144846, + "balance_loss_mlp": 1.00032139, + "epoch": 0.547225395298503, + "flos": 18150510389760.0, + "grad_norm": 1.9252023815850867, + "language_loss": 0.7890169, + "learning_rate": 1.7916144958067939e-06, + "loss": 0.81480038, + "num_input_tokens_seen": 98174590, + "step": 4551, + "time_per_iteration": 2.7259767055511475 + }, + { + "auxiliary_loss_clip": 0.01191026, + "auxiliary_loss_mlp": 0.01026391, + "balance_loss_clip": 1.05123138, + "balance_loss_mlp": 1.01883316, + "epoch": 0.5473456381891421, + "flos": 21361498790400.0, + "grad_norm": 1.7622275276468582, + "language_loss": 0.78989148, + "learning_rate": 1.7908397798320905e-06, + "loss": 0.8120656, + "num_input_tokens_seen": 98194325, + "step": 4552, + "time_per_iteration": 2.831275701522827 + }, + { + "auxiliary_loss_clip": 0.01187578, + "auxiliary_loss_mlp": 0.01379236, + "balance_loss_clip": 1.0514245, + "balance_loss_mlp": 1.00021052, + "epoch": 0.5474658810797811, + "flos": 19932145908480.0, + "grad_norm": 1.7180193845158562, + "language_loss": 0.74836737, + "learning_rate": 1.7900650955868265e-06, + "loss": 0.77403551, + "num_input_tokens_seen": 98213970, + "step": 4553, + "time_per_iteration": 2.6400866508483887 + }, + { + "auxiliary_loss_clip": 0.01192961, + "auxiliary_loss_mlp": 0.01378816, + "balance_loss_clip": 1.05510497, + "balance_loss_mlp": 1.00031853, + "epoch": 0.5475861239704203, + "flos": 50476217264640.0, + "grad_norm": 1.3168930129751562, + "language_loss": 0.76242566, + "learning_rate": 1.7892904431885202e-06, + "loss": 0.7881434, + "num_input_tokens_seen": 98241145, + "step": 4554, + "time_per_iteration": 2.8920328617095947 + }, + { + "auxiliary_loss_clip": 0.01212692, + "auxiliary_loss_mlp": 0.01023032, + "balance_loss_clip": 1.04474616, + "balance_loss_mlp": 1.0155158, + "epoch": 0.5477063668610593, + "flos": 20705123612160.0, + "grad_norm": 1.871287779434614, + "language_loss": 0.75336015, + "learning_rate": 1.788515822754686e-06, + "loss": 0.77571738, + "num_input_tokens_seen": 98261565, + "step": 4555, + "time_per_iteration": 2.7682554721832275 + }, + { + "auxiliary_loss_clip": 0.0120243, + "auxiliary_loss_mlp": 0.01025461, + "balance_loss_clip": 1.04419565, + "balance_loss_mlp": 1.01729572, + "epoch": 0.5478266097516984, + "flos": 19609740408960.0, + "grad_norm": 2.8099794813181322, + "language_loss": 0.78501016, + "learning_rate": 1.7877412344028335e-06, + "loss": 0.80728912, + "num_input_tokens_seen": 98281370, + "step": 4556, + "time_per_iteration": 2.7542026042938232 + }, + { + "auxiliary_loss_clip": 0.01189813, + "auxiliary_loss_mlp": 0.01030575, + "balance_loss_clip": 1.05002904, + "balance_loss_mlp": 1.02325559, + "epoch": 0.5479468526423376, + "flos": 12896599962240.0, + "grad_norm": 2.261088977273096, + "language_loss": 0.77357876, + "learning_rate": 1.7869666782504668e-06, + "loss": 0.79578263, + "num_input_tokens_seen": 98297950, + "step": 4557, + "time_per_iteration": 2.6069629192352295 + }, + { + "auxiliary_loss_clip": 0.01183288, + "auxiliary_loss_mlp": 0.01027021, + "balance_loss_clip": 1.04628479, + "balance_loss_mlp": 1.01935029, + "epoch": 0.5480670955329766, + "flos": 18588800142720.0, + "grad_norm": 1.8032340274179885, + "language_loss": 0.69002843, + "learning_rate": 1.7861921544150867e-06, + "loss": 0.71213156, + "num_input_tokens_seen": 98316800, + "step": 4558, + "time_per_iteration": 2.656235694885254 + }, + { + "auxiliary_loss_clip": 0.01208174, + "auxiliary_loss_mlp": 0.01378904, + "balance_loss_clip": 1.04391694, + "balance_loss_mlp": 1.0002991, + "epoch": 0.5481873384236157, + "flos": 15954608338560.0, + "grad_norm": 1.8114171968912949, + "language_loss": 0.76342201, + "learning_rate": 1.7854176630141856e-06, + "loss": 0.78929281, + "num_input_tokens_seen": 98333935, + "step": 4559, + "time_per_iteration": 2.7245821952819824 + }, + { + "auxiliary_loss_clip": 0.01183023, + "auxiliary_loss_mlp": 0.01037878, + "balance_loss_clip": 1.0537343, + "balance_loss_mlp": 1.02989125, + "epoch": 0.5483075813142548, + "flos": 22783812606720.0, + "grad_norm": 2.0962421098789994, + "language_loss": 0.84584945, + "learning_rate": 1.784643204165255e-06, + "loss": 0.86805844, + "num_input_tokens_seen": 98353255, + "step": 4560, + "time_per_iteration": 3.58347225189209 + }, + { + "auxiliary_loss_clip": 0.01183709, + "auxiliary_loss_mlp": 0.01028089, + "balance_loss_clip": 1.05196857, + "balance_loss_mlp": 1.02065599, + "epoch": 0.5484278242048939, + "flos": 19317212046720.0, + "grad_norm": 1.9967687278467803, + "language_loss": 0.77341276, + "learning_rate": 1.7838687779857783e-06, + "loss": 0.79553074, + "num_input_tokens_seen": 98371130, + "step": 4561, + "time_per_iteration": 3.7134902477264404 + }, + { + "auxiliary_loss_clip": 0.01189946, + "auxiliary_loss_mlp": 0.01026066, + "balance_loss_clip": 1.04798698, + "balance_loss_mlp": 1.01892519, + "epoch": 0.5485480670955329, + "flos": 22816024128000.0, + "grad_norm": 2.2807723628268293, + "language_loss": 0.63883823, + "learning_rate": 1.7830943845932366e-06, + "loss": 0.66099834, + "num_input_tokens_seen": 98390455, + "step": 4562, + "time_per_iteration": 3.579211950302124 + }, + { + "auxiliary_loss_clip": 0.01203333, + "auxiliary_loss_mlp": 0.01029125, + "balance_loss_clip": 1.05146861, + "balance_loss_mlp": 1.0212456, + "epoch": 0.5486683099861721, + "flos": 22671304231680.0, + "grad_norm": 1.5940838867430112, + "language_loss": 0.74909198, + "learning_rate": 1.7823200241051044e-06, + "loss": 0.77141654, + "num_input_tokens_seen": 98409370, + "step": 4563, + "time_per_iteration": 2.7095894813537598 + }, + { + "auxiliary_loss_clip": 0.01182703, + "auxiliary_loss_mlp": 0.01029345, + "balance_loss_clip": 1.05492151, + "balance_loss_mlp": 1.02206135, + "epoch": 0.5487885528768112, + "flos": 23149383275520.0, + "grad_norm": 2.2424069346939013, + "language_loss": 0.80682695, + "learning_rate": 1.7815456966388513e-06, + "loss": 0.82894742, + "num_input_tokens_seen": 98428465, + "step": 4564, + "time_per_iteration": 3.5513107776641846 + }, + { + "auxiliary_loss_clip": 0.01205643, + "auxiliary_loss_mlp": 0.01027742, + "balance_loss_clip": 1.04741609, + "balance_loss_mlp": 1.01983261, + "epoch": 0.5489087957674502, + "flos": 22053928245120.0, + "grad_norm": 2.8131580339460616, + "language_loss": 0.80770755, + "learning_rate": 1.780771402311943e-06, + "loss": 0.83004141, + "num_input_tokens_seen": 98447300, + "step": 4565, + "time_per_iteration": 2.7732269763946533 + }, + { + "auxiliary_loss_clip": 0.01198435, + "auxiliary_loss_mlp": 0.01030568, + "balance_loss_clip": 1.05173612, + "balance_loss_mlp": 1.02289152, + "epoch": 0.5490290386580894, + "flos": 24315977191680.0, + "grad_norm": 1.9336786902189187, + "language_loss": 0.78717685, + "learning_rate": 1.7799971412418374e-06, + "loss": 0.8094669, + "num_input_tokens_seen": 98468695, + "step": 4566, + "time_per_iteration": 2.7881054878234863 + }, + { + "auxiliary_loss_clip": 0.01202426, + "auxiliary_loss_mlp": 0.01028563, + "balance_loss_clip": 1.04598141, + "balance_loss_mlp": 1.02003407, + "epoch": 0.5491492815487284, + "flos": 18294942977280.0, + "grad_norm": 1.9522857167786267, + "language_loss": 0.74205053, + "learning_rate": 1.7792229135459918e-06, + "loss": 0.76436043, + "num_input_tokens_seen": 98485345, + "step": 4567, + "time_per_iteration": 2.755030632019043 + }, + { + "auxiliary_loss_clip": 0.01126779, + "auxiliary_loss_mlp": 0.01003379, + "balance_loss_clip": 1.02649796, + "balance_loss_mlp": 1.00224042, + "epoch": 0.5492695244393675, + "flos": 64550257050240.0, + "grad_norm": 0.7373572539241214, + "language_loss": 0.61634552, + "learning_rate": 1.7784487193418538e-06, + "loss": 0.63764709, + "num_input_tokens_seen": 98543195, + "step": 4568, + "time_per_iteration": 3.1960840225219727 + }, + { + "auxiliary_loss_clip": 0.01186583, + "auxiliary_loss_mlp": 0.01028023, + "balance_loss_clip": 1.04204655, + "balance_loss_mlp": 1.01976156, + "epoch": 0.5493897673300067, + "flos": 17379579761280.0, + "grad_norm": 2.012002219072116, + "language_loss": 0.61230743, + "learning_rate": 1.7776745587468698e-06, + "loss": 0.63445354, + "num_input_tokens_seen": 98560620, + "step": 4569, + "time_per_iteration": 2.7647180557250977 + }, + { + "auxiliary_loss_clip": 0.01178573, + "auxiliary_loss_mlp": 0.01027875, + "balance_loss_clip": 1.05043364, + "balance_loss_mlp": 1.01969433, + "epoch": 0.5495100102206457, + "flos": 19901765980800.0, + "grad_norm": 2.3063939208049926, + "language_loss": 0.81398821, + "learning_rate": 1.7769004318784776e-06, + "loss": 0.83605266, + "num_input_tokens_seen": 98578265, + "step": 4570, + "time_per_iteration": 2.6697194576263428 + }, + { + "auxiliary_loss_clip": 0.01189698, + "auxiliary_loss_mlp": 0.01029523, + "balance_loss_clip": 1.05129504, + "balance_loss_mlp": 1.02186394, + "epoch": 0.5496302531112848, + "flos": 16727190992640.0, + "grad_norm": 1.7731303406578127, + "language_loss": 0.80845845, + "learning_rate": 1.776126338854113e-06, + "loss": 0.83065063, + "num_input_tokens_seen": 98596055, + "step": 4571, + "time_per_iteration": 2.5982706546783447 + }, + { + "auxiliary_loss_clip": 0.01183935, + "auxiliary_loss_mlp": 0.01031397, + "balance_loss_clip": 1.05105329, + "balance_loss_mlp": 1.02358603, + "epoch": 0.5497504960019239, + "flos": 24572343536640.0, + "grad_norm": 1.896686630517489, + "language_loss": 0.8459574, + "learning_rate": 1.7753522797912044e-06, + "loss": 0.86811066, + "num_input_tokens_seen": 98616140, + "step": 4572, + "time_per_iteration": 2.6581180095672607 + }, + { + "auxiliary_loss_clip": 0.01202993, + "auxiliary_loss_mlp": 0.01030829, + "balance_loss_clip": 1.0493753, + "balance_loss_mlp": 1.02256227, + "epoch": 0.549870738892563, + "flos": 15450494912640.0, + "grad_norm": 3.2830749865021227, + "language_loss": 0.69925618, + "learning_rate": 1.7745782548071765e-06, + "loss": 0.72159445, + "num_input_tokens_seen": 98633035, + "step": 4573, + "time_per_iteration": 2.6831328868865967 + }, + { + "auxiliary_loss_clip": 0.01203704, + "auxiliary_loss_mlp": 0.01030188, + "balance_loss_clip": 1.05081046, + "balance_loss_mlp": 1.02277946, + "epoch": 0.549990981783202, + "flos": 21069114082560.0, + "grad_norm": 1.6448639434432208, + "language_loss": 0.74433362, + "learning_rate": 1.7738042640194482e-06, + "loss": 0.76667255, + "num_input_tokens_seen": 98652700, + "step": 4574, + "time_per_iteration": 2.7068169116973877 + }, + { + "auxiliary_loss_clip": 0.01179659, + "auxiliary_loss_mlp": 0.01029292, + "balance_loss_clip": 1.05268764, + "balance_loss_mlp": 1.02144253, + "epoch": 0.5501112246738411, + "flos": 21395901041280.0, + "grad_norm": 3.7947103937758198, + "language_loss": 0.7032876, + "learning_rate": 1.7730303075454335e-06, + "loss": 0.72537708, + "num_input_tokens_seen": 98671590, + "step": 4575, + "time_per_iteration": 2.616074800491333 + }, + { + "auxiliary_loss_clip": 0.01211331, + "auxiliary_loss_mlp": 0.01030932, + "balance_loss_clip": 1.04875743, + "balance_loss_mlp": 1.02280164, + "epoch": 0.5502314675644803, + "flos": 17456931699840.0, + "grad_norm": 1.9673985885065073, + "language_loss": 0.85078919, + "learning_rate": 1.7722563855025402e-06, + "loss": 0.87321186, + "num_input_tokens_seen": 98689620, + "step": 4576, + "time_per_iteration": 2.682798385620117 + }, + { + "auxiliary_loss_clip": 0.01195274, + "auxiliary_loss_mlp": 0.01024439, + "balance_loss_clip": 1.04645371, + "balance_loss_mlp": 1.01643109, + "epoch": 0.5503517104551193, + "flos": 24310410583680.0, + "grad_norm": 2.054111621476412, + "language_loss": 0.70769405, + "learning_rate": 1.7714824980081721e-06, + "loss": 0.72989112, + "num_input_tokens_seen": 98708915, + "step": 4577, + "time_per_iteration": 2.747821092605591 + }, + { + "auxiliary_loss_clip": 0.01186108, + "auxiliary_loss_mlp": 0.01026914, + "balance_loss_clip": 1.05253029, + "balance_loss_mlp": 1.01915383, + "epoch": 0.5504719533457584, + "flos": 22419427086720.0, + "grad_norm": 2.6819585382513096, + "language_loss": 0.73686922, + "learning_rate": 1.7707086451797276e-06, + "loss": 0.75899947, + "num_input_tokens_seen": 98729790, + "step": 4578, + "time_per_iteration": 2.708474636077881 + }, + { + "auxiliary_loss_clip": 0.01115661, + "auxiliary_loss_mlp": 0.01001093, + "balance_loss_clip": 1.02221966, + "balance_loss_mlp": 0.99983543, + "epoch": 0.5505921962363975, + "flos": 67294155968640.0, + "grad_norm": 0.7028897684089057, + "language_loss": 0.52370536, + "learning_rate": 1.7699348271345993e-06, + "loss": 0.54487288, + "num_input_tokens_seen": 98792415, + "step": 4579, + "time_per_iteration": 3.2432422637939453 + }, + { + "auxiliary_loss_clip": 0.01130942, + "auxiliary_loss_mlp": 0.01001992, + "balance_loss_clip": 1.01955938, + "balance_loss_mlp": 1.000705, + "epoch": 0.5507124391270366, + "flos": 45685125578880.0, + "grad_norm": 0.7176306433121864, + "language_loss": 0.54466224, + "learning_rate": 1.7691610439901753e-06, + "loss": 0.56599164, + "num_input_tokens_seen": 98855350, + "step": 4580, + "time_per_iteration": 3.361020565032959 + }, + { + "auxiliary_loss_clip": 0.01190808, + "auxiliary_loss_mlp": 0.01026393, + "balance_loss_clip": 1.05226326, + "balance_loss_mlp": 1.01881385, + "epoch": 0.5508326820176757, + "flos": 22273845264000.0, + "grad_norm": 1.7952085853129485, + "language_loss": 0.75314367, + "learning_rate": 1.7683872958638367e-06, + "loss": 0.77531564, + "num_input_tokens_seen": 98874230, + "step": 4581, + "time_per_iteration": 2.674431324005127 + }, + { + "auxiliary_loss_clip": 0.01194598, + "auxiliary_loss_mlp": 0.01028031, + "balance_loss_clip": 1.04786193, + "balance_loss_mlp": 1.02015746, + "epoch": 0.5509529249083148, + "flos": 20012442762240.0, + "grad_norm": 1.9451745224225647, + "language_loss": 0.84236938, + "learning_rate": 1.7676135828729614e-06, + "loss": 0.86459565, + "num_input_tokens_seen": 98893940, + "step": 4582, + "time_per_iteration": 2.7216460704803467 + }, + { + "auxiliary_loss_clip": 0.01188352, + "auxiliary_loss_mlp": 0.01025679, + "balance_loss_clip": 1.05154443, + "balance_loss_mlp": 1.0179069, + "epoch": 0.5510731677989539, + "flos": 21834801325440.0, + "grad_norm": 1.8286091385865915, + "language_loss": 0.83009779, + "learning_rate": 1.7668399051349205e-06, + "loss": 0.85223818, + "num_input_tokens_seen": 98913620, + "step": 4583, + "time_per_iteration": 2.6449828147888184 + }, + { + "auxiliary_loss_clip": 0.01202478, + "auxiliary_loss_mlp": 0.01026912, + "balance_loss_clip": 1.04821384, + "balance_loss_mlp": 1.01909196, + "epoch": 0.5511934106895929, + "flos": 21467901853440.0, + "grad_norm": 2.222772040681578, + "language_loss": 0.83006471, + "learning_rate": 1.766066262767081e-06, + "loss": 0.8523587, + "num_input_tokens_seen": 98931460, + "step": 4584, + "time_per_iteration": 2.7367448806762695 + }, + { + "auxiliary_loss_clip": 0.01191189, + "auxiliary_loss_mlp": 0.01026262, + "balance_loss_clip": 1.05007946, + "balance_loss_mlp": 1.01860607, + "epoch": 0.5513136535802321, + "flos": 21068934514560.0, + "grad_norm": 2.35150586329892, + "language_loss": 0.77269226, + "learning_rate": 1.765292655886803e-06, + "loss": 0.7948668, + "num_input_tokens_seen": 98950105, + "step": 4585, + "time_per_iteration": 2.6445446014404297 + }, + { + "auxiliary_loss_clip": 0.01215872, + "auxiliary_loss_mlp": 0.01026301, + "balance_loss_clip": 1.05036736, + "balance_loss_mlp": 1.01838589, + "epoch": 0.5514338964708712, + "flos": 27815004754560.0, + "grad_norm": 1.7290653147608843, + "language_loss": 0.70553553, + "learning_rate": 1.764519084611443e-06, + "loss": 0.72795725, + "num_input_tokens_seen": 98970560, + "step": 4586, + "time_per_iteration": 2.786937713623047 + }, + { + "auxiliary_loss_clip": 0.01199298, + "auxiliary_loss_mlp": 0.01033552, + "balance_loss_clip": 1.05150759, + "balance_loss_mlp": 1.02527928, + "epoch": 0.5515541393615102, + "flos": 21908525990400.0, + "grad_norm": 2.348702415350108, + "language_loss": 0.77758604, + "learning_rate": 1.7637455490583505e-06, + "loss": 0.79991454, + "num_input_tokens_seen": 98989885, + "step": 4587, + "time_per_iteration": 3.730424404144287 + }, + { + "auxiliary_loss_clip": 0.01192657, + "auxiliary_loss_mlp": 0.01028034, + "balance_loss_clip": 1.05507624, + "balance_loss_mlp": 1.02056587, + "epoch": 0.5516743822521494, + "flos": 20485422074880.0, + "grad_norm": 2.1858492886582397, + "language_loss": 0.77664441, + "learning_rate": 1.7629720493448701e-06, + "loss": 0.79885131, + "num_input_tokens_seen": 99007180, + "step": 4588, + "time_per_iteration": 4.4246532917022705 + }, + { + "auxiliary_loss_clip": 0.01203746, + "auxiliary_loss_mlp": 0.01036689, + "balance_loss_clip": 1.05019045, + "balance_loss_mlp": 1.0289166, + "epoch": 0.5517946251427884, + "flos": 14940383915520.0, + "grad_norm": 1.8064837306281298, + "language_loss": 0.84902024, + "learning_rate": 1.7621985855883418e-06, + "loss": 0.87142456, + "num_input_tokens_seen": 99023880, + "step": 4589, + "time_per_iteration": 2.6757631301879883 + }, + { + "auxiliary_loss_clip": 0.0119791, + "auxiliary_loss_mlp": 0.01027238, + "balance_loss_clip": 1.053864, + "balance_loss_mlp": 1.01963878, + "epoch": 0.5519148680334275, + "flos": 18404865573120.0, + "grad_norm": 1.8931313516659523, + "language_loss": 0.7245189, + "learning_rate": 1.7614251579060983e-06, + "loss": 0.74677038, + "num_input_tokens_seen": 99042475, + "step": 4590, + "time_per_iteration": 3.589261770248413 + }, + { + "auxiliary_loss_clip": 0.01206648, + "auxiliary_loss_mlp": 0.01027755, + "balance_loss_clip": 1.0486145, + "balance_loss_mlp": 1.02003574, + "epoch": 0.5520351109240667, + "flos": 25113337251840.0, + "grad_norm": 2.0770468088088783, + "language_loss": 0.84519124, + "learning_rate": 1.76065176641547e-06, + "loss": 0.86753523, + "num_input_tokens_seen": 99065185, + "step": 4591, + "time_per_iteration": 2.7746052742004395 + }, + { + "auxiliary_loss_clip": 0.01187811, + "auxiliary_loss_mlp": 0.0102487, + "balance_loss_clip": 1.04821301, + "balance_loss_mlp": 1.01723492, + "epoch": 0.5521553538147057, + "flos": 21069545045760.0, + "grad_norm": 1.7856041688063784, + "language_loss": 0.77913445, + "learning_rate": 1.759878411233777e-06, + "loss": 0.80126131, + "num_input_tokens_seen": 99083645, + "step": 4592, + "time_per_iteration": 2.7313647270202637 + }, + { + "auxiliary_loss_clip": 0.011879, + "auxiliary_loss_mlp": 0.01025094, + "balance_loss_clip": 1.0513587, + "balance_loss_mlp": 1.01709509, + "epoch": 0.5522755967053448, + "flos": 18879999701760.0, + "grad_norm": 2.584417960905068, + "language_loss": 0.75595659, + "learning_rate": 1.7591050924783388e-06, + "loss": 0.77808654, + "num_input_tokens_seen": 99100835, + "step": 4593, + "time_per_iteration": 2.6499521732330322 + }, + { + "auxiliary_loss_clip": 0.01126999, + "auxiliary_loss_mlp": 0.01000042, + "balance_loss_clip": 1.01968527, + "balance_loss_mlp": 0.99871331, + "epoch": 0.5523958395959839, + "flos": 64675622494080.0, + "grad_norm": 0.8353411144047057, + "language_loss": 0.57879913, + "learning_rate": 1.7583318102664661e-06, + "loss": 0.60006952, + "num_input_tokens_seen": 99168400, + "step": 4594, + "time_per_iteration": 3.4595134258270264 + }, + { + "auxiliary_loss_clip": 0.01188012, + "auxiliary_loss_mlp": 0.01022925, + "balance_loss_clip": 1.04735887, + "balance_loss_mlp": 1.01534879, + "epoch": 0.552516082486623, + "flos": 10889732211840.0, + "grad_norm": 1.8284796530857894, + "language_loss": 0.79168928, + "learning_rate": 1.757558564715466e-06, + "loss": 0.81379867, + "num_input_tokens_seen": 99186475, + "step": 4595, + "time_per_iteration": 2.809256076812744 + }, + { + "auxiliary_loss_clip": 0.01190969, + "auxiliary_loss_mlp": 0.01023613, + "balance_loss_clip": 1.05092192, + "balance_loss_mlp": 1.01560783, + "epoch": 0.552636325377262, + "flos": 22199797376640.0, + "grad_norm": 2.954275610987667, + "language_loss": 0.7385993, + "learning_rate": 1.7567853559426386e-06, + "loss": 0.76074517, + "num_input_tokens_seen": 99203525, + "step": 4596, + "time_per_iteration": 2.598959445953369 + }, + { + "auxiliary_loss_clip": 0.01191269, + "auxiliary_loss_mlp": 0.01024933, + "balance_loss_clip": 1.05143094, + "balance_loss_mlp": 1.01690125, + "epoch": 0.5527565682679012, + "flos": 23988184652160.0, + "grad_norm": 2.0902975375743185, + "language_loss": 0.75028604, + "learning_rate": 1.7560121840652797e-06, + "loss": 0.772448, + "num_input_tokens_seen": 99222910, + "step": 4597, + "time_per_iteration": 2.664628744125366 + }, + { + "auxiliary_loss_clip": 0.01192021, + "auxiliary_loss_mlp": 0.01023406, + "balance_loss_clip": 1.04513478, + "balance_loss_mlp": 1.01545477, + "epoch": 0.5528768111585403, + "flos": 19719267955200.0, + "grad_norm": 1.7589703457141945, + "language_loss": 0.6871413, + "learning_rate": 1.7552390492006782e-06, + "loss": 0.70929551, + "num_input_tokens_seen": 99241230, + "step": 4598, + "time_per_iteration": 2.673248291015625 + }, + { + "auxiliary_loss_clip": 0.01220453, + "auxiliary_loss_mlp": 0.01379152, + "balance_loss_clip": 1.0455873, + "balance_loss_mlp": 1.00034571, + "epoch": 0.5529970540491793, + "flos": 26215975002240.0, + "grad_norm": 1.9129585354638663, + "language_loss": 0.6498431, + "learning_rate": 1.7544659514661184e-06, + "loss": 0.67583919, + "num_input_tokens_seen": 99264320, + "step": 4599, + "time_per_iteration": 2.831949234008789 + }, + { + "auxiliary_loss_clip": 0.01190518, + "auxiliary_loss_mlp": 0.01027073, + "balance_loss_clip": 1.04603779, + "balance_loss_mlp": 1.01920533, + "epoch": 0.5531172969398185, + "flos": 24425971614720.0, + "grad_norm": 2.0557906378197717, + "language_loss": 0.79918826, + "learning_rate": 1.7536928909788786e-06, + "loss": 0.82136416, + "num_input_tokens_seen": 99283625, + "step": 4600, + "time_per_iteration": 2.676812171936035 + }, + { + "auxiliary_loss_clip": 0.01127707, + "auxiliary_loss_mlp": 0.01000726, + "balance_loss_clip": 1.017138, + "balance_loss_mlp": 0.99945015, + "epoch": 0.5532375398304575, + "flos": 64907316195840.0, + "grad_norm": 0.881009239251243, + "language_loss": 0.62023866, + "learning_rate": 1.752919867856231e-06, + "loss": 0.641523, + "num_input_tokens_seen": 99335270, + "step": 4601, + "time_per_iteration": 3.153716564178467 + }, + { + "auxiliary_loss_clip": 0.01187146, + "auxiliary_loss_mlp": 0.01032119, + "balance_loss_clip": 1.04640794, + "balance_loss_mlp": 1.02465677, + "epoch": 0.5533577827210966, + "flos": 19683105937920.0, + "grad_norm": 2.0953495712241708, + "language_loss": 0.7902776, + "learning_rate": 1.7521468822154436e-06, + "loss": 0.8124702, + "num_input_tokens_seen": 99354185, + "step": 4602, + "time_per_iteration": 2.6395251750946045 + }, + { + "auxiliary_loss_clip": 0.01196318, + "auxiliary_loss_mlp": 0.01026334, + "balance_loss_clip": 1.05369008, + "balance_loss_mlp": 1.01857948, + "epoch": 0.5534780256117358, + "flos": 32306496076800.0, + "grad_norm": 2.3181416056757738, + "language_loss": 0.7504257, + "learning_rate": 1.751373934173777e-06, + "loss": 0.77265215, + "num_input_tokens_seen": 99376930, + "step": 4603, + "time_per_iteration": 2.857050657272339 + }, + { + "auxiliary_loss_clip": 0.01179371, + "auxiliary_loss_mlp": 0.01027335, + "balance_loss_clip": 1.05028033, + "balance_loss_mlp": 1.01962829, + "epoch": 0.5535982685023748, + "flos": 23222425582080.0, + "grad_norm": 1.5507375706848958, + "language_loss": 0.72946846, + "learning_rate": 1.750601023848487e-06, + "loss": 0.75153553, + "num_input_tokens_seen": 99397655, + "step": 4604, + "time_per_iteration": 2.654961347579956 + }, + { + "auxiliary_loss_clip": 0.01178462, + "auxiliary_loss_mlp": 0.01378663, + "balance_loss_clip": 1.05296934, + "balance_loss_mlp": 1.00024235, + "epoch": 0.5537185113930139, + "flos": 24352534258560.0, + "grad_norm": 1.994104145798645, + "language_loss": 0.7365371, + "learning_rate": 1.749828151356823e-06, + "loss": 0.76210833, + "num_input_tokens_seen": 99417850, + "step": 4605, + "time_per_iteration": 2.6511895656585693 + }, + { + "auxiliary_loss_clip": 0.01197846, + "auxiliary_loss_mlp": 0.01024815, + "balance_loss_clip": 1.04972959, + "balance_loss_mlp": 1.01744175, + "epoch": 0.553838754283653, + "flos": 23549068886400.0, + "grad_norm": 1.644907269533533, + "language_loss": 0.75491381, + "learning_rate": 1.7490553168160297e-06, + "loss": 0.77714038, + "num_input_tokens_seen": 99438920, + "step": 4606, + "time_per_iteration": 2.7231545448303223 + }, + { + "auxiliary_loss_clip": 0.01194345, + "auxiliary_loss_mlp": 0.01025682, + "balance_loss_clip": 1.04800248, + "balance_loss_mlp": 1.01867199, + "epoch": 0.5539589971742921, + "flos": 17275044205440.0, + "grad_norm": 2.26441030144546, + "language_loss": 0.76722622, + "learning_rate": 1.748282520343345e-06, + "loss": 0.78942651, + "num_input_tokens_seen": 99457950, + "step": 4607, + "time_per_iteration": 2.665297746658325 + }, + { + "auxiliary_loss_clip": 0.01194196, + "auxiliary_loss_mlp": 0.01028157, + "balance_loss_clip": 1.05157042, + "balance_loss_mlp": 1.02043533, + "epoch": 0.5540792400649311, + "flos": 27564169104000.0, + "grad_norm": 2.0355169002525053, + "language_loss": 0.78707945, + "learning_rate": 1.7475097620560023e-06, + "loss": 0.80930293, + "num_input_tokens_seen": 99478015, + "step": 4608, + "time_per_iteration": 2.7565269470214844 + }, + { + "auxiliary_loss_clip": 0.01176478, + "auxiliary_loss_mlp": 0.01023083, + "balance_loss_clip": 1.05100632, + "balance_loss_mlp": 1.01575136, + "epoch": 0.5541994829555702, + "flos": 23878657105920.0, + "grad_norm": 1.83567038132216, + "language_loss": 0.71292019, + "learning_rate": 1.746737042071228e-06, + "loss": 0.73491585, + "num_input_tokens_seen": 99496520, + "step": 4609, + "time_per_iteration": 2.5956125259399414 + }, + { + "auxiliary_loss_clip": 0.01192098, + "auxiliary_loss_mlp": 0.01029792, + "balance_loss_clip": 1.05066597, + "balance_loss_mlp": 1.02266383, + "epoch": 0.5543197258462094, + "flos": 20115721342080.0, + "grad_norm": 1.9071393199946516, + "language_loss": 0.79265296, + "learning_rate": 1.7459643605062424e-06, + "loss": 0.81487191, + "num_input_tokens_seen": 99513780, + "step": 4610, + "time_per_iteration": 2.688256025314331 + }, + { + "auxiliary_loss_clip": 0.01216334, + "auxiliary_loss_mlp": 0.01025734, + "balance_loss_clip": 1.05138004, + "balance_loss_mlp": 1.01802111, + "epoch": 0.5544399687368484, + "flos": 20916565021440.0, + "grad_norm": 2.0887887334657598, + "language_loss": 0.80815077, + "learning_rate": 1.745191717478262e-06, + "loss": 0.83057141, + "num_input_tokens_seen": 99532360, + "step": 4611, + "time_per_iteration": 2.7876734733581543 + }, + { + "auxiliary_loss_clip": 0.01190112, + "auxiliary_loss_mlp": 0.01026274, + "balance_loss_clip": 1.0480926, + "balance_loss_mlp": 1.01821589, + "epoch": 0.5545602116274875, + "flos": 25518661297920.0, + "grad_norm": 1.751646332985938, + "language_loss": 0.79480463, + "learning_rate": 1.7444191131044948e-06, + "loss": 0.81696844, + "num_input_tokens_seen": 99552635, + "step": 4612, + "time_per_iteration": 2.754035472869873 + }, + { + "auxiliary_loss_clip": 0.01197135, + "auxiliary_loss_mlp": 0.01026756, + "balance_loss_clip": 1.05100894, + "balance_loss_mlp": 1.01874554, + "epoch": 0.5546804545181266, + "flos": 20995568985600.0, + "grad_norm": 1.7437665751793436, + "language_loss": 0.73201156, + "learning_rate": 1.7436465475021456e-06, + "loss": 0.75425053, + "num_input_tokens_seen": 99572685, + "step": 4613, + "time_per_iteration": 4.662269353866577 + }, + { + "auxiliary_loss_clip": 0.01196034, + "auxiliary_loss_mlp": 0.01024231, + "balance_loss_clip": 1.0462625, + "balance_loss_mlp": 1.01667297, + "epoch": 0.5548006974087657, + "flos": 26833638297600.0, + "grad_norm": 2.299812414896192, + "language_loss": 0.71433985, + "learning_rate": 1.7428740207884111e-06, + "loss": 0.73654246, + "num_input_tokens_seen": 99593565, + "step": 4614, + "time_per_iteration": 3.740806818008423 + }, + { + "auxiliary_loss_clip": 0.01217041, + "auxiliary_loss_mlp": 0.01025471, + "balance_loss_clip": 1.04719043, + "balance_loss_mlp": 1.01755559, + "epoch": 0.5549209402994048, + "flos": 33656414031360.0, + "grad_norm": 1.7275759470797425, + "language_loss": 0.60890484, + "learning_rate": 1.7421015330804833e-06, + "loss": 0.63133001, + "num_input_tokens_seen": 99613485, + "step": 4615, + "time_per_iteration": 2.7815072536468506 + }, + { + "auxiliary_loss_clip": 0.01177116, + "auxiliary_loss_mlp": 0.01024884, + "balance_loss_clip": 1.05051219, + "balance_loss_mlp": 1.01702213, + "epoch": 0.5550411831900439, + "flos": 23769524609280.0, + "grad_norm": 1.8476457451717478, + "language_loss": 0.72752076, + "learning_rate": 1.7413290844955475e-06, + "loss": 0.74954075, + "num_input_tokens_seen": 99633515, + "step": 4616, + "time_per_iteration": 3.4645230770111084 + }, + { + "auxiliary_loss_clip": 0.0117831, + "auxiliary_loss_mlp": 0.01030152, + "balance_loss_clip": 1.05132461, + "balance_loss_mlp": 1.02285016, + "epoch": 0.555161426080683, + "flos": 21651189978240.0, + "grad_norm": 2.131382564984793, + "language_loss": 0.7834838, + "learning_rate": 1.7405566751507843e-06, + "loss": 0.80556846, + "num_input_tokens_seen": 99651560, + "step": 4617, + "time_per_iteration": 2.697322130203247 + }, + { + "auxiliary_loss_clip": 0.01205739, + "auxiliary_loss_mlp": 0.010199, + "balance_loss_clip": 1.04963303, + "balance_loss_mlp": 1.01249719, + "epoch": 0.555281668971322, + "flos": 49563116605440.0, + "grad_norm": 2.0455563760712043, + "language_loss": 0.68000019, + "learning_rate": 1.7397843051633668e-06, + "loss": 0.70225656, + "num_input_tokens_seen": 99674255, + "step": 4618, + "time_per_iteration": 2.92587947845459 + }, + { + "auxiliary_loss_clip": 0.0118507, + "auxiliary_loss_mlp": 0.01025592, + "balance_loss_clip": 1.05164027, + "balance_loss_mlp": 1.01838613, + "epoch": 0.5554019118619612, + "flos": 20741608851840.0, + "grad_norm": 2.1335977962228707, + "language_loss": 0.71648061, + "learning_rate": 1.739011974650464e-06, + "loss": 0.73858726, + "num_input_tokens_seen": 99693585, + "step": 4619, + "time_per_iteration": 2.587373971939087 + }, + { + "auxiliary_loss_clip": 0.01224695, + "auxiliary_loss_mlp": 0.01034015, + "balance_loss_clip": 1.0514729, + "balance_loss_mlp": 1.02568841, + "epoch": 0.5555221547526003, + "flos": 25483217552640.0, + "grad_norm": 2.04072198252209, + "language_loss": 0.7647621, + "learning_rate": 1.7382396837292365e-06, + "loss": 0.78734922, + "num_input_tokens_seen": 99714045, + "step": 4620, + "time_per_iteration": 2.7853903770446777 + }, + { + "auxiliary_loss_clip": 0.01181174, + "auxiliary_loss_mlp": 0.01030614, + "balance_loss_clip": 1.05460668, + "balance_loss_mlp": 1.02232289, + "epoch": 0.5556423976432393, + "flos": 21762513204480.0, + "grad_norm": 1.6688508740765022, + "language_loss": 0.73562157, + "learning_rate": 1.737467432516841e-06, + "loss": 0.75773948, + "num_input_tokens_seen": 99734145, + "step": 4621, + "time_per_iteration": 2.6106584072113037 + }, + { + "auxiliary_loss_clip": 0.01194408, + "auxiliary_loss_mlp": 0.01028951, + "balance_loss_clip": 1.04729033, + "balance_loss_mlp": 1.02160192, + "epoch": 0.5557626405338785, + "flos": 24900171989760.0, + "grad_norm": 2.5525040374488586, + "language_loss": 0.74542958, + "learning_rate": 1.7366952211304274e-06, + "loss": 0.76766318, + "num_input_tokens_seen": 99751990, + "step": 4622, + "time_per_iteration": 2.7893149852752686 + }, + { + "auxiliary_loss_clip": 0.01185266, + "auxiliary_loss_mlp": 0.01024938, + "balance_loss_clip": 1.04649782, + "balance_loss_mlp": 1.01698053, + "epoch": 0.5558828834245175, + "flos": 18697501676160.0, + "grad_norm": 1.9926157205267918, + "language_loss": 0.8308562, + "learning_rate": 1.735923049687139e-06, + "loss": 0.85295826, + "num_input_tokens_seen": 99768565, + "step": 4623, + "time_per_iteration": 2.667311668395996 + }, + { + "auxiliary_loss_clip": 0.01188091, + "auxiliary_loss_mlp": 0.01027615, + "balance_loss_clip": 1.04744279, + "balance_loss_mlp": 1.01950264, + "epoch": 0.5560031263151566, + "flos": 27272179445760.0, + "grad_norm": 1.5649643205766421, + "language_loss": 0.73704135, + "learning_rate": 1.7351509183041144e-06, + "loss": 0.75919843, + "num_input_tokens_seen": 99788895, + "step": 4624, + "time_per_iteration": 2.6818912029266357 + }, + { + "auxiliary_loss_clip": 0.01181785, + "auxiliary_loss_mlp": 0.01027709, + "balance_loss_clip": 1.05353904, + "balance_loss_mlp": 1.01907182, + "epoch": 0.5561233692057957, + "flos": 23403738458880.0, + "grad_norm": 1.7280493226400975, + "language_loss": 0.717224, + "learning_rate": 1.7343788270984852e-06, + "loss": 0.73931897, + "num_input_tokens_seen": 99808035, + "step": 4625, + "time_per_iteration": 2.6496307849884033 + }, + { + "auxiliary_loss_clip": 0.01197303, + "auxiliary_loss_mlp": 0.01025365, + "balance_loss_clip": 1.05228305, + "balance_loss_mlp": 1.01725888, + "epoch": 0.5562436120964348, + "flos": 37670867804160.0, + "grad_norm": 2.230695898180457, + "language_loss": 0.74899745, + "learning_rate": 1.7336067761873764e-06, + "loss": 0.77122414, + "num_input_tokens_seen": 99830460, + "step": 4626, + "time_per_iteration": 2.773380994796753 + }, + { + "auxiliary_loss_clip": 0.0119388, + "auxiliary_loss_mlp": 0.01031273, + "balance_loss_clip": 1.05031955, + "balance_loss_mlp": 1.02269602, + "epoch": 0.5563638549870739, + "flos": 25155245445120.0, + "grad_norm": 2.072867997510411, + "language_loss": 0.76511657, + "learning_rate": 1.7328347656879076e-06, + "loss": 0.78736812, + "num_input_tokens_seen": 99850320, + "step": 4627, + "time_per_iteration": 2.6520979404449463 + }, + { + "auxiliary_loss_clip": 0.01202053, + "auxiliary_loss_mlp": 0.01025999, + "balance_loss_clip": 1.04707503, + "balance_loss_mlp": 1.01800621, + "epoch": 0.556484097877713, + "flos": 13581810783360.0, + "grad_norm": 2.1836758383730337, + "language_loss": 0.68390179, + "learning_rate": 1.7320627957171927e-06, + "loss": 0.7061823, + "num_input_tokens_seen": 99864980, + "step": 4628, + "time_per_iteration": 2.647186517715454 + }, + { + "auxiliary_loss_clip": 0.01179505, + "auxiliary_loss_mlp": 0.01021389, + "balance_loss_clip": 1.05345607, + "balance_loss_mlp": 1.01343155, + "epoch": 0.5566043407683521, + "flos": 24681368292480.0, + "grad_norm": 1.6820289160128548, + "language_loss": 0.81272006, + "learning_rate": 1.7312908663923382e-06, + "loss": 0.83472896, + "num_input_tokens_seen": 99881155, + "step": 4629, + "time_per_iteration": 2.6441421508789062 + }, + { + "auxiliary_loss_clip": 0.01178062, + "auxiliary_loss_mlp": 0.01031004, + "balance_loss_clip": 1.04689193, + "balance_loss_mlp": 1.02281392, + "epoch": 0.5567245836589911, + "flos": 20588161950720.0, + "grad_norm": 1.9672294283478988, + "language_loss": 0.6741907, + "learning_rate": 1.7305189778304463e-06, + "loss": 0.69628137, + "num_input_tokens_seen": 99899330, + "step": 4630, + "time_per_iteration": 2.630810499191284 + }, + { + "auxiliary_loss_clip": 0.0120276, + "auxiliary_loss_mlp": 0.0102803, + "balance_loss_clip": 1.05644596, + "balance_loss_mlp": 1.02041292, + "epoch": 0.5568448265496303, + "flos": 20704189858560.0, + "grad_norm": 2.630817886206473, + "language_loss": 0.79846507, + "learning_rate": 1.729747130148611e-06, + "loss": 0.82077301, + "num_input_tokens_seen": 99918525, + "step": 4631, + "time_per_iteration": 2.667227268218994 + }, + { + "auxiliary_loss_clip": 0.01214337, + "auxiliary_loss_mlp": 0.01028077, + "balance_loss_clip": 1.05028892, + "balance_loss_mlp": 1.01956522, + "epoch": 0.5569650694402694, + "flos": 25302910256640.0, + "grad_norm": 1.9327408196813807, + "language_loss": 0.77113485, + "learning_rate": 1.7289753234639208e-06, + "loss": 0.79355896, + "num_input_tokens_seen": 99937500, + "step": 4632, + "time_per_iteration": 2.7196621894836426 + }, + { + "auxiliary_loss_clip": 0.01195663, + "auxiliary_loss_mlp": 0.0102298, + "balance_loss_clip": 1.05225158, + "balance_loss_mlp": 1.01517749, + "epoch": 0.5570853123309084, + "flos": 19712623939200.0, + "grad_norm": 1.7888245923543538, + "language_loss": 0.76397938, + "learning_rate": 1.7282035578934592e-06, + "loss": 0.78616577, + "num_input_tokens_seen": 99955665, + "step": 4633, + "time_per_iteration": 2.6732187271118164 + }, + { + "auxiliary_loss_clip": 0.01195757, + "auxiliary_loss_mlp": 0.01029055, + "balance_loss_clip": 1.05524933, + "balance_loss_mlp": 1.02143788, + "epoch": 0.5572055552215476, + "flos": 16108091153280.0, + "grad_norm": 1.8428726273894507, + "language_loss": 0.79143739, + "learning_rate": 1.727431833554301e-06, + "loss": 0.81368548, + "num_input_tokens_seen": 99974140, + "step": 4634, + "time_per_iteration": 2.668886423110962 + }, + { + "auxiliary_loss_clip": 0.01230011, + "auxiliary_loss_mlp": 0.01028138, + "balance_loss_clip": 1.04562926, + "balance_loss_mlp": 1.02072656, + "epoch": 0.5573257981121866, + "flos": 17128815937920.0, + "grad_norm": 1.7647429568851274, + "language_loss": 0.77118933, + "learning_rate": 1.7266601505635175e-06, + "loss": 0.79377079, + "num_input_tokens_seen": 99991480, + "step": 4635, + "time_per_iteration": 2.773725748062134 + }, + { + "auxiliary_loss_clip": 0.01189208, + "auxiliary_loss_mlp": 0.01028251, + "balance_loss_clip": 1.05213714, + "balance_loss_mlp": 1.02007031, + "epoch": 0.5574460410028257, + "flos": 18807029222400.0, + "grad_norm": 1.889147706087297, + "language_loss": 0.75475854, + "learning_rate": 1.7258885090381717e-06, + "loss": 0.77693307, + "num_input_tokens_seen": 100009520, + "step": 4636, + "time_per_iteration": 2.6123993396759033 + }, + { + "auxiliary_loss_clip": 0.01198116, + "auxiliary_loss_mlp": 0.01023724, + "balance_loss_clip": 1.04849434, + "balance_loss_mlp": 1.01639843, + "epoch": 0.5575662838934649, + "flos": 29642678530560.0, + "grad_norm": 1.803533209581248, + "language_loss": 0.78515613, + "learning_rate": 1.7251169090953213e-06, + "loss": 0.80737448, + "num_input_tokens_seen": 100029995, + "step": 4637, + "time_per_iteration": 2.7456324100494385 + }, + { + "auxiliary_loss_clip": 0.01185329, + "auxiliary_loss_mlp": 0.01030664, + "balance_loss_clip": 1.04974091, + "balance_loss_mlp": 1.02274847, + "epoch": 0.5576865267841039, + "flos": 22054466949120.0, + "grad_norm": 2.3077793597962866, + "language_loss": 0.76017261, + "learning_rate": 1.7243453508520168e-06, + "loss": 0.78233254, + "num_input_tokens_seen": 100046980, + "step": 4638, + "time_per_iteration": 2.7117817401885986 + }, + { + "auxiliary_loss_clip": 0.01196172, + "auxiliary_loss_mlp": 0.0102856, + "balance_loss_clip": 1.04781294, + "balance_loss_mlp": 1.01950049, + "epoch": 0.557806769674743, + "flos": 17196040241280.0, + "grad_norm": 2.2608089403753047, + "language_loss": 0.84429049, + "learning_rate": 1.7235738344253038e-06, + "loss": 0.86653775, + "num_input_tokens_seen": 100060610, + "step": 4639, + "time_per_iteration": 4.665926456451416 + }, + { + "auxiliary_loss_clip": 0.0118594, + "auxiliary_loss_mlp": 0.01026354, + "balance_loss_clip": 1.05147672, + "balance_loss_mlp": 1.01808643, + "epoch": 0.557927012565382, + "flos": 24712717887360.0, + "grad_norm": 1.8751373251878405, + "language_loss": 0.82248175, + "learning_rate": 1.72280235993222e-06, + "loss": 0.84460467, + "num_input_tokens_seen": 100078915, + "step": 4640, + "time_per_iteration": 3.551297903060913 + }, + { + "auxiliary_loss_clip": 0.01187951, + "auxiliary_loss_mlp": 0.01379352, + "balance_loss_clip": 1.0524385, + "balance_loss_mlp": 1.00024462, + "epoch": 0.5580472554560212, + "flos": 16983090460800.0, + "grad_norm": 2.5093536908747103, + "language_loss": 0.69291514, + "learning_rate": 1.722030927489798e-06, + "loss": 0.71858823, + "num_input_tokens_seen": 100096195, + "step": 4641, + "time_per_iteration": 2.630371332168579 + }, + { + "auxiliary_loss_clip": 0.01205457, + "auxiliary_loss_mlp": 0.0102862, + "balance_loss_clip": 1.04876852, + "balance_loss_mlp": 1.02075565, + "epoch": 0.5581674983466602, + "flos": 23509100027520.0, + "grad_norm": 1.6801188306725556, + "language_loss": 0.7369858, + "learning_rate": 1.7212595372150634e-06, + "loss": 0.75932658, + "num_input_tokens_seen": 100116175, + "step": 4642, + "time_per_iteration": 3.7199108600616455 + }, + { + "auxiliary_loss_clip": 0.0117883, + "auxiliary_loss_mlp": 0.01023371, + "balance_loss_clip": 1.05315042, + "balance_loss_mlp": 1.01585519, + "epoch": 0.5582877412372993, + "flos": 13480291969920.0, + "grad_norm": 2.126063405886545, + "language_loss": 0.7258693, + "learning_rate": 1.720488189225035e-06, + "loss": 0.74789131, + "num_input_tokens_seen": 100133875, + "step": 4643, + "time_per_iteration": 2.5817832946777344 + }, + { + "auxiliary_loss_clip": 0.01192765, + "auxiliary_loss_mlp": 0.01026729, + "balance_loss_clip": 1.05262852, + "balance_loss_mlp": 1.01884329, + "epoch": 0.5584079841279385, + "flos": 21903605827200.0, + "grad_norm": 2.5664470734594205, + "language_loss": 0.79236299, + "learning_rate": 1.7197168836367265e-06, + "loss": 0.81455791, + "num_input_tokens_seen": 100150685, + "step": 4644, + "time_per_iteration": 2.665842294692993 + }, + { + "auxiliary_loss_clip": 0.01186485, + "auxiliary_loss_mlp": 0.01378966, + "balance_loss_clip": 1.05043364, + "balance_loss_mlp": 1.00020742, + "epoch": 0.5585282270185775, + "flos": 18843550375680.0, + "grad_norm": 1.9557441708403933, + "language_loss": 0.81940436, + "learning_rate": 1.7189456205671433e-06, + "loss": 0.84505886, + "num_input_tokens_seen": 100169530, + "step": 4645, + "time_per_iteration": 2.600283622741699 + }, + { + "auxiliary_loss_clip": 0.01196077, + "auxiliary_loss_mlp": 0.01030013, + "balance_loss_clip": 1.05313063, + "balance_loss_mlp": 1.02137041, + "epoch": 0.5586484699092166, + "flos": 21868449390720.0, + "grad_norm": 1.858024603957946, + "language_loss": 0.82463807, + "learning_rate": 1.7181744001332866e-06, + "loss": 0.84689897, + "num_input_tokens_seen": 100188140, + "step": 4646, + "time_per_iteration": 2.661710023880005 + }, + { + "auxiliary_loss_clip": 0.01180797, + "auxiliary_loss_mlp": 0.01024439, + "balance_loss_clip": 1.05450821, + "balance_loss_mlp": 1.01713181, + "epoch": 0.5587687127998557, + "flos": 22893232412160.0, + "grad_norm": 2.066126958370382, + "language_loss": 0.63190091, + "learning_rate": 1.7174032224521493e-06, + "loss": 0.65395325, + "num_input_tokens_seen": 100206850, + "step": 4647, + "time_per_iteration": 2.572096824645996 + }, + { + "auxiliary_loss_clip": 0.01187959, + "auxiliary_loss_mlp": 0.01025103, + "balance_loss_clip": 1.05202174, + "balance_loss_mlp": 1.01789665, + "epoch": 0.5588889556904948, + "flos": 20303067703680.0, + "grad_norm": 1.6771150089708393, + "language_loss": 0.69585657, + "learning_rate": 1.7166320876407184e-06, + "loss": 0.71798718, + "num_input_tokens_seen": 100226270, + "step": 4648, + "time_per_iteration": 2.6809322834014893 + }, + { + "auxiliary_loss_clip": 0.01182439, + "auxiliary_loss_mlp": 0.0137914, + "balance_loss_clip": 1.05448437, + "balance_loss_mlp": 1.00027132, + "epoch": 0.5590091985811338, + "flos": 16472153450880.0, + "grad_norm": 2.738640178610574, + "language_loss": 0.67544401, + "learning_rate": 1.7158609958159742e-06, + "loss": 0.70105976, + "num_input_tokens_seen": 100243675, + "step": 4649, + "time_per_iteration": 2.6236321926116943 + }, + { + "auxiliary_loss_clip": 0.01229864, + "auxiliary_loss_mlp": 0.01035815, + "balance_loss_clip": 1.04521787, + "balance_loss_mlp": 1.02746403, + "epoch": 0.559129441471773, + "flos": 14532186781440.0, + "grad_norm": 2.5660398242332856, + "language_loss": 0.78419644, + "learning_rate": 1.7150899470948911e-06, + "loss": 0.80685318, + "num_input_tokens_seen": 100258940, + "step": 4650, + "time_per_iteration": 2.7200827598571777 + }, + { + "auxiliary_loss_clip": 0.01094407, + "auxiliary_loss_mlp": 0.0100458, + "balance_loss_clip": 1.01679027, + "balance_loss_mlp": 1.00338793, + "epoch": 0.5592496843624121, + "flos": 60521009852160.0, + "grad_norm": 0.8037208332271406, + "language_loss": 0.56586361, + "learning_rate": 1.7143189415944365e-06, + "loss": 0.5868535, + "num_input_tokens_seen": 100323400, + "step": 4651, + "time_per_iteration": 3.2533490657806396 + }, + { + "auxiliary_loss_clip": 0.01184625, + "auxiliary_loss_mlp": 0.01031853, + "balance_loss_clip": 1.05058575, + "balance_loss_mlp": 1.02423906, + "epoch": 0.5593699272530511, + "flos": 20886256920960.0, + "grad_norm": 1.9691191236011423, + "language_loss": 0.76321828, + "learning_rate": 1.7135479794315714e-06, + "loss": 0.78538311, + "num_input_tokens_seen": 100340355, + "step": 4652, + "time_per_iteration": 2.603571653366089 + }, + { + "auxiliary_loss_clip": 0.01205164, + "auxiliary_loss_mlp": 0.01025178, + "balance_loss_clip": 1.05006337, + "balance_loss_mlp": 1.01763153, + "epoch": 0.5594901701436903, + "flos": 12896743616640.0, + "grad_norm": 2.1304961018725077, + "language_loss": 0.78891504, + "learning_rate": 1.7127770607232502e-06, + "loss": 0.8112185, + "num_input_tokens_seen": 100358900, + "step": 4653, + "time_per_iteration": 2.6754844188690186 + }, + { + "auxiliary_loss_clip": 0.01210309, + "auxiliary_loss_mlp": 0.01030311, + "balance_loss_clip": 1.0473963, + "balance_loss_mlp": 1.02294385, + "epoch": 0.5596104130343293, + "flos": 23112107936640.0, + "grad_norm": 1.8846642607313773, + "language_loss": 0.79830921, + "learning_rate": 1.7120061855864204e-06, + "loss": 0.82071537, + "num_input_tokens_seen": 100378910, + "step": 4654, + "time_per_iteration": 2.762981653213501 + }, + { + "auxiliary_loss_clip": 0.01187967, + "auxiliary_loss_mlp": 0.0102956, + "balance_loss_clip": 1.05324674, + "balance_loss_mlp": 1.02158499, + "epoch": 0.5597306559249684, + "flos": 25957812977280.0, + "grad_norm": 2.2802600130979283, + "language_loss": 0.71071219, + "learning_rate": 1.7112353541380233e-06, + "loss": 0.73288751, + "num_input_tokens_seen": 100398770, + "step": 4655, + "time_per_iteration": 2.7155048847198486 + }, + { + "auxiliary_loss_clip": 0.01196228, + "auxiliary_loss_mlp": 0.01027098, + "balance_loss_clip": 1.05108929, + "balance_loss_mlp": 1.0191524, + "epoch": 0.5598508988156076, + "flos": 22492289825280.0, + "grad_norm": 1.5847525902624378, + "language_loss": 0.72201216, + "learning_rate": 1.7104645664949931e-06, + "loss": 0.74424541, + "num_input_tokens_seen": 100421240, + "step": 4656, + "time_per_iteration": 2.6980180740356445 + }, + { + "auxiliary_loss_clip": 0.01196681, + "auxiliary_loss_mlp": 0.0102796, + "balance_loss_clip": 1.04730701, + "balance_loss_mlp": 1.0196507, + "epoch": 0.5599711417062466, + "flos": 23112538899840.0, + "grad_norm": 2.0892021714691613, + "language_loss": 0.71435004, + "learning_rate": 1.7096938227742584e-06, + "loss": 0.73659647, + "num_input_tokens_seen": 100442370, + "step": 4657, + "time_per_iteration": 2.855888843536377 + }, + { + "auxiliary_loss_clip": 0.01178735, + "auxiliary_loss_mlp": 0.01022254, + "balance_loss_clip": 1.05204916, + "balance_loss_mlp": 1.01460719, + "epoch": 0.5600913845968857, + "flos": 22339345714560.0, + "grad_norm": 1.7811795042791634, + "language_loss": 0.84074074, + "learning_rate": 1.70892312309274e-06, + "loss": 0.86275065, + "num_input_tokens_seen": 100460260, + "step": 4658, + "time_per_iteration": 2.592379570007324 + }, + { + "auxiliary_loss_clip": 0.01192769, + "auxiliary_loss_mlp": 0.01023955, + "balance_loss_clip": 1.04383183, + "balance_loss_mlp": 1.0166235, + "epoch": 0.5602116274875248, + "flos": 17633791290240.0, + "grad_norm": 1.9863480619466438, + "language_loss": 0.68547082, + "learning_rate": 1.7081524675673523e-06, + "loss": 0.70763803, + "num_input_tokens_seen": 100475750, + "step": 4659, + "time_per_iteration": 2.6514077186584473 + }, + { + "auxiliary_loss_clip": 0.01097427, + "auxiliary_loss_mlp": 0.01002743, + "balance_loss_clip": 1.01631808, + "balance_loss_mlp": 1.00152159, + "epoch": 0.5603318703781639, + "flos": 70115945529600.0, + "grad_norm": 0.7741477303482105, + "language_loss": 0.5963608, + "learning_rate": 1.7073818563150026e-06, + "loss": 0.6173625, + "num_input_tokens_seen": 100537830, + "step": 4660, + "time_per_iteration": 3.296078681945801 + }, + { + "auxiliary_loss_clip": 0.01181763, + "auxiliary_loss_mlp": 0.01024848, + "balance_loss_clip": 1.04793596, + "balance_loss_mlp": 1.01725423, + "epoch": 0.560452113268803, + "flos": 18545850455040.0, + "grad_norm": 3.1466856666491987, + "language_loss": 0.86497056, + "learning_rate": 1.7066112894525935e-06, + "loss": 0.88703668, + "num_input_tokens_seen": 100555910, + "step": 4661, + "time_per_iteration": 2.6337428092956543 + }, + { + "auxiliary_loss_clip": 0.01185299, + "auxiliary_loss_mlp": 0.01029678, + "balance_loss_clip": 1.04748285, + "balance_loss_mlp": 1.02164865, + "epoch": 0.5605723561594421, + "flos": 25264665250560.0, + "grad_norm": 1.9691069403126635, + "language_loss": 0.72686219, + "learning_rate": 1.7058407670970177e-06, + "loss": 0.74901193, + "num_input_tokens_seen": 100577385, + "step": 4662, + "time_per_iteration": 2.7117674350738525 + }, + { + "auxiliary_loss_clip": 0.01193488, + "auxiliary_loss_mlp": 0.01025609, + "balance_loss_clip": 1.05116153, + "balance_loss_mlp": 1.01828408, + "epoch": 0.5606925990500812, + "flos": 20594949621120.0, + "grad_norm": 1.8033277433045256, + "language_loss": 0.6128071, + "learning_rate": 1.7050702893651643e-06, + "loss": 0.63499808, + "num_input_tokens_seen": 100596965, + "step": 4663, + "time_per_iteration": 2.626025915145874 + }, + { + "auxiliary_loss_clip": 0.01187662, + "auxiliary_loss_mlp": 0.01026278, + "balance_loss_clip": 1.05184734, + "balance_loss_mlp": 1.01873779, + "epoch": 0.5608128419407202, + "flos": 35006044677120.0, + "grad_norm": 2.4261432132070873, + "language_loss": 0.75642133, + "learning_rate": 1.7042998563739134e-06, + "loss": 0.7785607, + "num_input_tokens_seen": 100615315, + "step": 4664, + "time_per_iteration": 2.7722949981689453 + }, + { + "auxiliary_loss_clip": 0.01202663, + "auxiliary_loss_mlp": 0.01030001, + "balance_loss_clip": 1.04818833, + "balance_loss_mlp": 1.02163887, + "epoch": 0.5609330848313594, + "flos": 24639819235200.0, + "grad_norm": 2.593551326020325, + "language_loss": 0.71786857, + "learning_rate": 1.703529468240139e-06, + "loss": 0.74019516, + "num_input_tokens_seen": 100634185, + "step": 4665, + "time_per_iteration": 3.5799922943115234 + }, + { + "auxiliary_loss_clip": 0.01189295, + "auxiliary_loss_mlp": 0.01026807, + "balance_loss_clip": 1.04811502, + "balance_loss_mlp": 1.01886749, + "epoch": 0.5610533277219985, + "flos": 18762894385920.0, + "grad_norm": 2.242952949744467, + "language_loss": 0.73397911, + "learning_rate": 1.7027591250807088e-06, + "loss": 0.75614017, + "num_input_tokens_seen": 100651360, + "step": 4666, + "time_per_iteration": 3.588545799255371 + }, + { + "auxiliary_loss_clip": 0.01180396, + "auxiliary_loss_mlp": 0.01027094, + "balance_loss_clip": 1.05335319, + "balance_loss_mlp": 1.01917839, + "epoch": 0.5611735706126375, + "flos": 15012384727680.0, + "grad_norm": 2.4975285250507184, + "language_loss": 0.84582877, + "learning_rate": 1.7019888270124825e-06, + "loss": 0.86790371, + "num_input_tokens_seen": 100668525, + "step": 4667, + "time_per_iteration": 2.605821371078491 + }, + { + "auxiliary_loss_clip": 0.01189822, + "auxiliary_loss_mlp": 0.01037965, + "balance_loss_clip": 1.05203032, + "balance_loss_mlp": 1.03036869, + "epoch": 0.5612938135032767, + "flos": 16468167041280.0, + "grad_norm": 2.035289586716538, + "language_loss": 0.81896031, + "learning_rate": 1.7012185741523147e-06, + "loss": 0.84123814, + "num_input_tokens_seen": 100684850, + "step": 4668, + "time_per_iteration": 3.585791826248169 + }, + { + "auxiliary_loss_clip": 0.01178398, + "auxiliary_loss_mlp": 0.01031653, + "balance_loss_clip": 1.05364633, + "balance_loss_mlp": 1.02381539, + "epoch": 0.5614140563939157, + "flos": 25666433850240.0, + "grad_norm": 2.0649709770612934, + "language_loss": 0.62158233, + "learning_rate": 1.7004483666170514e-06, + "loss": 0.64368284, + "num_input_tokens_seen": 100705345, + "step": 4669, + "time_per_iteration": 2.6376729011535645 + }, + { + "auxiliary_loss_clip": 0.01185362, + "auxiliary_loss_mlp": 0.01029626, + "balance_loss_clip": 1.05088663, + "balance_loss_mlp": 1.02185917, + "epoch": 0.5615342992845548, + "flos": 24717566223360.0, + "grad_norm": 1.8079998379933702, + "language_loss": 0.79796034, + "learning_rate": 1.699678204523533e-06, + "loss": 0.8201102, + "num_input_tokens_seen": 100725210, + "step": 4670, + "time_per_iteration": 2.5908474922180176 + }, + { + "auxiliary_loss_clip": 0.01199949, + "auxiliary_loss_mlp": 0.01028516, + "balance_loss_clip": 1.05265808, + "balance_loss_mlp": 1.0209316, + "epoch": 0.5616545421751938, + "flos": 22015934634240.0, + "grad_norm": 2.7031658696373624, + "language_loss": 0.68310541, + "learning_rate": 1.6989080879885918e-06, + "loss": 0.7053901, + "num_input_tokens_seen": 100743070, + "step": 4671, + "time_per_iteration": 2.681229829788208 + }, + { + "auxiliary_loss_clip": 0.01112248, + "auxiliary_loss_mlp": 0.01006824, + "balance_loss_clip": 1.01902556, + "balance_loss_mlp": 1.00562048, + "epoch": 0.561774785065833, + "flos": 53760358690560.0, + "grad_norm": 0.9139652638089119, + "language_loss": 0.61051512, + "learning_rate": 1.6981380171290544e-06, + "loss": 0.63170588, + "num_input_tokens_seen": 100804095, + "step": 4672, + "time_per_iteration": 3.239626407623291 + }, + { + "auxiliary_loss_clip": 0.01187235, + "auxiliary_loss_mlp": 0.01025247, + "balance_loss_clip": 1.04573143, + "balance_loss_mlp": 1.01782632, + "epoch": 0.5618950279564721, + "flos": 19750007018880.0, + "grad_norm": 1.9563742510157367, + "language_loss": 0.7448194, + "learning_rate": 1.6973679920617396e-06, + "loss": 0.76694417, + "num_input_tokens_seen": 100821630, + "step": 4673, + "time_per_iteration": 2.6868226528167725 + }, + { + "auxiliary_loss_clip": 0.01194022, + "auxiliary_loss_mlp": 0.01022534, + "balance_loss_clip": 1.05007899, + "balance_loss_mlp": 1.01501143, + "epoch": 0.5620152708471111, + "flos": 16800592435200.0, + "grad_norm": 2.0615607517455143, + "language_loss": 0.84965062, + "learning_rate": 1.6965980129034603e-06, + "loss": 0.87181616, + "num_input_tokens_seen": 100839015, + "step": 4674, + "time_per_iteration": 2.6463096141815186 + }, + { + "auxiliary_loss_clip": 0.01197937, + "auxiliary_loss_mlp": 0.01030326, + "balance_loss_clip": 1.05170155, + "balance_loss_mlp": 1.02280951, + "epoch": 0.5621355137377503, + "flos": 26797799502720.0, + "grad_norm": 1.4929352524283488, + "language_loss": 0.76491058, + "learning_rate": 1.6958280797710209e-06, + "loss": 0.78719318, + "num_input_tokens_seen": 100860940, + "step": 4675, + "time_per_iteration": 2.7184700965881348 + }, + { + "auxiliary_loss_clip": 0.01097099, + "auxiliary_loss_mlp": 0.01001661, + "balance_loss_clip": 1.0162828, + "balance_loss_mlp": 1.00048661, + "epoch": 0.5622557566283893, + "flos": 61207046686080.0, + "grad_norm": 0.7216748039201613, + "language_loss": 0.547328, + "learning_rate": 1.6950581927812198e-06, + "loss": 0.56831563, + "num_input_tokens_seen": 100920510, + "step": 4676, + "time_per_iteration": 3.1256117820739746 + }, + { + "auxiliary_loss_clip": 0.01188538, + "auxiliary_loss_mlp": 0.01021049, + "balance_loss_clip": 1.0513463, + "balance_loss_mlp": 1.01362491, + "epoch": 0.5623759995190284, + "flos": 26468534505600.0, + "grad_norm": 2.0965648208689975, + "language_loss": 0.79391789, + "learning_rate": 1.6942883520508486e-06, + "loss": 0.81601381, + "num_input_tokens_seen": 100939245, + "step": 4677, + "time_per_iteration": 2.736186981201172 + }, + { + "auxiliary_loss_clip": 0.01188054, + "auxiliary_loss_mlp": 0.01022081, + "balance_loss_clip": 1.05131602, + "balance_loss_mlp": 1.01495576, + "epoch": 0.5624962424096676, + "flos": 19390900798080.0, + "grad_norm": 2.66952114866511, + "language_loss": 0.77082872, + "learning_rate": 1.693518557696691e-06, + "loss": 0.79293001, + "num_input_tokens_seen": 100958385, + "step": 4678, + "time_per_iteration": 2.64693021774292 + }, + { + "auxiliary_loss_clip": 0.011813, + "auxiliary_loss_mlp": 0.01023051, + "balance_loss_clip": 1.0476675, + "balance_loss_mlp": 1.01530218, + "epoch": 0.5626164853003066, + "flos": 20667345482880.0, + "grad_norm": 2.0522632041269557, + "language_loss": 0.89212102, + "learning_rate": 1.6927488098355252e-06, + "loss": 0.91416448, + "num_input_tokens_seen": 100976015, + "step": 4679, + "time_per_iteration": 2.596606969833374 + }, + { + "auxiliary_loss_clip": 0.01125859, + "auxiliary_loss_mlp": 0.01002273, + "balance_loss_clip": 1.01500154, + "balance_loss_mlp": 1.00096726, + "epoch": 0.5627367281909457, + "flos": 62766071665920.0, + "grad_norm": 0.9154567475766638, + "language_loss": 0.63182545, + "learning_rate": 1.6919791085841201e-06, + "loss": 0.65310669, + "num_input_tokens_seen": 101033425, + "step": 4680, + "time_per_iteration": 3.2400002479553223 + }, + { + "auxiliary_loss_clip": 0.0117826, + "auxiliary_loss_mlp": 0.01031153, + "balance_loss_clip": 1.04656434, + "balance_loss_mlp": 1.02319002, + "epoch": 0.5628569710815848, + "flos": 12787144243200.0, + "grad_norm": 2.277281710030721, + "language_loss": 0.78757888, + "learning_rate": 1.6912094540592396e-06, + "loss": 0.80967295, + "num_input_tokens_seen": 101048945, + "step": 4681, + "time_per_iteration": 2.551325798034668 + }, + { + "auxiliary_loss_clip": 0.01183963, + "auxiliary_loss_mlp": 0.01026819, + "balance_loss_clip": 1.04994273, + "balance_loss_mlp": 1.01954687, + "epoch": 0.5629772139722239, + "flos": 13762082165760.0, + "grad_norm": 2.6451415820808637, + "language_loss": 0.80931532, + "learning_rate": 1.6904398463776393e-06, + "loss": 0.83142316, + "num_input_tokens_seen": 101062745, + "step": 4682, + "time_per_iteration": 2.6265225410461426 + }, + { + "auxiliary_loss_clip": 0.01187639, + "auxiliary_loss_mlp": 0.01023003, + "balance_loss_clip": 1.04920578, + "balance_loss_mlp": 1.01605296, + "epoch": 0.5630974568628629, + "flos": 21467830026240.0, + "grad_norm": 1.714165604103692, + "language_loss": 0.72558618, + "learning_rate": 1.6896702856560683e-06, + "loss": 0.74769264, + "num_input_tokens_seen": 101081840, + "step": 4683, + "time_per_iteration": 2.621053457260132 + }, + { + "auxiliary_loss_clip": 0.01197805, + "auxiliary_loss_mlp": 0.0102755, + "balance_loss_clip": 1.04282475, + "balance_loss_mlp": 1.01960468, + "epoch": 0.5632176997535021, + "flos": 14245907385600.0, + "grad_norm": 2.9775259633727402, + "language_loss": 0.69274217, + "learning_rate": 1.6889007720112677e-06, + "loss": 0.71499574, + "num_input_tokens_seen": 101099585, + "step": 4684, + "time_per_iteration": 2.667989492416382 + }, + { + "auxiliary_loss_clip": 0.01188034, + "auxiliary_loss_mlp": 0.01026977, + "balance_loss_clip": 1.05073178, + "balance_loss_mlp": 1.01969075, + "epoch": 0.5633379426441412, + "flos": 20812244947200.0, + "grad_norm": 1.7247116254297548, + "language_loss": 0.77114666, + "learning_rate": 1.6881313055599734e-06, + "loss": 0.79329675, + "num_input_tokens_seen": 101119515, + "step": 4685, + "time_per_iteration": 2.636867046356201 + }, + { + "auxiliary_loss_clip": 0.0117652, + "auxiliary_loss_mlp": 0.01026547, + "balance_loss_clip": 1.04307997, + "balance_loss_mlp": 1.01838136, + "epoch": 0.5634581855347802, + "flos": 22600883617920.0, + "grad_norm": 2.3160626820558243, + "language_loss": 0.82221228, + "learning_rate": 1.6873618864189117e-06, + "loss": 0.84424293, + "num_input_tokens_seen": 101135285, + "step": 4686, + "time_per_iteration": 2.63897705078125 + }, + { + "auxiliary_loss_clip": 0.01186284, + "auxiliary_loss_mlp": 0.01026471, + "balance_loss_clip": 1.05031466, + "balance_loss_mlp": 1.01918697, + "epoch": 0.5635784284254194, + "flos": 21506972872320.0, + "grad_norm": 9.32020089765182, + "language_loss": 0.78020597, + "learning_rate": 1.686592514704803e-06, + "loss": 0.80233347, + "num_input_tokens_seen": 101152680, + "step": 4687, + "time_per_iteration": 2.630263328552246 + }, + { + "auxiliary_loss_clip": 0.01193433, + "auxiliary_loss_mlp": 0.01022552, + "balance_loss_clip": 1.05022669, + "balance_loss_mlp": 1.01532817, + "epoch": 0.5636986713160584, + "flos": 19827466698240.0, + "grad_norm": 2.2803217012834716, + "language_loss": 0.70250565, + "learning_rate": 1.685823190534361e-06, + "loss": 0.72466552, + "num_input_tokens_seen": 101170920, + "step": 4688, + "time_per_iteration": 2.6373345851898193 + }, + { + "auxiliary_loss_clip": 0.01180372, + "auxiliary_loss_mlp": 0.01024154, + "balance_loss_clip": 1.05170751, + "balance_loss_mlp": 1.01611328, + "epoch": 0.5638189142066975, + "flos": 19792453916160.0, + "grad_norm": 2.047998982651886, + "language_loss": 0.83509701, + "learning_rate": 1.6850539140242907e-06, + "loss": 0.85714227, + "num_input_tokens_seen": 101190180, + "step": 4689, + "time_per_iteration": 2.611609935760498 + }, + { + "auxiliary_loss_clip": 0.01190303, + "auxiliary_loss_mlp": 0.01027673, + "balance_loss_clip": 1.05084121, + "balance_loss_mlp": 1.02015102, + "epoch": 0.5639391570973367, + "flos": 22893771116160.0, + "grad_norm": 1.7728269967969363, + "language_loss": 0.82214105, + "learning_rate": 1.684284685291292e-06, + "loss": 0.84432089, + "num_input_tokens_seen": 101211825, + "step": 4690, + "time_per_iteration": 2.6738977432250977 + }, + { + "auxiliary_loss_clip": 0.01178146, + "auxiliary_loss_mlp": 0.01025694, + "balance_loss_clip": 1.05194497, + "balance_loss_mlp": 1.01808894, + "epoch": 0.5640593999879757, + "flos": 23727077712000.0, + "grad_norm": 1.9582331982498058, + "language_loss": 0.81162798, + "learning_rate": 1.683515504452055e-06, + "loss": 0.83366632, + "num_input_tokens_seen": 101229200, + "step": 4691, + "time_per_iteration": 4.450700044631958 + }, + { + "auxiliary_loss_clip": 0.01195686, + "auxiliary_loss_mlp": 0.01027808, + "balance_loss_clip": 1.0450927, + "balance_loss_mlp": 1.02003551, + "epoch": 0.5641796428786148, + "flos": 22710123855360.0, + "grad_norm": 1.763151651690187, + "language_loss": 0.66606534, + "learning_rate": 1.6827463716232648e-06, + "loss": 0.68830025, + "num_input_tokens_seen": 101249860, + "step": 4692, + "time_per_iteration": 3.6335630416870117 + }, + { + "auxiliary_loss_clip": 0.01185179, + "auxiliary_loss_mlp": 0.01378521, + "balance_loss_clip": 1.05097902, + "balance_loss_mlp": 1.00016749, + "epoch": 0.5642998857692539, + "flos": 19791987039360.0, + "grad_norm": 1.7496111374665757, + "language_loss": 0.75867361, + "learning_rate": 1.6819772869215972e-06, + "loss": 0.78431058, + "num_input_tokens_seen": 101268940, + "step": 4693, + "time_per_iteration": 2.7004873752593994 + }, + { + "auxiliary_loss_clip": 0.01201111, + "auxiliary_loss_mlp": 0.01022496, + "balance_loss_clip": 1.04951024, + "balance_loss_mlp": 1.01508999, + "epoch": 0.564420128659893, + "flos": 23185904428800.0, + "grad_norm": 1.75554306911985, + "language_loss": 0.82399809, + "learning_rate": 1.6812082504637228e-06, + "loss": 0.84623408, + "num_input_tokens_seen": 101290260, + "step": 4694, + "time_per_iteration": 3.6020593643188477 + }, + { + "auxiliary_loss_clip": 0.01182881, + "auxiliary_loss_mlp": 0.01023767, + "balance_loss_clip": 1.05085015, + "balance_loss_mlp": 1.01638818, + "epoch": 0.564540371550532, + "flos": 23258264376960.0, + "grad_norm": 1.4592352244844795, + "language_loss": 0.74429524, + "learning_rate": 1.6804392623663025e-06, + "loss": 0.76636171, + "num_input_tokens_seen": 101311465, + "step": 4695, + "time_per_iteration": 2.6950035095214844 + }, + { + "auxiliary_loss_clip": 0.01180019, + "auxiliary_loss_mlp": 0.01021534, + "balance_loss_clip": 1.0486927, + "balance_loss_mlp": 1.01428628, + "epoch": 0.5646606144411712, + "flos": 25010058672000.0, + "grad_norm": 1.9562120344744225, + "language_loss": 0.78301603, + "learning_rate": 1.6796703227459935e-06, + "loss": 0.80503154, + "num_input_tokens_seen": 101329420, + "step": 4696, + "time_per_iteration": 2.6516659259796143 + }, + { + "auxiliary_loss_clip": 0.01201276, + "auxiliary_loss_mlp": 0.01028629, + "balance_loss_clip": 1.04182649, + "balance_loss_mlp": 1.02091038, + "epoch": 0.5647808573318103, + "flos": 36539645806080.0, + "grad_norm": 1.7781475821577184, + "language_loss": 0.76240063, + "learning_rate": 1.6789014317194407e-06, + "loss": 0.78469968, + "num_input_tokens_seen": 101350900, + "step": 4697, + "time_per_iteration": 2.916534185409546 + }, + { + "auxiliary_loss_clip": 0.01204291, + "auxiliary_loss_mlp": 0.01025538, + "balance_loss_clip": 1.05152023, + "balance_loss_mlp": 1.01719904, + "epoch": 0.5649011002224493, + "flos": 22528451842560.0, + "grad_norm": 2.1553704535704212, + "language_loss": 0.72772586, + "learning_rate": 1.6781325894032853e-06, + "loss": 0.7500242, + "num_input_tokens_seen": 101369860, + "step": 4698, + "time_per_iteration": 2.6952898502349854 + }, + { + "auxiliary_loss_clip": 0.01188497, + "auxiliary_loss_mlp": 0.0102769, + "balance_loss_clip": 1.04774761, + "balance_loss_mlp": 1.01957798, + "epoch": 0.5650213431130885, + "flos": 18515147304960.0, + "grad_norm": 2.357038664202, + "language_loss": 0.91879141, + "learning_rate": 1.6773637959141608e-06, + "loss": 0.94095325, + "num_input_tokens_seen": 101386835, + "step": 4699, + "time_per_iteration": 2.6569759845733643 + }, + { + "auxiliary_loss_clip": 0.01184303, + "auxiliary_loss_mlp": 0.01025091, + "balance_loss_clip": 1.04695427, + "balance_loss_mlp": 1.01764619, + "epoch": 0.5651415860037275, + "flos": 17526310819200.0, + "grad_norm": 2.3294719639520265, + "language_loss": 0.66483784, + "learning_rate": 1.6765950513686915e-06, + "loss": 0.68693185, + "num_input_tokens_seen": 101404945, + "step": 4700, + "time_per_iteration": 2.6550447940826416 + }, + { + "auxiliary_loss_clip": 0.01210767, + "auxiliary_loss_mlp": 0.01027046, + "balance_loss_clip": 1.04430032, + "balance_loss_mlp": 1.01919019, + "epoch": 0.5652618288943666, + "flos": 25520026014720.0, + "grad_norm": 1.7787068860109008, + "language_loss": 0.76122963, + "learning_rate": 1.675826355883496e-06, + "loss": 0.78360772, + "num_input_tokens_seen": 101424160, + "step": 4701, + "time_per_iteration": 2.7795374393463135 + }, + { + "auxiliary_loss_clip": 0.01193279, + "auxiliary_loss_mlp": 0.01020411, + "balance_loss_clip": 1.05279732, + "balance_loss_mlp": 1.01291823, + "epoch": 0.5653820717850057, + "flos": 19683105937920.0, + "grad_norm": 1.8742344827305066, + "language_loss": 0.78891891, + "learning_rate": 1.6750577095751848e-06, + "loss": 0.81105578, + "num_input_tokens_seen": 101443270, + "step": 4702, + "time_per_iteration": 2.6995296478271484 + }, + { + "auxiliary_loss_clip": 0.01175554, + "auxiliary_loss_mlp": 0.01022995, + "balance_loss_clip": 1.05112171, + "balance_loss_mlp": 1.01561594, + "epoch": 0.5655023146756448, + "flos": 26979722910720.0, + "grad_norm": 1.7154345227349075, + "language_loss": 0.72847933, + "learning_rate": 1.6742891125603605e-06, + "loss": 0.75046486, + "num_input_tokens_seen": 101464175, + "step": 4703, + "time_per_iteration": 2.6859140396118164 + }, + { + "auxiliary_loss_clip": 0.01185205, + "auxiliary_loss_mlp": 0.01026138, + "balance_loss_clip": 1.05112922, + "balance_loss_mlp": 1.0180198, + "epoch": 0.5656225575662839, + "flos": 27669351104640.0, + "grad_norm": 2.0884347446373615, + "language_loss": 0.7246809, + "learning_rate": 1.6735205649556185e-06, + "loss": 0.74679434, + "num_input_tokens_seen": 101484045, + "step": 4704, + "time_per_iteration": 2.7629354000091553 + }, + { + "auxiliary_loss_clip": 0.01207776, + "auxiliary_loss_mlp": 0.01021598, + "balance_loss_clip": 1.04815412, + "balance_loss_mlp": 1.0141356, + "epoch": 0.5657428004569229, + "flos": 24349732997760.0, + "grad_norm": 1.6193473541291175, + "language_loss": 0.84971273, + "learning_rate": 1.6727520668775476e-06, + "loss": 0.87200642, + "num_input_tokens_seen": 101504330, + "step": 4705, + "time_per_iteration": 2.7716777324676514 + }, + { + "auxiliary_loss_clip": 0.01179229, + "auxiliary_loss_mlp": 0.01030664, + "balance_loss_clip": 1.05135596, + "balance_loss_mlp": 1.0226711, + "epoch": 0.5658630433475621, + "flos": 21944041562880.0, + "grad_norm": 1.5625421161843651, + "language_loss": 0.74966884, + "learning_rate": 1.6719836184427275e-06, + "loss": 0.7717678, + "num_input_tokens_seen": 101524635, + "step": 4706, + "time_per_iteration": 2.719562530517578 + }, + { + "auxiliary_loss_clip": 0.0118976, + "auxiliary_loss_mlp": 0.01026265, + "balance_loss_clip": 1.04796529, + "balance_loss_mlp": 1.01893377, + "epoch": 0.5659832862382012, + "flos": 30409012218240.0, + "grad_norm": 1.901362259558202, + "language_loss": 0.64416224, + "learning_rate": 1.671215219767733e-06, + "loss": 0.66632247, + "num_input_tokens_seen": 101544095, + "step": 4707, + "time_per_iteration": 2.7367639541625977 + }, + { + "auxiliary_loss_clip": 0.01212673, + "auxiliary_loss_mlp": 0.01029007, + "balance_loss_clip": 1.04604197, + "balance_loss_mlp": 1.02145529, + "epoch": 0.5661035291288402, + "flos": 13188194570880.0, + "grad_norm": 2.0984459117957033, + "language_loss": 0.76340812, + "learning_rate": 1.670446870969127e-06, + "loss": 0.78582495, + "num_input_tokens_seen": 101561760, + "step": 4708, + "time_per_iteration": 2.7265570163726807 + }, + { + "auxiliary_loss_clip": 0.0119746, + "auxiliary_loss_mlp": 0.01027404, + "balance_loss_clip": 1.04896164, + "balance_loss_mlp": 1.01931548, + "epoch": 0.5662237720194794, + "flos": 16143032108160.0, + "grad_norm": 2.43638896041354, + "language_loss": 0.80078804, + "learning_rate": 1.6696785721634685e-06, + "loss": 0.82303667, + "num_input_tokens_seen": 101576245, + "step": 4709, + "time_per_iteration": 2.601170778274536 + }, + { + "auxiliary_loss_clip": 0.01188033, + "auxiliary_loss_mlp": 0.01025452, + "balance_loss_clip": 1.05144143, + "balance_loss_mlp": 1.01736331, + "epoch": 0.5663440149101184, + "flos": 17676848718720.0, + "grad_norm": 2.3245354461810597, + "language_loss": 0.73939264, + "learning_rate": 1.6689103234673086e-06, + "loss": 0.76152748, + "num_input_tokens_seen": 101594565, + "step": 4710, + "time_per_iteration": 2.6434593200683594 + }, + { + "auxiliary_loss_clip": 0.01196872, + "auxiliary_loss_mlp": 0.01027207, + "balance_loss_clip": 1.05337656, + "balance_loss_mlp": 1.01963758, + "epoch": 0.5664642578007575, + "flos": 23368330627200.0, + "grad_norm": 2.1112809300148525, + "language_loss": 0.77376205, + "learning_rate": 1.668142124997189e-06, + "loss": 0.79600286, + "num_input_tokens_seen": 101614225, + "step": 4711, + "time_per_iteration": 2.6548595428466797 + }, + { + "auxiliary_loss_clip": 0.01094461, + "auxiliary_loss_mlp": 0.01002762, + "balance_loss_clip": 1.02014399, + "balance_loss_mlp": 1.00140333, + "epoch": 0.5665845006913967, + "flos": 65516470945920.0, + "grad_norm": 0.7252111722995853, + "language_loss": 0.59837472, + "learning_rate": 1.6673739768696453e-06, + "loss": 0.61934698, + "num_input_tokens_seen": 101680795, + "step": 4712, + "time_per_iteration": 3.244798183441162 + }, + { + "auxiliary_loss_clip": 0.01200285, + "auxiliary_loss_mlp": 0.01027954, + "balance_loss_clip": 1.04886246, + "balance_loss_mlp": 1.02047336, + "epoch": 0.5667047435820357, + "flos": 26140885620480.0, + "grad_norm": 1.7839976587017945, + "language_loss": 0.77337909, + "learning_rate": 1.6666058792012052e-06, + "loss": 0.79566145, + "num_input_tokens_seen": 101701680, + "step": 4713, + "time_per_iteration": 2.719442367553711 + }, + { + "auxiliary_loss_clip": 0.01088083, + "auxiliary_loss_mlp": 0.01000729, + "balance_loss_clip": 1.01809657, + "balance_loss_mlp": 0.99945372, + "epoch": 0.5668249864726748, + "flos": 71866949725440.0, + "grad_norm": 0.8940712721357841, + "language_loss": 0.68712121, + "learning_rate": 1.6658378321083878e-06, + "loss": 0.7080093, + "num_input_tokens_seen": 101766010, + "step": 4714, + "time_per_iteration": 3.2868449687957764 + }, + { + "auxiliary_loss_clip": 0.0122606, + "auxiliary_loss_mlp": 0.01026001, + "balance_loss_clip": 1.04550612, + "balance_loss_mlp": 1.01886082, + "epoch": 0.5669452293633139, + "flos": 22195667312640.0, + "grad_norm": 1.7367178334085103, + "language_loss": 0.82541895, + "learning_rate": 1.6650698357077055e-06, + "loss": 0.84793955, + "num_input_tokens_seen": 101783055, + "step": 4715, + "time_per_iteration": 2.877772808074951 + }, + { + "auxiliary_loss_clip": 0.0119726, + "auxiliary_loss_mlp": 0.01027078, + "balance_loss_clip": 1.04908633, + "balance_loss_mlp": 1.01971054, + "epoch": 0.567065472253953, + "flos": 18223193560320.0, + "grad_norm": 6.595369046809832, + "language_loss": 0.80956358, + "learning_rate": 1.6643018901156632e-06, + "loss": 0.83180696, + "num_input_tokens_seen": 101802150, + "step": 4716, + "time_per_iteration": 2.624027967453003 + }, + { + "auxiliary_loss_clip": 0.01198483, + "auxiliary_loss_mlp": 0.01028234, + "balance_loss_clip": 1.04736066, + "balance_loss_mlp": 1.02088177, + "epoch": 0.567185715144592, + "flos": 20371548983040.0, + "grad_norm": 3.4240971794734887, + "language_loss": 0.79373258, + "learning_rate": 1.6635339954487566e-06, + "loss": 0.81599975, + "num_input_tokens_seen": 101818025, + "step": 4717, + "time_per_iteration": 4.486262083053589 + }, + { + "auxiliary_loss_clip": 0.01198375, + "auxiliary_loss_mlp": 0.01031069, + "balance_loss_clip": 1.04938269, + "balance_loss_mlp": 1.02328491, + "epoch": 0.5673059580352312, + "flos": 23221348174080.0, + "grad_norm": 1.7930060087393058, + "language_loss": 0.8227213, + "learning_rate": 1.6627661518234765e-06, + "loss": 0.84501576, + "num_input_tokens_seen": 101837280, + "step": 4718, + "time_per_iteration": 3.5464024543762207 + }, + { + "auxiliary_loss_clip": 0.01216586, + "auxiliary_loss_mlp": 0.01030936, + "balance_loss_clip": 1.04810858, + "balance_loss_mlp": 1.0232892, + "epoch": 0.5674262009258703, + "flos": 21719599430400.0, + "grad_norm": 2.119012075138681, + "language_loss": 0.85308123, + "learning_rate": 1.661998359356302e-06, + "loss": 0.87555641, + "num_input_tokens_seen": 101856310, + "step": 4719, + "time_per_iteration": 2.7196109294891357 + }, + { + "auxiliary_loss_clip": 0.01073481, + "auxiliary_loss_mlp": 0.01002598, + "balance_loss_clip": 1.01705027, + "balance_loss_mlp": 1.00129879, + "epoch": 0.5675464438165093, + "flos": 67470369114240.0, + "grad_norm": 0.7529489494027298, + "language_loss": 0.55803692, + "learning_rate": 1.6612306181637077e-06, + "loss": 0.5787977, + "num_input_tokens_seen": 101915635, + "step": 4720, + "time_per_iteration": 4.026615381240845 + }, + { + "auxiliary_loss_clip": 0.01203429, + "auxiliary_loss_mlp": 0.01025819, + "balance_loss_clip": 1.04853404, + "balance_loss_mlp": 1.01808858, + "epoch": 0.5676666867071485, + "flos": 18879173688960.0, + "grad_norm": 2.5575184148354566, + "language_loss": 0.65230542, + "learning_rate": 1.6604629283621598e-06, + "loss": 0.67459792, + "num_input_tokens_seen": 101933565, + "step": 4721, + "time_per_iteration": 2.7277939319610596 + }, + { + "auxiliary_loss_clip": 0.01180487, + "auxiliary_loss_mlp": 0.01026923, + "balance_loss_clip": 1.05197465, + "balance_loss_mlp": 1.01894784, + "epoch": 0.5677869295977875, + "flos": 33546778744320.0, + "grad_norm": 1.7029334385735158, + "language_loss": 0.74301267, + "learning_rate": 1.6596952900681152e-06, + "loss": 0.76508677, + "num_input_tokens_seen": 101954325, + "step": 4722, + "time_per_iteration": 2.7801811695098877 + }, + { + "auxiliary_loss_clip": 0.01201538, + "auxiliary_loss_mlp": 0.01027128, + "balance_loss_clip": 1.0477159, + "balance_loss_mlp": 1.01825261, + "epoch": 0.5679071724884266, + "flos": 28037256157440.0, + "grad_norm": 1.9866323644826513, + "language_loss": 0.82014555, + "learning_rate": 1.658927703398025e-06, + "loss": 0.8424322, + "num_input_tokens_seen": 101974390, + "step": 4723, + "time_per_iteration": 2.8509647846221924 + }, + { + "auxiliary_loss_clip": 0.01208629, + "auxiliary_loss_mlp": 0.01027739, + "balance_loss_clip": 1.04224372, + "balance_loss_mlp": 1.02048492, + "epoch": 0.5680274153790658, + "flos": 23550110380800.0, + "grad_norm": 12.996413606737951, + "language_loss": 0.78076148, + "learning_rate": 1.6581601684683309e-06, + "loss": 0.80312514, + "num_input_tokens_seen": 101994815, + "step": 4724, + "time_per_iteration": 2.767723321914673 + }, + { + "auxiliary_loss_clip": 0.01189885, + "auxiliary_loss_mlp": 0.01025856, + "balance_loss_clip": 1.05229187, + "balance_loss_mlp": 1.01776147, + "epoch": 0.5681476582697048, + "flos": 22455158140800.0, + "grad_norm": 2.70530101162823, + "language_loss": 0.68625128, + "learning_rate": 1.6573926853954674e-06, + "loss": 0.70840871, + "num_input_tokens_seen": 102012400, + "step": 4725, + "time_per_iteration": 2.668689012527466 + }, + { + "auxiliary_loss_clip": 0.01187449, + "auxiliary_loss_mlp": 0.0102861, + "balance_loss_clip": 1.04626489, + "balance_loss_mlp": 1.02078962, + "epoch": 0.5682679011603439, + "flos": 19536913584000.0, + "grad_norm": 1.7692444104538416, + "language_loss": 0.82944596, + "learning_rate": 1.6566252542958608e-06, + "loss": 0.85160661, + "num_input_tokens_seen": 102031900, + "step": 4726, + "time_per_iteration": 2.7086734771728516 + }, + { + "auxiliary_loss_clip": 0.01194461, + "auxiliary_loss_mlp": 0.01026874, + "balance_loss_clip": 1.04662776, + "balance_loss_mlp": 1.01890469, + "epoch": 0.568388144050983, + "flos": 28765488493440.0, + "grad_norm": 2.152744254554076, + "language_loss": 0.78225911, + "learning_rate": 1.6558578752859305e-06, + "loss": 0.80447251, + "num_input_tokens_seen": 102050860, + "step": 4727, + "time_per_iteration": 2.762875556945801 + }, + { + "auxiliary_loss_clip": 0.01201691, + "auxiliary_loss_mlp": 0.01024191, + "balance_loss_clip": 1.04527974, + "balance_loss_mlp": 1.01673698, + "epoch": 0.5685083869416221, + "flos": 21209452519680.0, + "grad_norm": 2.349359438086861, + "language_loss": 0.78564596, + "learning_rate": 1.6550905484820865e-06, + "loss": 0.80790484, + "num_input_tokens_seen": 102069320, + "step": 4728, + "time_per_iteration": 2.7381036281585693 + }, + { + "auxiliary_loss_clip": 0.01177247, + "auxiliary_loss_mlp": 0.01027205, + "balance_loss_clip": 1.04974687, + "balance_loss_mlp": 1.01943243, + "epoch": 0.5686286298322611, + "flos": 24827021942400.0, + "grad_norm": 2.6323732399555837, + "language_loss": 0.78560752, + "learning_rate": 1.6543232740007328e-06, + "loss": 0.80765206, + "num_input_tokens_seen": 102086435, + "step": 4729, + "time_per_iteration": 2.6075539588928223 + }, + { + "auxiliary_loss_clip": 0.01189667, + "auxiliary_loss_mlp": 0.01023645, + "balance_loss_clip": 1.05133009, + "balance_loss_mlp": 1.01590228, + "epoch": 0.5687488727229003, + "flos": 26615121909120.0, + "grad_norm": 2.660536355806763, + "language_loss": 0.67399198, + "learning_rate": 1.653556051958263e-06, + "loss": 0.69612503, + "num_input_tokens_seen": 102106115, + "step": 4730, + "time_per_iteration": 2.730152130126953 + }, + { + "auxiliary_loss_clip": 0.01201036, + "auxiliary_loss_mlp": 0.01027549, + "balance_loss_clip": 1.04213285, + "balance_loss_mlp": 1.01988721, + "epoch": 0.5688691156135394, + "flos": 20808725414400.0, + "grad_norm": 2.392290494183486, + "language_loss": 0.73760843, + "learning_rate": 1.6527888824710642e-06, + "loss": 0.75989425, + "num_input_tokens_seen": 102125715, + "step": 4731, + "time_per_iteration": 2.77139949798584 + }, + { + "auxiliary_loss_clip": 0.01193407, + "auxiliary_loss_mlp": 0.01027921, + "balance_loss_clip": 1.04319251, + "balance_loss_mlp": 1.01954699, + "epoch": 0.5689893585041784, + "flos": 25880963829120.0, + "grad_norm": 2.385331367965475, + "language_loss": 0.7644484, + "learning_rate": 1.6520217656555166e-06, + "loss": 0.78666162, + "num_input_tokens_seen": 102145005, + "step": 4732, + "time_per_iteration": 2.792149782180786 + }, + { + "auxiliary_loss_clip": 0.01181521, + "auxiliary_loss_mlp": 0.01028118, + "balance_loss_clip": 1.04587233, + "balance_loss_mlp": 1.02114475, + "epoch": 0.5691096013948175, + "flos": 23477463123840.0, + "grad_norm": 1.5649775069582368, + "language_loss": 0.70804095, + "learning_rate": 1.65125470162799e-06, + "loss": 0.73013735, + "num_input_tokens_seen": 102165360, + "step": 4733, + "time_per_iteration": 2.726076602935791 + }, + { + "auxiliary_loss_clip": 0.01206767, + "auxiliary_loss_mlp": 0.01026429, + "balance_loss_clip": 1.0469352, + "balance_loss_mlp": 1.0182811, + "epoch": 0.5692298442854566, + "flos": 18075600576000.0, + "grad_norm": 2.187775355448344, + "language_loss": 0.6983428, + "learning_rate": 1.6504876905048485e-06, + "loss": 0.72067469, + "num_input_tokens_seen": 102182320, + "step": 4734, + "time_per_iteration": 2.72792387008667 + }, + { + "auxiliary_loss_clip": 0.01178193, + "auxiliary_loss_mlp": 0.01024687, + "balance_loss_clip": 1.05339479, + "balance_loss_mlp": 1.01727843, + "epoch": 0.5693500871760957, + "flos": 23039317025280.0, + "grad_norm": 1.6340362889136006, + "language_loss": 0.72252274, + "learning_rate": 1.6497207324024464e-06, + "loss": 0.74455154, + "num_input_tokens_seen": 102201220, + "step": 4735, + "time_per_iteration": 2.579802989959717 + }, + { + "auxiliary_loss_clip": 0.01205423, + "auxiliary_loss_mlp": 0.01026066, + "balance_loss_clip": 1.05032897, + "balance_loss_mlp": 1.01807332, + "epoch": 0.5694703300667348, + "flos": 18989670902400.0, + "grad_norm": 1.9000645440685004, + "language_loss": 0.82710713, + "learning_rate": 1.6489538274371305e-06, + "loss": 0.84942198, + "num_input_tokens_seen": 102219825, + "step": 4736, + "time_per_iteration": 2.731051206588745 + }, + { + "auxiliary_loss_clip": 0.0118399, + "auxiliary_loss_mlp": 0.01023892, + "balance_loss_clip": 1.05292058, + "balance_loss_mlp": 1.01641166, + "epoch": 0.5695905729573739, + "flos": 21908705558400.0, + "grad_norm": 2.0471975448646345, + "language_loss": 0.83285344, + "learning_rate": 1.6481869757252396e-06, + "loss": 0.85493231, + "num_input_tokens_seen": 102238160, + "step": 4737, + "time_per_iteration": 2.647428035736084 + }, + { + "auxiliary_loss_clip": 0.0118819, + "auxiliary_loss_mlp": 0.01033304, + "balance_loss_clip": 1.05199575, + "balance_loss_mlp": 1.0264163, + "epoch": 0.569710815848013, + "flos": 28476659232000.0, + "grad_norm": 1.4803359326383196, + "language_loss": 0.71709144, + "learning_rate": 1.647420177383105e-06, + "loss": 0.73930633, + "num_input_tokens_seen": 102261030, + "step": 4738, + "time_per_iteration": 2.6609644889831543 + }, + { + "auxiliary_loss_clip": 0.0118175, + "auxiliary_loss_mlp": 0.01024049, + "balance_loss_clip": 1.05182791, + "balance_loss_mlp": 1.01641321, + "epoch": 0.569831058738652, + "flos": 28366162018560.0, + "grad_norm": 2.257294642305495, + "language_loss": 0.72609919, + "learning_rate": 1.646653432527049e-06, + "loss": 0.74815714, + "num_input_tokens_seen": 102281670, + "step": 4739, + "time_per_iteration": 2.6719486713409424 + }, + { + "auxiliary_loss_clip": 0.01204504, + "auxiliary_loss_mlp": 0.01025666, + "balance_loss_clip": 1.04757261, + "balance_loss_mlp": 1.01856685, + "epoch": 0.5699513016292912, + "flos": 25849973370240.0, + "grad_norm": 1.6427869849442969, + "language_loss": 0.74070281, + "learning_rate": 1.645886741273387e-06, + "loss": 0.76300454, + "num_input_tokens_seen": 102303485, + "step": 4740, + "time_per_iteration": 2.7495362758636475 + }, + { + "auxiliary_loss_clip": 0.0119706, + "auxiliary_loss_mlp": 0.01027449, + "balance_loss_clip": 1.04806709, + "balance_loss_mlp": 1.01973629, + "epoch": 0.5700715445199303, + "flos": 18037858360320.0, + "grad_norm": 2.9731307722686657, + "language_loss": 0.74049711, + "learning_rate": 1.645120103738424e-06, + "loss": 0.76274228, + "num_input_tokens_seen": 102320995, + "step": 4741, + "time_per_iteration": 2.7187256813049316 + }, + { + "auxiliary_loss_clip": 0.01172673, + "auxiliary_loss_mlp": 0.01378214, + "balance_loss_clip": 1.04824758, + "balance_loss_mlp": 1.00016379, + "epoch": 0.5701917874105693, + "flos": 11473352392320.0, + "grad_norm": 1.930321399063842, + "language_loss": 0.83715481, + "learning_rate": 1.6443535200384591e-06, + "loss": 0.86266375, + "num_input_tokens_seen": 102339170, + "step": 4742, + "time_per_iteration": 2.711185932159424 + }, + { + "auxiliary_loss_clip": 0.01178833, + "auxiliary_loss_mlp": 0.01025698, + "balance_loss_clip": 1.05366015, + "balance_loss_mlp": 1.01809812, + "epoch": 0.5703120303012085, + "flos": 21761759018880.0, + "grad_norm": 2.9774648009047797, + "language_loss": 0.70580578, + "learning_rate": 1.6435869902897827e-06, + "loss": 0.72785103, + "num_input_tokens_seen": 102357750, + "step": 4743, + "time_per_iteration": 4.569341659545898 + }, + { + "auxiliary_loss_clip": 0.01105556, + "auxiliary_loss_mlp": 0.01006091, + "balance_loss_clip": 1.01491201, + "balance_loss_mlp": 1.00476766, + "epoch": 0.5704322731918475, + "flos": 56746258513920.0, + "grad_norm": 0.7975801900444153, + "language_loss": 0.62057233, + "learning_rate": 1.6428205146086764e-06, + "loss": 0.64168882, + "num_input_tokens_seen": 102419730, + "step": 4744, + "time_per_iteration": 4.274521589279175 + }, + { + "auxiliary_loss_clip": 0.0120036, + "auxiliary_loss_mlp": 0.0103498, + "balance_loss_clip": 1.04865265, + "balance_loss_mlp": 1.02750552, + "epoch": 0.5705525160824866, + "flos": 20741141975040.0, + "grad_norm": 1.4994812365568706, + "language_loss": 0.70754373, + "learning_rate": 1.6420540931114142e-06, + "loss": 0.72989714, + "num_input_tokens_seen": 102440320, + "step": 4745, + "time_per_iteration": 2.774373769760132 + }, + { + "auxiliary_loss_clip": 0.01198809, + "auxiliary_loss_mlp": 0.01030258, + "balance_loss_clip": 1.05069017, + "balance_loss_mlp": 1.02232444, + "epoch": 0.5706727589731257, + "flos": 18771262254720.0, + "grad_norm": 2.8739494859419397, + "language_loss": 0.79271501, + "learning_rate": 1.6412877259142616e-06, + "loss": 0.81500566, + "num_input_tokens_seen": 102460240, + "step": 4746, + "time_per_iteration": 3.6549899578094482 + }, + { + "auxiliary_loss_clip": 0.01193673, + "auxiliary_loss_mlp": 0.01032463, + "balance_loss_clip": 1.05090594, + "balance_loss_mlp": 1.02421975, + "epoch": 0.5707930018637648, + "flos": 27634733372160.0, + "grad_norm": 2.0085180207323514, + "language_loss": 0.73808742, + "learning_rate": 1.6405214131334757e-06, + "loss": 0.7603488, + "num_input_tokens_seen": 102478765, + "step": 4747, + "time_per_iteration": 2.859628677368164 + }, + { + "auxiliary_loss_clip": 0.01209107, + "auxiliary_loss_mlp": 0.01028031, + "balance_loss_clip": 1.04904771, + "balance_loss_mlp": 1.02082777, + "epoch": 0.5709132447544039, + "flos": 27597673514880.0, + "grad_norm": 1.6640745579715495, + "language_loss": 0.79615873, + "learning_rate": 1.6397551548853052e-06, + "loss": 0.81853008, + "num_input_tokens_seen": 102496930, + "step": 4748, + "time_per_iteration": 2.810631036758423 + }, + { + "auxiliary_loss_clip": 0.01197664, + "auxiliary_loss_mlp": 0.01023704, + "balance_loss_clip": 1.05165589, + "balance_loss_mlp": 1.01540756, + "epoch": 0.571033487645043, + "flos": 21686095019520.0, + "grad_norm": 1.6762504246671335, + "language_loss": 0.70685935, + "learning_rate": 1.6389889512859917e-06, + "loss": 0.72907305, + "num_input_tokens_seen": 102516590, + "step": 4749, + "time_per_iteration": 2.754183530807495 + }, + { + "auxiliary_loss_clip": 0.01095733, + "auxiliary_loss_mlp": 0.01003132, + "balance_loss_clip": 1.01650977, + "balance_loss_mlp": 1.0017848, + "epoch": 0.5711537305356821, + "flos": 70181445980160.0, + "grad_norm": 0.8150568152821756, + "language_loss": 0.60382938, + "learning_rate": 1.638222802451767e-06, + "loss": 0.62481803, + "num_input_tokens_seen": 102578070, + "step": 4750, + "time_per_iteration": 3.274487257003784 + }, + { + "auxiliary_loss_clip": 0.01182891, + "auxiliary_loss_mlp": 0.01023692, + "balance_loss_clip": 1.05290127, + "balance_loss_mlp": 1.01651525, + "epoch": 0.5712739734263211, + "flos": 24717494396160.0, + "grad_norm": 2.0748095159938713, + "language_loss": 0.75231946, + "learning_rate": 1.6374567084988561e-06, + "loss": 0.77438527, + "num_input_tokens_seen": 102599255, + "step": 4751, + "time_per_iteration": 2.6633858680725098 + }, + { + "auxiliary_loss_clip": 0.01201549, + "auxiliary_loss_mlp": 0.01034087, + "balance_loss_clip": 1.05292201, + "balance_loss_mlp": 1.02520037, + "epoch": 0.5713942163169603, + "flos": 26578169792640.0, + "grad_norm": 1.889077786193702, + "language_loss": 0.76619709, + "learning_rate": 1.6366906695434738e-06, + "loss": 0.78855348, + "num_input_tokens_seen": 102621775, + "step": 4752, + "time_per_iteration": 2.75439453125 + }, + { + "auxiliary_loss_clip": 0.01189483, + "auxiliary_loss_mlp": 0.01028068, + "balance_loss_clip": 1.054268, + "balance_loss_mlp": 1.02015829, + "epoch": 0.5715144592075994, + "flos": 21142443697920.0, + "grad_norm": 1.9808534029718272, + "language_loss": 0.86116379, + "learning_rate": 1.6359246857018275e-06, + "loss": 0.88333929, + "num_input_tokens_seen": 102639305, + "step": 4753, + "time_per_iteration": 2.6631178855895996 + }, + { + "auxiliary_loss_clip": 0.0120859, + "auxiliary_loss_mlp": 0.01027748, + "balance_loss_clip": 1.04374099, + "balance_loss_mlp": 1.02026749, + "epoch": 0.5716347020982384, + "flos": 23330265189120.0, + "grad_norm": 2.151612245147142, + "language_loss": 0.78257805, + "learning_rate": 1.6351587570901178e-06, + "loss": 0.80494148, + "num_input_tokens_seen": 102659430, + "step": 4754, + "time_per_iteration": 2.7838029861450195 + }, + { + "auxiliary_loss_clip": 0.01202701, + "auxiliary_loss_mlp": 0.01026492, + "balance_loss_clip": 1.04733717, + "balance_loss_mlp": 1.01880264, + "epoch": 0.5717549449888776, + "flos": 17009555806080.0, + "grad_norm": 2.5376569068441324, + "language_loss": 0.75083899, + "learning_rate": 1.634392883824534e-06, + "loss": 0.77313089, + "num_input_tokens_seen": 102671430, + "step": 4755, + "time_per_iteration": 2.6631762981414795 + }, + { + "auxiliary_loss_clip": 0.01213284, + "auxiliary_loss_mlp": 0.01030567, + "balance_loss_clip": 1.04456019, + "balance_loss_mlp": 1.02232361, + "epoch": 0.5718751878795166, + "flos": 35518130922240.0, + "grad_norm": 1.5767903381967439, + "language_loss": 0.67686957, + "learning_rate": 1.6336270660212595e-06, + "loss": 0.69930804, + "num_input_tokens_seen": 102693025, + "step": 4756, + "time_per_iteration": 2.860032081604004 + }, + { + "auxiliary_loss_clip": 0.01191602, + "auxiliary_loss_mlp": 0.01028018, + "balance_loss_clip": 1.05068398, + "balance_loss_mlp": 1.02008414, + "epoch": 0.5719954307701557, + "flos": 38613989255040.0, + "grad_norm": 5.582013934611639, + "language_loss": 0.65549731, + "learning_rate": 1.6328613037964676e-06, + "loss": 0.67769355, + "num_input_tokens_seen": 102716090, + "step": 4757, + "time_per_iteration": 2.803731918334961 + }, + { + "auxiliary_loss_clip": 0.01185823, + "auxiliary_loss_mlp": 0.01025799, + "balance_loss_clip": 1.050385, + "balance_loss_mlp": 1.01853967, + "epoch": 0.5721156736607949, + "flos": 20631111638400.0, + "grad_norm": 3.7685320118645724, + "language_loss": 0.68010682, + "learning_rate": 1.6320955972663241e-06, + "loss": 0.70222306, + "num_input_tokens_seen": 102735685, + "step": 4758, + "time_per_iteration": 2.6573455333709717 + }, + { + "auxiliary_loss_clip": 0.0118551, + "auxiliary_loss_mlp": 0.01025701, + "balance_loss_clip": 1.04955029, + "balance_loss_mlp": 1.01802957, + "epoch": 0.5722359165514339, + "flos": 37415076076800.0, + "grad_norm": 3.9215366144033164, + "language_loss": 0.6520794, + "learning_rate": 1.6313299465469857e-06, + "loss": 0.67419147, + "num_input_tokens_seen": 102758415, + "step": 4759, + "time_per_iteration": 2.802989959716797 + }, + { + "auxiliary_loss_clip": 0.01182133, + "auxiliary_loss_mlp": 0.01026418, + "balance_loss_clip": 1.05028605, + "balance_loss_mlp": 1.01839495, + "epoch": 0.572356159442073, + "flos": 21972877205760.0, + "grad_norm": 2.8958351109044913, + "language_loss": 0.79449862, + "learning_rate": 1.6305643517546014e-06, + "loss": 0.81658423, + "num_input_tokens_seen": 102773795, + "step": 4760, + "time_per_iteration": 2.700590133666992 + }, + { + "auxiliary_loss_clip": 0.01178747, + "auxiliary_loss_mlp": 0.01025123, + "balance_loss_clip": 1.05412197, + "balance_loss_mlp": 1.01790202, + "epoch": 0.5724764023327121, + "flos": 19135540033920.0, + "grad_norm": 2.0282755366224046, + "language_loss": 0.84758902, + "learning_rate": 1.629798813005311e-06, + "loss": 0.86962771, + "num_input_tokens_seen": 102793515, + "step": 4761, + "time_per_iteration": 2.631070613861084 + }, + { + "auxiliary_loss_clip": 0.01213629, + "auxiliary_loss_mlp": 0.01020959, + "balance_loss_clip": 1.04878497, + "balance_loss_mlp": 1.01396728, + "epoch": 0.5725966452233512, + "flos": 22819759142400.0, + "grad_norm": 2.50267772363466, + "language_loss": 0.70868027, + "learning_rate": 1.6290333304152473e-06, + "loss": 0.73102617, + "num_input_tokens_seen": 102813390, + "step": 4762, + "time_per_iteration": 2.7801406383514404 + }, + { + "auxiliary_loss_clip": 0.01191164, + "auxiliary_loss_mlp": 0.0102971, + "balance_loss_clip": 1.05054832, + "balance_loss_mlp": 1.02210426, + "epoch": 0.5727168881139902, + "flos": 41496610498560.0, + "grad_norm": 2.009604681091092, + "language_loss": 0.56894839, + "learning_rate": 1.6282679041005314e-06, + "loss": 0.5911572, + "num_input_tokens_seen": 102838980, + "step": 4763, + "time_per_iteration": 3.0059618949890137 + }, + { + "auxiliary_loss_clip": 0.01184999, + "auxiliary_loss_mlp": 0.01028542, + "balance_loss_clip": 1.04628909, + "balance_loss_mlp": 1.02047455, + "epoch": 0.5728371310046293, + "flos": 14647675985280.0, + "grad_norm": 2.1285934513924105, + "language_loss": 0.87072009, + "learning_rate": 1.6275025341772789e-06, + "loss": 0.89285547, + "num_input_tokens_seen": 102855285, + "step": 4764, + "time_per_iteration": 2.7160232067108154 + }, + { + "auxiliary_loss_clip": 0.01194854, + "auxiliary_loss_mlp": 0.01030397, + "balance_loss_clip": 1.04862869, + "balance_loss_mlp": 1.02215672, + "epoch": 0.5729573738952685, + "flos": 21506613736320.0, + "grad_norm": 2.2136668723257085, + "language_loss": 0.81815732, + "learning_rate": 1.626737220761596e-06, + "loss": 0.84040976, + "num_input_tokens_seen": 102872750, + "step": 4765, + "time_per_iteration": 2.7283742427825928 + }, + { + "auxiliary_loss_clip": 0.0118402, + "auxiliary_loss_mlp": 0.01026634, + "balance_loss_clip": 1.05199504, + "balance_loss_mlp": 1.01923752, + "epoch": 0.5730776167859075, + "flos": 23621680229760.0, + "grad_norm": 2.1771583333708464, + "language_loss": 0.78581923, + "learning_rate": 1.62597196396958e-06, + "loss": 0.80792576, + "num_input_tokens_seen": 102890920, + "step": 4766, + "time_per_iteration": 2.65975022315979 + }, + { + "auxiliary_loss_clip": 0.01187065, + "auxiliary_loss_mlp": 0.01026618, + "balance_loss_clip": 1.05267859, + "balance_loss_mlp": 1.01838064, + "epoch": 0.5731978596765466, + "flos": 25739224761600.0, + "grad_norm": 2.1913488273143797, + "language_loss": 0.85307133, + "learning_rate": 1.6252067639173197e-06, + "loss": 0.87520814, + "num_input_tokens_seen": 102912830, + "step": 4767, + "time_per_iteration": 2.685936212539673 + }, + { + "auxiliary_loss_clip": 0.01187978, + "auxiliary_loss_mlp": 0.01025455, + "balance_loss_clip": 1.05113828, + "balance_loss_mlp": 1.01772451, + "epoch": 0.5733181025671857, + "flos": 26359509749760.0, + "grad_norm": 1.9015254829317498, + "language_loss": 0.69895118, + "learning_rate": 1.6244416207208956e-06, + "loss": 0.72108555, + "num_input_tokens_seen": 102933765, + "step": 4768, + "time_per_iteration": 2.651395797729492 + }, + { + "auxiliary_loss_clip": 0.01202848, + "auxiliary_loss_mlp": 0.01026228, + "balance_loss_clip": 1.04772592, + "balance_loss_mlp": 1.01869404, + "epoch": 0.5734383454578248, + "flos": 29423874833280.0, + "grad_norm": 1.845047190745912, + "language_loss": 0.73823094, + "learning_rate": 1.6236765344963787e-06, + "loss": 0.76052171, + "num_input_tokens_seen": 102955025, + "step": 4769, + "time_per_iteration": 4.479487895965576 + }, + { + "auxiliary_loss_clip": 0.01195433, + "auxiliary_loss_mlp": 0.01025111, + "balance_loss_clip": 1.05133271, + "balance_loss_mlp": 1.01727343, + "epoch": 0.5735585883484638, + "flos": 34969954487040.0, + "grad_norm": 3.383500056224, + "language_loss": 0.69275862, + "learning_rate": 1.6229115053598322e-06, + "loss": 0.71496409, + "num_input_tokens_seen": 102976780, + "step": 4770, + "time_per_iteration": 3.6855883598327637 + }, + { + "auxiliary_loss_clip": 0.01189603, + "auxiliary_loss_mlp": 0.01030177, + "balance_loss_clip": 1.05371332, + "balance_loss_mlp": 1.0227623, + "epoch": 0.573678831239103, + "flos": 18770759464320.0, + "grad_norm": 1.9995419433947477, + "language_loss": 0.72244662, + "learning_rate": 1.6221465334273108e-06, + "loss": 0.7446444, + "num_input_tokens_seen": 102995990, + "step": 4771, + "time_per_iteration": 2.6783673763275146 + }, + { + "auxiliary_loss_clip": 0.01209394, + "auxiliary_loss_mlp": 0.01026684, + "balance_loss_clip": 1.04892349, + "balance_loss_mlp": 1.0191319, + "epoch": 0.5737990741297421, + "flos": 25702883176320.0, + "grad_norm": 2.6158964096297335, + "language_loss": 0.61759734, + "learning_rate": 1.6213816188148593e-06, + "loss": 0.63995814, + "num_input_tokens_seen": 103014695, + "step": 4772, + "time_per_iteration": 2.733689546585083 + }, + { + "auxiliary_loss_clip": 0.01183647, + "auxiliary_loss_mlp": 0.0102838, + "balance_loss_clip": 1.04807007, + "balance_loss_mlp": 1.02126598, + "epoch": 0.5739193170203811, + "flos": 27269234530560.0, + "grad_norm": 1.5734794879468268, + "language_loss": 0.77103269, + "learning_rate": 1.6206167616385162e-06, + "loss": 0.79315293, + "num_input_tokens_seen": 103035760, + "step": 4773, + "time_per_iteration": 3.6787490844726562 + }, + { + "auxiliary_loss_clip": 0.01201383, + "auxiliary_loss_mlp": 0.01029302, + "balance_loss_clip": 1.05229211, + "balance_loss_mlp": 1.02124906, + "epoch": 0.5740395599110203, + "flos": 12239721993600.0, + "grad_norm": 1.89094182883935, + "language_loss": 0.73745644, + "learning_rate": 1.6198519620143078e-06, + "loss": 0.7597633, + "num_input_tokens_seen": 103052915, + "step": 4774, + "time_per_iteration": 2.67010235786438 + }, + { + "auxiliary_loss_clip": 0.01207455, + "auxiliary_loss_mlp": 0.01027462, + "balance_loss_clip": 1.04839683, + "balance_loss_mlp": 1.01957083, + "epoch": 0.5741598028016593, + "flos": 25921399564800.0, + "grad_norm": 1.6494065125497137, + "language_loss": 0.78192514, + "learning_rate": 1.6190872200582546e-06, + "loss": 0.80427426, + "num_input_tokens_seen": 103074655, + "step": 4775, + "time_per_iteration": 2.787111759185791 + }, + { + "auxiliary_loss_clip": 0.01189624, + "auxiliary_loss_mlp": 0.01378678, + "balance_loss_clip": 1.04821658, + "balance_loss_mlp": 1.00015914, + "epoch": 0.5742800456922984, + "flos": 19244133826560.0, + "grad_norm": 3.0148482803565213, + "language_loss": 0.7788949, + "learning_rate": 1.6183225358863676e-06, + "loss": 0.80457795, + "num_input_tokens_seen": 103091550, + "step": 4776, + "time_per_iteration": 2.664595365524292 + }, + { + "auxiliary_loss_clip": 0.01185311, + "auxiliary_loss_mlp": 0.01025465, + "balance_loss_clip": 1.04607415, + "balance_loss_mlp": 1.01744807, + "epoch": 0.5744002885829376, + "flos": 30920487932160.0, + "grad_norm": 2.3954749933987207, + "language_loss": 0.72038364, + "learning_rate": 1.617557909614648e-06, + "loss": 0.74249136, + "num_input_tokens_seen": 103110985, + "step": 4777, + "time_per_iteration": 2.750925302505493 + }, + { + "auxiliary_loss_clip": 0.01198096, + "auxiliary_loss_mlp": 0.01030668, + "balance_loss_clip": 1.0458504, + "balance_loss_mlp": 1.02319109, + "epoch": 0.5745205314735766, + "flos": 23840017050240.0, + "grad_norm": 1.9679160790447816, + "language_loss": 0.86385322, + "learning_rate": 1.6167933413590899e-06, + "loss": 0.88614094, + "num_input_tokens_seen": 103129890, + "step": 4778, + "time_per_iteration": 2.7711572647094727 + }, + { + "auxiliary_loss_clip": 0.01186896, + "auxiliary_loss_mlp": 0.0103204, + "balance_loss_clip": 1.05247998, + "balance_loss_mlp": 1.0238626, + "epoch": 0.5746407743642157, + "flos": 12311902373760.0, + "grad_norm": 2.4258942818025657, + "language_loss": 0.90584075, + "learning_rate": 1.6160288312356773e-06, + "loss": 0.92803013, + "num_input_tokens_seen": 103147020, + "step": 4779, + "time_per_iteration": 2.5931711196899414 + }, + { + "auxiliary_loss_clip": 0.01189001, + "auxiliary_loss_mlp": 0.01025498, + "balance_loss_clip": 1.05039752, + "balance_loss_mlp": 1.01803541, + "epoch": 0.5747610172548548, + "flos": 24133658734080.0, + "grad_norm": 1.8305776733731465, + "language_loss": 0.81754726, + "learning_rate": 1.6152643793603857e-06, + "loss": 0.83969223, + "num_input_tokens_seen": 103167370, + "step": 4780, + "time_per_iteration": 2.7874908447265625 + }, + { + "auxiliary_loss_clip": 0.01176968, + "auxiliary_loss_mlp": 0.01028863, + "balance_loss_clip": 1.05164444, + "balance_loss_mlp": 1.02088833, + "epoch": 0.5748812601454939, + "flos": 25408451393280.0, + "grad_norm": 1.8717545566221356, + "language_loss": 0.87908888, + "learning_rate": 1.6144999858491815e-06, + "loss": 0.90114719, + "num_input_tokens_seen": 103186000, + "step": 4781, + "time_per_iteration": 2.624628782272339 + }, + { + "auxiliary_loss_clip": 0.01197938, + "auxiliary_loss_mlp": 0.01022936, + "balance_loss_clip": 1.04863095, + "balance_loss_mlp": 1.01469851, + "epoch": 0.575001503036133, + "flos": 30624942827520.0, + "grad_norm": 3.416795237685919, + "language_loss": 0.85910976, + "learning_rate": 1.6137356508180232e-06, + "loss": 0.88131845, + "num_input_tokens_seen": 103207710, + "step": 4782, + "time_per_iteration": 2.7461915016174316 + }, + { + "auxiliary_loss_clip": 0.01177689, + "auxiliary_loss_mlp": 0.01378833, + "balance_loss_clip": 1.05106807, + "balance_loss_mlp": 1.00015235, + "epoch": 0.5751217459267721, + "flos": 21726566668800.0, + "grad_norm": 2.293829302182643, + "language_loss": 0.80993915, + "learning_rate": 1.6129713743828593e-06, + "loss": 0.83550429, + "num_input_tokens_seen": 103226720, + "step": 4783, + "time_per_iteration": 2.585479497909546 + }, + { + "auxiliary_loss_clip": 0.01192809, + "auxiliary_loss_mlp": 0.01024937, + "balance_loss_clip": 1.04661238, + "balance_loss_mlp": 1.01780534, + "epoch": 0.5752419888174112, + "flos": 21651620941440.0, + "grad_norm": 3.010472948757005, + "language_loss": 0.75591862, + "learning_rate": 1.6122071566596306e-06, + "loss": 0.77809608, + "num_input_tokens_seen": 103246995, + "step": 4784, + "time_per_iteration": 2.8080015182495117 + }, + { + "auxiliary_loss_clip": 0.0118797, + "auxiliary_loss_mlp": 0.01029895, + "balance_loss_clip": 1.05102348, + "balance_loss_mlp": 1.02228332, + "epoch": 0.5753622317080502, + "flos": 17775997234560.0, + "grad_norm": 2.4040024470966874, + "language_loss": 0.83082616, + "learning_rate": 1.6114429977642674e-06, + "loss": 0.85300481, + "num_input_tokens_seen": 103261500, + "step": 4785, + "time_per_iteration": 2.530289649963379 + }, + { + "auxiliary_loss_clip": 0.01187319, + "auxiliary_loss_mlp": 0.01030533, + "balance_loss_clip": 1.05324411, + "balance_loss_mlp": 1.02353287, + "epoch": 0.5754824745986894, + "flos": 19789616741760.0, + "grad_norm": 1.6639193636083833, + "language_loss": 0.74153411, + "learning_rate": 1.6106788978126926e-06, + "loss": 0.76371264, + "num_input_tokens_seen": 103280475, + "step": 4786, + "time_per_iteration": 2.6612958908081055 + }, + { + "auxiliary_loss_clip": 0.01206462, + "auxiliary_loss_mlp": 0.01026435, + "balance_loss_clip": 1.04225278, + "balance_loss_mlp": 1.01841259, + "epoch": 0.5756027174893285, + "flos": 30985665160320.0, + "grad_norm": 2.8503491016850924, + "language_loss": 0.78528178, + "learning_rate": 1.6099148569208196e-06, + "loss": 0.80761075, + "num_input_tokens_seen": 103297695, + "step": 4787, + "time_per_iteration": 2.804701805114746 + }, + { + "auxiliary_loss_clip": 0.01194272, + "auxiliary_loss_mlp": 0.0102775, + "balance_loss_clip": 1.05114949, + "balance_loss_mlp": 1.01984012, + "epoch": 0.5757229603799675, + "flos": 28546864364160.0, + "grad_norm": 1.8965882151813527, + "language_loss": 0.62709117, + "learning_rate": 1.6091508752045523e-06, + "loss": 0.64931136, + "num_input_tokens_seen": 103318575, + "step": 4788, + "time_per_iteration": 2.7485127449035645 + }, + { + "auxiliary_loss_clip": 0.01191422, + "auxiliary_loss_mlp": 0.01028736, + "balance_loss_clip": 1.04231083, + "balance_loss_mlp": 1.02064121, + "epoch": 0.5758432032706067, + "flos": 22999024944000.0, + "grad_norm": 1.5979812133747526, + "language_loss": 0.86290514, + "learning_rate": 1.608386952779787e-06, + "loss": 0.88510674, + "num_input_tokens_seen": 103337945, + "step": 4789, + "time_per_iteration": 2.728635787963867 + }, + { + "auxiliary_loss_clip": 0.01199626, + "auxiliary_loss_mlp": 0.01027373, + "balance_loss_clip": 1.05001712, + "balance_loss_mlp": 1.01931143, + "epoch": 0.5759634461612457, + "flos": 25739727552000.0, + "grad_norm": 1.587811981808485, + "language_loss": 0.74758101, + "learning_rate": 1.6076230897624098e-06, + "loss": 0.76985097, + "num_input_tokens_seen": 103360150, + "step": 4790, + "time_per_iteration": 2.722787380218506 + }, + { + "auxiliary_loss_clip": 0.01186291, + "auxiliary_loss_mlp": 0.01028729, + "balance_loss_clip": 1.04908121, + "balance_loss_mlp": 1.02068245, + "epoch": 0.5760836890518848, + "flos": 30591761639040.0, + "grad_norm": 2.1323656091621834, + "language_loss": 0.77352846, + "learning_rate": 1.6068592862682974e-06, + "loss": 0.79567868, + "num_input_tokens_seen": 103378305, + "step": 4791, + "time_per_iteration": 2.7700233459472656 + }, + { + "auxiliary_loss_clip": 0.01199254, + "auxiliary_loss_mlp": 0.01028573, + "balance_loss_clip": 1.05172861, + "balance_loss_mlp": 1.02111936, + "epoch": 0.576203931942524, + "flos": 36538963447680.0, + "grad_norm": 1.7999128603595542, + "language_loss": 0.73818582, + "learning_rate": 1.6060955424133187e-06, + "loss": 0.76046407, + "num_input_tokens_seen": 103399230, + "step": 4792, + "time_per_iteration": 2.8996293544769287 + }, + { + "auxiliary_loss_clip": 0.0118486, + "auxiliary_loss_mlp": 0.01024578, + "balance_loss_clip": 1.0511241, + "balance_loss_mlp": 1.01635814, + "epoch": 0.576324174833163, + "flos": 25516937445120.0, + "grad_norm": 3.6879083043451217, + "language_loss": 0.893677, + "learning_rate": 1.6053318583133332e-06, + "loss": 0.91577142, + "num_input_tokens_seen": 103420100, + "step": 4793, + "time_per_iteration": 2.686967372894287 + }, + { + "auxiliary_loss_clip": 0.01183409, + "auxiliary_loss_mlp": 0.01022962, + "balance_loss_clip": 1.04962075, + "balance_loss_mlp": 1.01551771, + "epoch": 0.5764444177238021, + "flos": 25119262995840.0, + "grad_norm": 2.369787117982701, + "language_loss": 0.74926293, + "learning_rate": 1.6045682340841907e-06, + "loss": 0.77132666, + "num_input_tokens_seen": 103439025, + "step": 4794, + "time_per_iteration": 2.752265453338623 + }, + { + "auxiliary_loss_clip": 0.01112912, + "auxiliary_loss_mlp": 0.01373794, + "balance_loss_clip": 1.01751304, + "balance_loss_mlp": 0.9998706, + "epoch": 0.5765646606144411, + "flos": 62212687758720.0, + "grad_norm": 0.7493410378436927, + "language_loss": 0.58024514, + "learning_rate": 1.6038046698417336e-06, + "loss": 0.6051122, + "num_input_tokens_seen": 103499920, + "step": 4795, + "time_per_iteration": 5.037665367126465 + }, + { + "auxiliary_loss_clip": 0.01184823, + "auxiliary_loss_mlp": 0.01028481, + "balance_loss_clip": 1.0498625, + "balance_loss_mlp": 1.02017832, + "epoch": 0.5766849035050803, + "flos": 25118760205440.0, + "grad_norm": 1.7637198589306777, + "language_loss": 0.68365031, + "learning_rate": 1.6030411657017919e-06, + "loss": 0.70578337, + "num_input_tokens_seen": 103519575, + "step": 4796, + "time_per_iteration": 3.623316526412964 + }, + { + "auxiliary_loss_clip": 0.0117685, + "auxiliary_loss_mlp": 0.01024551, + "balance_loss_clip": 1.04971242, + "balance_loss_mlp": 1.01747012, + "epoch": 0.5768051463957193, + "flos": 15991093578240.0, + "grad_norm": 1.890727893079794, + "language_loss": 0.84441018, + "learning_rate": 1.6022777217801903e-06, + "loss": 0.86642414, + "num_input_tokens_seen": 103536530, + "step": 4797, + "time_per_iteration": 2.5901365280151367 + }, + { + "auxiliary_loss_clip": 0.01209863, + "auxiliary_loss_mlp": 0.01027575, + "balance_loss_clip": 1.05383301, + "balance_loss_mlp": 1.0194931, + "epoch": 0.5769253892863584, + "flos": 22163635359360.0, + "grad_norm": 2.346129741176678, + "language_loss": 0.73858368, + "learning_rate": 1.601514338192742e-06, + "loss": 0.76095808, + "num_input_tokens_seen": 103556460, + "step": 4798, + "time_per_iteration": 3.5992770195007324 + }, + { + "auxiliary_loss_clip": 0.01173882, + "auxiliary_loss_mlp": 0.01022884, + "balance_loss_clip": 1.05162072, + "balance_loss_mlp": 1.01575243, + "epoch": 0.5770456321769976, + "flos": 22856388036480.0, + "grad_norm": 2.8230430258099735, + "language_loss": 0.71565241, + "learning_rate": 1.6007510150552514e-06, + "loss": 0.73762012, + "num_input_tokens_seen": 103574520, + "step": 4799, + "time_per_iteration": 2.589027166366577 + }, + { + "auxiliary_loss_clip": 0.01190207, + "auxiliary_loss_mlp": 0.01027796, + "balance_loss_clip": 1.05013537, + "balance_loss_mlp": 1.01920366, + "epoch": 0.5771658750676366, + "flos": 46353672489600.0, + "grad_norm": 1.495351583141527, + "language_loss": 0.62303042, + "learning_rate": 1.599987752483515e-06, + "loss": 0.64521044, + "num_input_tokens_seen": 103598965, + "step": 4800, + "time_per_iteration": 2.8211605548858643 + }, + { + "auxiliary_loss_clip": 0.01193535, + "auxiliary_loss_mlp": 0.01026678, + "balance_loss_clip": 1.04477549, + "balance_loss_mlp": 1.01927173, + "epoch": 0.5772861179582757, + "flos": 22159972172160.0, + "grad_norm": 1.6830468488260446, + "language_loss": 0.67977065, + "learning_rate": 1.5992245505933184e-06, + "loss": 0.70197272, + "num_input_tokens_seen": 103618665, + "step": 4801, + "time_per_iteration": 2.735440731048584 + }, + { + "auxiliary_loss_clip": 0.01180318, + "auxiliary_loss_mlp": 0.0103177, + "balance_loss_clip": 1.05390286, + "balance_loss_mlp": 1.02383053, + "epoch": 0.5774063608489148, + "flos": 31248926916480.0, + "grad_norm": 2.1427884506284185, + "language_loss": 0.71234202, + "learning_rate": 1.5984614095004388e-06, + "loss": 0.73446286, + "num_input_tokens_seen": 103639800, + "step": 4802, + "time_per_iteration": 2.770792007446289 + }, + { + "auxiliary_loss_clip": 0.01176545, + "auxiliary_loss_mlp": 0.01031622, + "balance_loss_clip": 1.0478543, + "balance_loss_mlp": 1.02375376, + "epoch": 0.5775266037395539, + "flos": 22527123039360.0, + "grad_norm": 2.0727116919953623, + "language_loss": 0.80911523, + "learning_rate": 1.5976983293206438e-06, + "loss": 0.8311969, + "num_input_tokens_seen": 103655605, + "step": 4803, + "time_per_iteration": 2.6991071701049805 + }, + { + "auxiliary_loss_clip": 0.01188382, + "auxiliary_loss_mlp": 0.01026689, + "balance_loss_clip": 1.04557467, + "balance_loss_mlp": 1.01905942, + "epoch": 0.577646846630193, + "flos": 21068790860160.0, + "grad_norm": 14.274625439716555, + "language_loss": 0.7160399, + "learning_rate": 1.5969353101696928e-06, + "loss": 0.73819059, + "num_input_tokens_seen": 103674045, + "step": 4804, + "time_per_iteration": 2.65514874458313 + }, + { + "auxiliary_loss_clip": 0.01184465, + "auxiliary_loss_mlp": 0.01030243, + "balance_loss_clip": 1.04875863, + "balance_loss_mlp": 1.02271461, + "epoch": 0.5777670895208321, + "flos": 29714284293120.0, + "grad_norm": 1.6881081364594006, + "language_loss": 0.79349661, + "learning_rate": 1.5961723521633341e-06, + "loss": 0.81564373, + "num_input_tokens_seen": 103695285, + "step": 4805, + "time_per_iteration": 2.7793328762054443 + }, + { + "auxiliary_loss_clip": 0.0118773, + "auxiliary_loss_mlp": 0.01024255, + "balance_loss_clip": 1.04736722, + "balance_loss_mlp": 1.01679575, + "epoch": 0.5778873324114712, + "flos": 19500428344320.0, + "grad_norm": 2.181900412966423, + "language_loss": 0.9042663, + "learning_rate": 1.5954094554173097e-06, + "loss": 0.92638612, + "num_input_tokens_seen": 103713275, + "step": 4806, + "time_per_iteration": 2.6690189838409424 + }, + { + "auxiliary_loss_clip": 0.01201002, + "auxiliary_loss_mlp": 0.01019287, + "balance_loss_clip": 1.05111563, + "balance_loss_mlp": 1.01245689, + "epoch": 0.5780075753021102, + "flos": 14136846716160.0, + "grad_norm": 1.9433650574981685, + "language_loss": 0.79119712, + "learning_rate": 1.5946466200473482e-06, + "loss": 0.81340003, + "num_input_tokens_seen": 103731185, + "step": 4807, + "time_per_iteration": 2.633751153945923 + }, + { + "auxiliary_loss_clip": 0.01197251, + "auxiliary_loss_mlp": 0.01025772, + "balance_loss_clip": 1.0471499, + "balance_loss_mlp": 1.01870322, + "epoch": 0.5781278181927494, + "flos": 15262178883840.0, + "grad_norm": 1.9041963853773922, + "language_loss": 0.8341338, + "learning_rate": 1.5938838461691723e-06, + "loss": 0.85636407, + "num_input_tokens_seen": 103748095, + "step": 4808, + "time_per_iteration": 2.695213794708252 + }, + { + "auxiliary_loss_clip": 0.01180428, + "auxiliary_loss_mlp": 0.01029715, + "balance_loss_clip": 1.05444908, + "balance_loss_mlp": 1.0220443, + "epoch": 0.5782480610833884, + "flos": 16726831856640.0, + "grad_norm": 3.296296451106676, + "language_loss": 0.82680118, + "learning_rate": 1.593121133898494e-06, + "loss": 0.84890264, + "num_input_tokens_seen": 103765300, + "step": 4809, + "time_per_iteration": 2.5641534328460693 + }, + { + "auxiliary_loss_clip": 0.01194744, + "auxiliary_loss_mlp": 0.01029784, + "balance_loss_clip": 1.05298936, + "balance_loss_mlp": 1.02154088, + "epoch": 0.5783683039740275, + "flos": 25482140144640.0, + "grad_norm": 2.108882324526585, + "language_loss": 0.79220885, + "learning_rate": 1.592358483351016e-06, + "loss": 0.8144542, + "num_input_tokens_seen": 103785475, + "step": 4810, + "time_per_iteration": 2.704834461212158 + }, + { + "auxiliary_loss_clip": 0.01184614, + "auxiliary_loss_mlp": 0.01024593, + "balance_loss_clip": 1.0500896, + "balance_loss_mlp": 1.01729107, + "epoch": 0.5784885468646667, + "flos": 18405835240320.0, + "grad_norm": 2.0528379940961354, + "language_loss": 0.7231313, + "learning_rate": 1.5915958946424326e-06, + "loss": 0.7452234, + "num_input_tokens_seen": 103804160, + "step": 4811, + "time_per_iteration": 2.600417137145996 + }, + { + "auxiliary_loss_clip": 0.01204357, + "auxiliary_loss_mlp": 0.01379246, + "balance_loss_clip": 1.04771352, + "balance_loss_mlp": 1.0001725, + "epoch": 0.5786087897553057, + "flos": 46100717936640.0, + "grad_norm": 1.8553706284583507, + "language_loss": 0.74427634, + "learning_rate": 1.5908333678884271e-06, + "loss": 0.77011245, + "num_input_tokens_seen": 103830580, + "step": 4812, + "time_per_iteration": 2.9886412620544434 + }, + { + "auxiliary_loss_clip": 0.01185526, + "auxiliary_loss_mlp": 0.01024475, + "balance_loss_clip": 1.05058384, + "balance_loss_mlp": 1.01696491, + "epoch": 0.5787290326459448, + "flos": 12385950261120.0, + "grad_norm": 1.9508367475598687, + "language_loss": 0.7380631, + "learning_rate": 1.5900709032046743e-06, + "loss": 0.76016307, + "num_input_tokens_seen": 103848655, + "step": 4813, + "time_per_iteration": 2.685457229614258 + }, + { + "auxiliary_loss_clip": 0.01196039, + "auxiliary_loss_mlp": 0.01023584, + "balance_loss_clip": 1.05353725, + "balance_loss_mlp": 1.01615751, + "epoch": 0.5788492755365839, + "flos": 23290332243840.0, + "grad_norm": 6.91880045363904, + "language_loss": 0.78126955, + "learning_rate": 1.5893085007068391e-06, + "loss": 0.80346584, + "num_input_tokens_seen": 103866215, + "step": 4814, + "time_per_iteration": 2.7484130859375 + }, + { + "auxiliary_loss_clip": 0.01179964, + "auxiliary_loss_mlp": 0.0103486, + "balance_loss_clip": 1.04487288, + "balance_loss_mlp": 1.02638388, + "epoch": 0.578969518427223, + "flos": 24061047390720.0, + "grad_norm": 2.1706790827454174, + "language_loss": 0.70701647, + "learning_rate": 1.5885461605105786e-06, + "loss": 0.72916472, + "num_input_tokens_seen": 103887815, + "step": 4815, + "time_per_iteration": 2.6977715492248535 + }, + { + "auxiliary_loss_clip": 0.01195141, + "auxiliary_loss_mlp": 0.010307, + "balance_loss_clip": 1.04869175, + "balance_loss_mlp": 1.0226351, + "epoch": 0.579089761317862, + "flos": 21871825269120.0, + "grad_norm": 2.1615622256303326, + "language_loss": 0.76957691, + "learning_rate": 1.5877838827315375e-06, + "loss": 0.79183531, + "num_input_tokens_seen": 103906360, + "step": 4816, + "time_per_iteration": 2.7050514221191406 + }, + { + "auxiliary_loss_clip": 0.0117613, + "auxiliary_loss_mlp": 0.01025908, + "balance_loss_clip": 1.05182481, + "balance_loss_mlp": 1.01788473, + "epoch": 0.5792100042085012, + "flos": 22929681738240.0, + "grad_norm": 1.7225727698253972, + "language_loss": 0.70664269, + "learning_rate": 1.587021667485355e-06, + "loss": 0.72866309, + "num_input_tokens_seen": 103925730, + "step": 4817, + "time_per_iteration": 2.667398452758789 + }, + { + "auxiliary_loss_clip": 0.01197439, + "auxiliary_loss_mlp": 0.01030254, + "balance_loss_clip": 1.0474751, + "balance_loss_mlp": 1.02276707, + "epoch": 0.5793302470991403, + "flos": 21470056669440.0, + "grad_norm": 1.7116972545038946, + "language_loss": 0.78421128, + "learning_rate": 1.5862595148876559e-06, + "loss": 0.80648816, + "num_input_tokens_seen": 103945835, + "step": 4818, + "time_per_iteration": 2.7048046588897705 + }, + { + "auxiliary_loss_clip": 0.01211178, + "auxiliary_loss_mlp": 0.01027304, + "balance_loss_clip": 1.04553461, + "balance_loss_mlp": 1.01910853, + "epoch": 0.5794504899897793, + "flos": 12711013367040.0, + "grad_norm": 2.457878824908087, + "language_loss": 0.76279104, + "learning_rate": 1.58549742505406e-06, + "loss": 0.78517592, + "num_input_tokens_seen": 103960580, + "step": 4819, + "time_per_iteration": 2.7765660285949707 + }, + { + "auxiliary_loss_clip": 0.01178369, + "auxiliary_loss_mlp": 0.0102734, + "balance_loss_clip": 1.05169666, + "balance_loss_mlp": 1.01878047, + "epoch": 0.5795707328804185, + "flos": 14867054300160.0, + "grad_norm": 1.955390528661184, + "language_loss": 0.75794625, + "learning_rate": 1.5847353981001747e-06, + "loss": 0.78000331, + "num_input_tokens_seen": 103977760, + "step": 4820, + "time_per_iteration": 3.5173113346099854 + }, + { + "auxiliary_loss_clip": 0.01183482, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.04469347, + "balance_loss_mlp": 1.01927829, + "epoch": 0.5796909757710575, + "flos": 36430046432640.0, + "grad_norm": 1.9036018484228647, + "language_loss": 0.69890356, + "learning_rate": 1.5839734341415993e-06, + "loss": 0.72101229, + "num_input_tokens_seen": 103999960, + "step": 4821, + "time_per_iteration": 3.7130117416381836 + }, + { + "auxiliary_loss_clip": 0.01178894, + "auxiliary_loss_mlp": 0.01024574, + "balance_loss_clip": 1.05147088, + "balance_loss_mlp": 1.01721549, + "epoch": 0.5798112186616966, + "flos": 23039891642880.0, + "grad_norm": 2.467906495536079, + "language_loss": 0.76385826, + "learning_rate": 1.5832115332939238e-06, + "loss": 0.78589296, + "num_input_tokens_seen": 104018400, + "step": 4822, + "time_per_iteration": 3.5508084297180176 + }, + { + "auxiliary_loss_clip": 0.01188115, + "auxiliary_loss_mlp": 0.0102355, + "balance_loss_clip": 1.0508604, + "balance_loss_mlp": 1.01598322, + "epoch": 0.5799314615523358, + "flos": 16652604401280.0, + "grad_norm": 1.6234978556490467, + "language_loss": 0.74938208, + "learning_rate": 1.5824496956727272e-06, + "loss": 0.7714988, + "num_input_tokens_seen": 104035605, + "step": 4823, + "time_per_iteration": 2.6096272468566895 + }, + { + "auxiliary_loss_clip": 0.01194812, + "auxiliary_loss_mlp": 0.0102442, + "balance_loss_clip": 1.04914641, + "balance_loss_mlp": 1.01773787, + "epoch": 0.5800517044429748, + "flos": 20485673470080.0, + "grad_norm": 2.226324518453105, + "language_loss": 0.73417664, + "learning_rate": 1.5816879213935797e-06, + "loss": 0.75636899, + "num_input_tokens_seen": 104054415, + "step": 4824, + "time_per_iteration": 2.721134901046753 + }, + { + "auxiliary_loss_clip": 0.01181411, + "auxiliary_loss_mlp": 0.0102741, + "balance_loss_clip": 1.0504775, + "balance_loss_mlp": 1.01999533, + "epoch": 0.5801719473336139, + "flos": 31538258968320.0, + "grad_norm": 1.5503388870247385, + "language_loss": 0.79654193, + "learning_rate": 1.5809262105720416e-06, + "loss": 0.8186301, + "num_input_tokens_seen": 104075455, + "step": 4825, + "time_per_iteration": 3.640547513961792 + }, + { + "auxiliary_loss_clip": 0.01174401, + "auxiliary_loss_mlp": 0.01025046, + "balance_loss_clip": 1.05147195, + "balance_loss_mlp": 1.01753628, + "epoch": 0.580292190224253, + "flos": 20375966355840.0, + "grad_norm": 1.5330811698691795, + "language_loss": 0.79349494, + "learning_rate": 1.5801645633236644e-06, + "loss": 0.81548941, + "num_input_tokens_seen": 104096440, + "step": 4826, + "time_per_iteration": 2.578209400177002 + }, + { + "auxiliary_loss_clip": 0.01186107, + "auxiliary_loss_mlp": 0.01030607, + "balance_loss_clip": 1.04639518, + "balance_loss_mlp": 1.02310276, + "epoch": 0.5804124331148921, + "flos": 26615373304320.0, + "grad_norm": 2.0401035283760587, + "language_loss": 0.77037597, + "learning_rate": 1.579402979763989e-06, + "loss": 0.79254311, + "num_input_tokens_seen": 104116775, + "step": 4827, + "time_per_iteration": 2.7139346599578857 + }, + { + "auxiliary_loss_clip": 0.01235609, + "auxiliary_loss_mlp": 0.01027302, + "balance_loss_clip": 1.04872441, + "balance_loss_mlp": 1.01989341, + "epoch": 0.5805326760055312, + "flos": 13478496289920.0, + "grad_norm": 2.22041693299378, + "language_loss": 0.81688166, + "learning_rate": 1.578641460008548e-06, + "loss": 0.8395108, + "num_input_tokens_seen": 104134510, + "step": 4828, + "time_per_iteration": 2.71773099899292 + }, + { + "auxiliary_loss_clip": 0.01183943, + "auxiliary_loss_mlp": 0.0102968, + "balance_loss_clip": 1.05019486, + "balance_loss_mlp": 1.022223, + "epoch": 0.5806529188961702, + "flos": 12091374823680.0, + "grad_norm": 2.2612288181952085, + "language_loss": 0.68177164, + "learning_rate": 1.5778800041728613e-06, + "loss": 0.70390785, + "num_input_tokens_seen": 104150800, + "step": 4829, + "time_per_iteration": 2.584672212600708 + }, + { + "auxiliary_loss_clip": 0.01179295, + "auxiliary_loss_mlp": 0.01024194, + "balance_loss_clip": 1.04878092, + "balance_loss_mlp": 1.01710057, + "epoch": 0.5807731617868094, + "flos": 26214107495040.0, + "grad_norm": 1.7343921105485103, + "language_loss": 0.65866363, + "learning_rate": 1.577118612372443e-06, + "loss": 0.68069851, + "num_input_tokens_seen": 104172640, + "step": 4830, + "time_per_iteration": 2.666156530380249 + }, + { + "auxiliary_loss_clip": 0.01189339, + "auxiliary_loss_mlp": 0.01379091, + "balance_loss_clip": 1.04708457, + "balance_loss_mlp": 1.0000912, + "epoch": 0.5808934046774484, + "flos": 37962139190400.0, + "grad_norm": 1.6313122091219616, + "language_loss": 0.70823133, + "learning_rate": 1.5763572847227943e-06, + "loss": 0.73391563, + "num_input_tokens_seen": 104193525, + "step": 4831, + "time_per_iteration": 2.808462619781494 + }, + { + "auxiliary_loss_clip": 0.0118544, + "auxiliary_loss_mlp": 0.01023019, + "balance_loss_clip": 1.04982591, + "balance_loss_mlp": 1.01582432, + "epoch": 0.5810136475680875, + "flos": 20485853038080.0, + "grad_norm": 1.9970361059303787, + "language_loss": 0.81237626, + "learning_rate": 1.5755960213394091e-06, + "loss": 0.83446079, + "num_input_tokens_seen": 104210625, + "step": 4832, + "time_per_iteration": 2.627732753753662 + }, + { + "auxiliary_loss_clip": 0.0120466, + "auxiliary_loss_mlp": 0.01022511, + "balance_loss_clip": 1.04730606, + "balance_loss_mlp": 1.01554012, + "epoch": 0.5811338904587267, + "flos": 17530153574400.0, + "grad_norm": 1.8043146192666475, + "language_loss": 0.78264725, + "learning_rate": 1.5748348223377703e-06, + "loss": 0.804919, + "num_input_tokens_seen": 104228180, + "step": 4833, + "time_per_iteration": 2.7221221923828125 + }, + { + "auxiliary_loss_clip": 0.0119597, + "auxiliary_loss_mlp": 0.01023003, + "balance_loss_clip": 1.05219889, + "balance_loss_mlp": 1.01574874, + "epoch": 0.5812541333493657, + "flos": 19458017360640.0, + "grad_norm": 1.5339914669634982, + "language_loss": 0.7803793, + "learning_rate": 1.5740736878333507e-06, + "loss": 0.80256903, + "num_input_tokens_seen": 104246020, + "step": 4834, + "time_per_iteration": 2.634526491165161 + }, + { + "auxiliary_loss_clip": 0.01195856, + "auxiliary_loss_mlp": 0.01026326, + "balance_loss_clip": 1.04737139, + "balance_loss_mlp": 1.01913524, + "epoch": 0.5813743762400048, + "flos": 20594949621120.0, + "grad_norm": 10.934106729296534, + "language_loss": 0.77676874, + "learning_rate": 1.5733126179416143e-06, + "loss": 0.79899055, + "num_input_tokens_seen": 104260505, + "step": 4835, + "time_per_iteration": 2.736595392227173 + }, + { + "auxiliary_loss_clip": 0.01183438, + "auxiliary_loss_mlp": 0.01020985, + "balance_loss_clip": 1.05025256, + "balance_loss_mlp": 1.01360297, + "epoch": 0.5814946191306439, + "flos": 33178227246720.0, + "grad_norm": 1.9397033120024165, + "language_loss": 0.72699046, + "learning_rate": 1.5725516127780137e-06, + "loss": 0.7490347, + "num_input_tokens_seen": 104282640, + "step": 4836, + "time_per_iteration": 2.751741886138916 + }, + { + "auxiliary_loss_clip": 0.0119304, + "auxiliary_loss_mlp": 0.01029431, + "balance_loss_clip": 1.05050516, + "balance_loss_mlp": 1.02111602, + "epoch": 0.581614862021283, + "flos": 16143283503360.0, + "grad_norm": 2.3221356701370044, + "language_loss": 0.88010895, + "learning_rate": 1.5717906724579943e-06, + "loss": 0.90233362, + "num_input_tokens_seen": 104299700, + "step": 4837, + "time_per_iteration": 2.6494925022125244 + }, + { + "auxiliary_loss_clip": 0.01209916, + "auxiliary_loss_mlp": 0.01028793, + "balance_loss_clip": 1.04798639, + "balance_loss_mlp": 1.02167583, + "epoch": 0.581735104911922, + "flos": 33802642298880.0, + "grad_norm": 2.6813502519702395, + "language_loss": 0.68134302, + "learning_rate": 1.571029797096989e-06, + "loss": 0.70373011, + "num_input_tokens_seen": 104320805, + "step": 4838, + "time_per_iteration": 2.8542933464050293 + }, + { + "auxiliary_loss_clip": 0.01171948, + "auxiliary_loss_mlp": 0.01027865, + "balance_loss_clip": 1.04948366, + "balance_loss_mlp": 1.02028656, + "epoch": 0.5818553478025612, + "flos": 23331163029120.0, + "grad_norm": 1.9638468770582664, + "language_loss": 0.78773975, + "learning_rate": 1.570268986810423e-06, + "loss": 0.80973798, + "num_input_tokens_seen": 104340700, + "step": 4839, + "time_per_iteration": 2.59895396232605 + }, + { + "auxiliary_loss_clip": 0.01190548, + "auxiliary_loss_mlp": 0.01024476, + "balance_loss_clip": 1.04732168, + "balance_loss_mlp": 1.0169065, + "epoch": 0.5819755906932003, + "flos": 20996143603200.0, + "grad_norm": 2.5715140094094315, + "language_loss": 0.74369788, + "learning_rate": 1.5695082417137096e-06, + "loss": 0.76584804, + "num_input_tokens_seen": 104358575, + "step": 4840, + "time_per_iteration": 2.8211193084716797 + }, + { + "auxiliary_loss_clip": 0.01187607, + "auxiliary_loss_mlp": 0.01030814, + "balance_loss_clip": 1.04379702, + "balance_loss_mlp": 1.02409399, + "epoch": 0.5820958335838393, + "flos": 21431668008960.0, + "grad_norm": 2.1888603765485586, + "language_loss": 0.7509467, + "learning_rate": 1.5687475619222539e-06, + "loss": 0.77313089, + "num_input_tokens_seen": 104378530, + "step": 4841, + "time_per_iteration": 2.6931798458099365 + }, + { + "auxiliary_loss_clip": 0.01180984, + "auxiliary_loss_mlp": 0.0102443, + "balance_loss_clip": 1.04337835, + "balance_loss_mlp": 1.01651454, + "epoch": 0.5822160764744785, + "flos": 17967473660160.0, + "grad_norm": 2.776963621413354, + "language_loss": 0.72990531, + "learning_rate": 1.5679869475514496e-06, + "loss": 0.75195944, + "num_input_tokens_seen": 104395465, + "step": 4842, + "time_per_iteration": 2.661700487136841 + }, + { + "auxiliary_loss_clip": 0.01184462, + "auxiliary_loss_mlp": 0.01026038, + "balance_loss_clip": 1.04931712, + "balance_loss_mlp": 1.0187217, + "epoch": 0.5823363193651175, + "flos": 23033858158080.0, + "grad_norm": 2.4806663064898675, + "language_loss": 0.81046623, + "learning_rate": 1.567226398716682e-06, + "loss": 0.83257127, + "num_input_tokens_seen": 104415380, + "step": 4843, + "time_per_iteration": 2.623974084854126 + }, + { + "auxiliary_loss_clip": 0.01201232, + "auxiliary_loss_mlp": 0.01026304, + "balance_loss_clip": 1.04920971, + "balance_loss_mlp": 1.01810288, + "epoch": 0.5824565622557566, + "flos": 32891840110080.0, + "grad_norm": 4.859982576040207, + "language_loss": 0.62006754, + "learning_rate": 1.566465915533326e-06, + "loss": 0.64234293, + "num_input_tokens_seen": 104437410, + "step": 4844, + "time_per_iteration": 2.856506586074829 + }, + { + "auxiliary_loss_clip": 0.01181072, + "auxiliary_loss_mlp": 0.01025409, + "balance_loss_clip": 1.04954839, + "balance_loss_mlp": 1.01784468, + "epoch": 0.5825768051463958, + "flos": 22229674513920.0, + "grad_norm": 1.8419350428303076, + "language_loss": 0.88355112, + "learning_rate": 1.5657054981167458e-06, + "loss": 0.90561593, + "num_input_tokens_seen": 104456305, + "step": 4845, + "time_per_iteration": 2.635943651199341 + }, + { + "auxiliary_loss_clip": 0.01177758, + "auxiliary_loss_mlp": 0.01027414, + "balance_loss_clip": 1.04742932, + "balance_loss_mlp": 1.02017212, + "epoch": 0.5826970480370348, + "flos": 28001561016960.0, + "grad_norm": 1.698160452577204, + "language_loss": 0.67972159, + "learning_rate": 1.5649451465822965e-06, + "loss": 0.70177323, + "num_input_tokens_seen": 104477695, + "step": 4846, + "time_per_iteration": 3.627601146697998 + }, + { + "auxiliary_loss_clip": 0.01210167, + "auxiliary_loss_mlp": 0.01029011, + "balance_loss_clip": 1.0486269, + "balance_loss_mlp": 1.0216316, + "epoch": 0.5828172909276739, + "flos": 17858053854720.0, + "grad_norm": 1.6095921780484277, + "language_loss": 0.83627629, + "learning_rate": 1.5641848610453218e-06, + "loss": 0.85866809, + "num_input_tokens_seen": 104496355, + "step": 4847, + "time_per_iteration": 3.734710216522217 + }, + { + "auxiliary_loss_clip": 0.01181862, + "auxiliary_loss_mlp": 0.01027603, + "balance_loss_clip": 1.05126429, + "balance_loss_mlp": 1.02004528, + "epoch": 0.582937533818313, + "flos": 19865244827520.0, + "grad_norm": 2.072400498922807, + "language_loss": 0.85673451, + "learning_rate": 1.563424641621158e-06, + "loss": 0.87882918, + "num_input_tokens_seen": 104515535, + "step": 4848, + "time_per_iteration": 3.5718181133270264 + }, + { + "auxiliary_loss_clip": 0.01195195, + "auxiliary_loss_mlp": 0.01025315, + "balance_loss_clip": 1.04736865, + "balance_loss_mlp": 1.01736939, + "epoch": 0.5830577767089521, + "flos": 26870734068480.0, + "grad_norm": 1.8441546943370064, + "language_loss": 0.70176154, + "learning_rate": 1.5626644884251282e-06, + "loss": 0.72396672, + "num_input_tokens_seen": 104535055, + "step": 4849, + "time_per_iteration": 2.7213635444641113 + }, + { + "auxiliary_loss_clip": 0.01171318, + "auxiliary_loss_mlp": 0.0102317, + "balance_loss_clip": 1.04900229, + "balance_loss_mlp": 1.01615191, + "epoch": 0.5831780195995911, + "flos": 25298205575040.0, + "grad_norm": 1.5533564572639507, + "language_loss": 0.87817246, + "learning_rate": 1.5619044015725488e-06, + "loss": 0.9001174, + "num_input_tokens_seen": 104554745, + "step": 4850, + "time_per_iteration": 2.660586357116699 + }, + { + "auxiliary_loss_clip": 0.01183438, + "auxiliary_loss_mlp": 0.01024529, + "balance_loss_clip": 1.0559175, + "balance_loss_mlp": 1.01653552, + "epoch": 0.5832982624902303, + "flos": 14756988049920.0, + "grad_norm": 2.2583595271879715, + "language_loss": 0.86968774, + "learning_rate": 1.5611443811787224e-06, + "loss": 0.89176732, + "num_input_tokens_seen": 104568870, + "step": 4851, + "time_per_iteration": 3.477557897567749 + }, + { + "auxiliary_loss_clip": 0.0118104, + "auxiliary_loss_mlp": 0.01025561, + "balance_loss_clip": 1.04948282, + "balance_loss_mlp": 1.0186677, + "epoch": 0.5834185053808694, + "flos": 20444555376000.0, + "grad_norm": 2.235763048429158, + "language_loss": 0.69337231, + "learning_rate": 1.560384427358945e-06, + "loss": 0.71543837, + "num_input_tokens_seen": 104588415, + "step": 4852, + "time_per_iteration": 2.6310524940490723 + }, + { + "auxiliary_loss_clip": 0.01181624, + "auxiliary_loss_mlp": 0.01026488, + "balance_loss_clip": 1.04264009, + "balance_loss_mlp": 1.01930547, + "epoch": 0.5835387482715084, + "flos": 27200394115200.0, + "grad_norm": 1.4467175607997855, + "language_loss": 0.72680473, + "learning_rate": 1.5596245402284998e-06, + "loss": 0.74888587, + "num_input_tokens_seen": 104611940, + "step": 4853, + "time_per_iteration": 2.696119785308838 + }, + { + "auxiliary_loss_clip": 0.01186384, + "auxiliary_loss_mlp": 0.01028206, + "balance_loss_clip": 1.05204558, + "balance_loss_mlp": 1.02068377, + "epoch": 0.5836589911621476, + "flos": 16654615562880.0, + "grad_norm": 1.7414298969092812, + "language_loss": 0.82100284, + "learning_rate": 1.5588647199026619e-06, + "loss": 0.84314871, + "num_input_tokens_seen": 104629675, + "step": 4854, + "time_per_iteration": 2.641765594482422 + }, + { + "auxiliary_loss_clip": 0.01180191, + "auxiliary_loss_mlp": 0.01027529, + "balance_loss_clip": 1.05420887, + "balance_loss_mlp": 1.01958954, + "epoch": 0.5837792340527866, + "flos": 20446817932800.0, + "grad_norm": 2.380193852947612, + "language_loss": 0.87155586, + "learning_rate": 1.5581049664966956e-06, + "loss": 0.89363307, + "num_input_tokens_seen": 104647435, + "step": 4855, + "time_per_iteration": 2.5473251342773438 + }, + { + "auxiliary_loss_clip": 0.01137067, + "auxiliary_loss_mlp": 0.01002209, + "balance_loss_clip": 1.01548326, + "balance_loss_mlp": 1.00074291, + "epoch": 0.5838994769434257, + "flos": 65995480765440.0, + "grad_norm": 0.9851960704627538, + "language_loss": 0.65123177, + "learning_rate": 1.5573452801258545e-06, + "loss": 0.67262459, + "num_input_tokens_seen": 104694605, + "step": 4856, + "time_per_iteration": 3.074672222137451 + }, + { + "auxiliary_loss_clip": 0.01189153, + "auxiliary_loss_mlp": 0.01025083, + "balance_loss_clip": 1.05038548, + "balance_loss_mlp": 1.01756716, + "epoch": 0.5840197198340649, + "flos": 21470523546240.0, + "grad_norm": 1.9948178406110058, + "language_loss": 0.63401371, + "learning_rate": 1.5565856609053824e-06, + "loss": 0.65615606, + "num_input_tokens_seen": 104713400, + "step": 4857, + "time_per_iteration": 2.7106540203094482 + }, + { + "auxiliary_loss_clip": 0.01177501, + "auxiliary_loss_mlp": 0.01024888, + "balance_loss_clip": 1.05163288, + "balance_loss_mlp": 1.01709139, + "epoch": 0.5841399627247039, + "flos": 19135144984320.0, + "grad_norm": 1.784677463079084, + "language_loss": 0.80291796, + "learning_rate": 1.5558261089505127e-06, + "loss": 0.82494181, + "num_input_tokens_seen": 104732130, + "step": 4858, + "time_per_iteration": 2.545693874359131 + }, + { + "auxiliary_loss_clip": 0.011878, + "auxiliary_loss_mlp": 0.01026496, + "balance_loss_clip": 1.05279589, + "balance_loss_mlp": 1.01898623, + "epoch": 0.584260205615343, + "flos": 26425692558720.0, + "grad_norm": 1.885345348569696, + "language_loss": 0.79777807, + "learning_rate": 1.5550666243764697e-06, + "loss": 0.81992102, + "num_input_tokens_seen": 104750290, + "step": 4859, + "time_per_iteration": 2.6878652572631836 + }, + { + "auxiliary_loss_clip": 0.0118813, + "auxiliary_loss_mlp": 0.01028631, + "balance_loss_clip": 1.05249977, + "balance_loss_mlp": 1.02098978, + "epoch": 0.584380448505982, + "flos": 13881809174400.0, + "grad_norm": 2.0524066010976165, + "language_loss": 0.77324939, + "learning_rate": 1.554307207298465e-06, + "loss": 0.79541701, + "num_input_tokens_seen": 104768550, + "step": 4860, + "time_per_iteration": 2.6185011863708496 + }, + { + "auxiliary_loss_clip": 0.01176968, + "auxiliary_loss_mlp": 0.01026922, + "balance_loss_clip": 1.05183721, + "balance_loss_mlp": 1.01884246, + "epoch": 0.5845006913966212, + "flos": 21543709507200.0, + "grad_norm": 2.028196277564857, + "language_loss": 0.78713149, + "learning_rate": 1.553547857831704e-06, + "loss": 0.80917037, + "num_input_tokens_seen": 104785060, + "step": 4861, + "time_per_iteration": 2.6018574237823486 + }, + { + "auxiliary_loss_clip": 0.01072896, + "auxiliary_loss_mlp": 0.01002648, + "balance_loss_clip": 1.01668, + "balance_loss_mlp": 1.00134897, + "epoch": 0.5846209342872603, + "flos": 58375452712320.0, + "grad_norm": 0.8871750242934636, + "language_loss": 0.64215672, + "learning_rate": 1.5527885760913771e-06, + "loss": 0.66291213, + "num_input_tokens_seen": 104834950, + "step": 4862, + "time_per_iteration": 2.9832000732421875 + }, + { + "auxiliary_loss_clip": 0.01193909, + "auxiliary_loss_mlp": 0.01025338, + "balance_loss_clip": 1.05112875, + "balance_loss_mlp": 1.0185883, + "epoch": 0.5847411771778993, + "flos": 18588045957120.0, + "grad_norm": 1.5400927699063411, + "language_loss": 0.76549459, + "learning_rate": 1.552029362192668e-06, + "loss": 0.787687, + "num_input_tokens_seen": 104854210, + "step": 4863, + "time_per_iteration": 2.653881788253784 + }, + { + "auxiliary_loss_clip": 0.01197316, + "auxiliary_loss_mlp": 0.01024133, + "balance_loss_clip": 1.04727793, + "balance_loss_mlp": 1.01677203, + "epoch": 0.5848614200685385, + "flos": 24240780069120.0, + "grad_norm": 2.6239223934619336, + "language_loss": 0.72430557, + "learning_rate": 1.5512702162507478e-06, + "loss": 0.74652004, + "num_input_tokens_seen": 104874525, + "step": 4864, + "time_per_iteration": 2.7079696655273438 + }, + { + "auxiliary_loss_clip": 0.01096932, + "auxiliary_loss_mlp": 0.0100106, + "balance_loss_clip": 1.01543617, + "balance_loss_mlp": 0.99976045, + "epoch": 0.5849816629591775, + "flos": 71660245933440.0, + "grad_norm": 1.1307008950543531, + "language_loss": 0.55755633, + "learning_rate": 1.5505111383807792e-06, + "loss": 0.57853621, + "num_input_tokens_seen": 104937195, + "step": 4865, + "time_per_iteration": 3.2816059589385986 + }, + { + "auxiliary_loss_clip": 0.0120951, + "auxiliary_loss_mlp": 0.01027568, + "balance_loss_clip": 1.04283023, + "balance_loss_mlp": 1.02030861, + "epoch": 0.5851019058498166, + "flos": 23802095266560.0, + "grad_norm": 1.7900652840352809, + "language_loss": 0.80482519, + "learning_rate": 1.5497521286979138e-06, + "loss": 0.82719588, + "num_input_tokens_seen": 104957435, + "step": 4866, + "time_per_iteration": 2.7309088706970215 + }, + { + "auxiliary_loss_clip": 0.01205487, + "auxiliary_loss_mlp": 0.01028943, + "balance_loss_clip": 1.04567051, + "balance_loss_mlp": 1.02062857, + "epoch": 0.5852221487404557, + "flos": 24388516707840.0, + "grad_norm": 1.959504779706754, + "language_loss": 0.74263656, + "learning_rate": 1.5489931873172927e-06, + "loss": 0.76498085, + "num_input_tokens_seen": 104978755, + "step": 4867, + "time_per_iteration": 2.702197313308716 + }, + { + "auxiliary_loss_clip": 0.01197452, + "auxiliary_loss_mlp": 0.0102674, + "balance_loss_clip": 1.03872156, + "balance_loss_mlp": 1.01899791, + "epoch": 0.5853423916310948, + "flos": 27271425260160.0, + "grad_norm": 1.621314166342712, + "language_loss": 0.79208881, + "learning_rate": 1.5482343143540467e-06, + "loss": 0.8143307, + "num_input_tokens_seen": 105000020, + "step": 4868, + "time_per_iteration": 2.8234126567840576 + }, + { + "auxiliary_loss_clip": 0.01203675, + "auxiliary_loss_mlp": 0.01378125, + "balance_loss_clip": 1.04612207, + "balance_loss_mlp": 1.00008988, + "epoch": 0.5854626345217339, + "flos": 11983786611840.0, + "grad_norm": 2.581387551630424, + "language_loss": 0.82619458, + "learning_rate": 1.547475509923295e-06, + "loss": 0.85201257, + "num_input_tokens_seen": 105017060, + "step": 4869, + "time_per_iteration": 2.7942123413085938 + }, + { + "auxiliary_loss_clip": 0.01122571, + "auxiliary_loss_mlp": 0.00999512, + "balance_loss_clip": 1.0148648, + "balance_loss_mlp": 0.99813509, + "epoch": 0.585582877412373, + "flos": 64342335173760.0, + "grad_norm": 0.7256745423376223, + "language_loss": 0.56047952, + "learning_rate": 1.5467167741401495e-06, + "loss": 0.58170033, + "num_input_tokens_seen": 105078540, + "step": 4870, + "time_per_iteration": 3.319235324859619 + }, + { + "auxiliary_loss_clip": 0.01189266, + "auxiliary_loss_mlp": 0.01027398, + "balance_loss_clip": 1.04690504, + "balance_loss_mlp": 1.01912475, + "epoch": 0.5857031203030121, + "flos": 17011926103680.0, + "grad_norm": 2.309740679961388, + "language_loss": 0.71239281, + "learning_rate": 1.5459581071197083e-06, + "loss": 0.73455942, + "num_input_tokens_seen": 105094200, + "step": 4871, + "time_per_iteration": 2.653958559036255 + }, + { + "auxiliary_loss_clip": 0.01188015, + "auxiliary_loss_mlp": 0.01025987, + "balance_loss_clip": 1.05222654, + "balance_loss_mlp": 1.01833379, + "epoch": 0.5858233631936511, + "flos": 20885682303360.0, + "grad_norm": 2.2152507438662816, + "language_loss": 0.83042067, + "learning_rate": 1.5451995089770624e-06, + "loss": 0.8525607, + "num_input_tokens_seen": 105113985, + "step": 4872, + "time_per_iteration": 3.6209311485290527 + }, + { + "auxiliary_loss_clip": 0.01175393, + "auxiliary_loss_mlp": 0.01023281, + "balance_loss_clip": 1.05211985, + "balance_loss_mlp": 1.01674557, + "epoch": 0.5859436060842903, + "flos": 23191902000000.0, + "grad_norm": 1.3507240333272335, + "language_loss": 0.71721065, + "learning_rate": 1.5444409798272885e-06, + "loss": 0.73919737, + "num_input_tokens_seen": 105138075, + "step": 4873, + "time_per_iteration": 3.6690468788146973 + }, + { + "auxiliary_loss_clip": 0.01204605, + "auxiliary_loss_mlp": 0.0102259, + "balance_loss_clip": 1.04791474, + "balance_loss_mlp": 1.01490736, + "epoch": 0.5860638489749294, + "flos": 22492648961280.0, + "grad_norm": 1.9381795754469806, + "language_loss": 0.80582458, + "learning_rate": 1.543682519785456e-06, + "loss": 0.82809657, + "num_input_tokens_seen": 105156555, + "step": 4874, + "time_per_iteration": 3.660700798034668 + }, + { + "auxiliary_loss_clip": 0.01190959, + "auxiliary_loss_mlp": 0.01026173, + "balance_loss_clip": 1.04825926, + "balance_loss_mlp": 1.0182997, + "epoch": 0.5861840918655684, + "flos": 17566243764480.0, + "grad_norm": 3.5240934223445066, + "language_loss": 0.79885304, + "learning_rate": 1.5429241289666219e-06, + "loss": 0.82102436, + "num_input_tokens_seen": 105174055, + "step": 4875, + "time_per_iteration": 2.683790445327759 + }, + { + "auxiliary_loss_clip": 0.01185723, + "auxiliary_loss_mlp": 0.01023676, + "balance_loss_clip": 1.04687119, + "balance_loss_mlp": 1.01631212, + "epoch": 0.5863043347562076, + "flos": 25556152118400.0, + "grad_norm": 1.9734183691470555, + "language_loss": 0.69774848, + "learning_rate": 1.5421658074858342e-06, + "loss": 0.71984243, + "num_input_tokens_seen": 105192160, + "step": 4876, + "time_per_iteration": 2.727687120437622 + }, + { + "auxiliary_loss_clip": 0.01189973, + "auxiliary_loss_mlp": 0.01024136, + "balance_loss_clip": 1.0489701, + "balance_loss_mlp": 1.0163517, + "epoch": 0.5864245776468466, + "flos": 20667525050880.0, + "grad_norm": 2.6476262989959722, + "language_loss": 0.66449606, + "learning_rate": 1.5414075554581298e-06, + "loss": 0.6866371, + "num_input_tokens_seen": 105210205, + "step": 4877, + "time_per_iteration": 3.5154473781585693 + }, + { + "auxiliary_loss_clip": 0.01176179, + "auxiliary_loss_mlp": 0.01028745, + "balance_loss_clip": 1.04960823, + "balance_loss_mlp": 1.02121115, + "epoch": 0.5865448205374857, + "flos": 28913907490560.0, + "grad_norm": 2.635644820410014, + "language_loss": 0.78735292, + "learning_rate": 1.5406493729985348e-06, + "loss": 0.80940211, + "num_input_tokens_seen": 105229400, + "step": 4878, + "time_per_iteration": 2.608513116836548 + }, + { + "auxiliary_loss_clip": 0.0121607, + "auxiliary_loss_mlp": 0.013789, + "balance_loss_clip": 1.04770291, + "balance_loss_mlp": 1.00016236, + "epoch": 0.5866650634281249, + "flos": 25842575168640.0, + "grad_norm": 2.1795970102892066, + "language_loss": 0.71919042, + "learning_rate": 1.5398912602220644e-06, + "loss": 0.74514008, + "num_input_tokens_seen": 105248675, + "step": 4879, + "time_per_iteration": 2.845057487487793 + }, + { + "auxiliary_loss_clip": 0.01222671, + "auxiliary_loss_mlp": 0.01024787, + "balance_loss_clip": 1.04697871, + "balance_loss_mlp": 1.01679444, + "epoch": 0.5867853063187639, + "flos": 17052325925760.0, + "grad_norm": 2.1988013631947982, + "language_loss": 0.78759611, + "learning_rate": 1.539133217243724e-06, + "loss": 0.81007069, + "num_input_tokens_seen": 105265695, + "step": 4880, + "time_per_iteration": 2.69659161567688 + }, + { + "auxiliary_loss_clip": 0.01207074, + "auxiliary_loss_mlp": 0.01028355, + "balance_loss_clip": 1.04693079, + "balance_loss_mlp": 1.02003431, + "epoch": 0.586905549209403, + "flos": 24645026707200.0, + "grad_norm": 2.2659555035145726, + "language_loss": 0.76425028, + "learning_rate": 1.5383752441785081e-06, + "loss": 0.78660458, + "num_input_tokens_seen": 105284920, + "step": 4881, + "time_per_iteration": 2.747098684310913 + }, + { + "auxiliary_loss_clip": 0.01190485, + "auxiliary_loss_mlp": 0.01028848, + "balance_loss_clip": 1.05193508, + "balance_loss_mlp": 1.02056313, + "epoch": 0.5870257921000421, + "flos": 14720538723840.0, + "grad_norm": 2.0409455262955216, + "language_loss": 0.85557038, + "learning_rate": 1.5376173411414003e-06, + "loss": 0.87776375, + "num_input_tokens_seen": 105302960, + "step": 4882, + "time_per_iteration": 2.590705633163452 + }, + { + "auxiliary_loss_clip": 0.01191806, + "auxiliary_loss_mlp": 0.01022306, + "balance_loss_clip": 1.04563951, + "balance_loss_mlp": 1.01470637, + "epoch": 0.5871460349906812, + "flos": 23914998691200.0, + "grad_norm": 1.9887558430285261, + "language_loss": 0.79355538, + "learning_rate": 1.5368595082473753e-06, + "loss": 0.81569654, + "num_input_tokens_seen": 105321260, + "step": 4883, + "time_per_iteration": 2.6891181468963623 + }, + { + "auxiliary_loss_clip": 0.0118651, + "auxiliary_loss_mlp": 0.01024913, + "balance_loss_clip": 1.04945004, + "balance_loss_mlp": 1.01765597, + "epoch": 0.5872662778813202, + "flos": 22164174063360.0, + "grad_norm": 1.7963619600666978, + "language_loss": 0.77854705, + "learning_rate": 1.5361017456113935e-06, + "loss": 0.80066121, + "num_input_tokens_seen": 105341610, + "step": 4884, + "time_per_iteration": 2.615159034729004 + }, + { + "auxiliary_loss_clip": 0.01184743, + "auxiliary_loss_mlp": 0.01028338, + "balance_loss_clip": 1.04885566, + "balance_loss_mlp": 1.02083409, + "epoch": 0.5873865207719594, + "flos": 18441925430400.0, + "grad_norm": 1.9003211409964442, + "language_loss": 0.86006391, + "learning_rate": 1.5353440533484085e-06, + "loss": 0.88219476, + "num_input_tokens_seen": 105360465, + "step": 4885, + "time_per_iteration": 2.7642979621887207 + }, + { + "auxiliary_loss_clip": 0.01195744, + "auxiliary_loss_mlp": 0.01029774, + "balance_loss_clip": 1.04835296, + "balance_loss_mlp": 1.02156591, + "epoch": 0.5875067636625985, + "flos": 54015321427200.0, + "grad_norm": 1.771784593372174, + "language_loss": 0.65918803, + "learning_rate": 1.534586431573361e-06, + "loss": 0.68144321, + "num_input_tokens_seen": 105385405, + "step": 4886, + "time_per_iteration": 2.943300485610962 + }, + { + "auxiliary_loss_clip": 0.01216551, + "auxiliary_loss_mlp": 0.0102936, + "balance_loss_clip": 1.04199529, + "balance_loss_mlp": 1.02108705, + "epoch": 0.5876270065532375, + "flos": 27995707100160.0, + "grad_norm": 2.091257437835265, + "language_loss": 0.79089093, + "learning_rate": 1.5338288804011817e-06, + "loss": 0.81335002, + "num_input_tokens_seen": 105404905, + "step": 4887, + "time_per_iteration": 2.8546030521392822 + }, + { + "auxiliary_loss_clip": 0.01184722, + "auxiliary_loss_mlp": 0.01026729, + "balance_loss_clip": 1.04446745, + "balance_loss_mlp": 1.01877165, + "epoch": 0.5877472494438767, + "flos": 21361462876800.0, + "grad_norm": 1.8815274140704696, + "language_loss": 0.71161306, + "learning_rate": 1.533071399946791e-06, + "loss": 0.73372757, + "num_input_tokens_seen": 105423650, + "step": 4888, + "time_per_iteration": 2.6376094818115234 + }, + { + "auxiliary_loss_clip": 0.01196665, + "auxiliary_loss_mlp": 0.01030165, + "balance_loss_clip": 1.04762363, + "balance_loss_mlp": 1.02277422, + "epoch": 0.5878674923345157, + "flos": 22383013674240.0, + "grad_norm": 1.9323954555474923, + "language_loss": 0.57212341, + "learning_rate": 1.5323139903250977e-06, + "loss": 0.5943917, + "num_input_tokens_seen": 105444255, + "step": 4889, + "time_per_iteration": 2.7016048431396484 + }, + { + "auxiliary_loss_clip": 0.01198742, + "auxiliary_loss_mlp": 0.01025372, + "balance_loss_clip": 1.05222154, + "balance_loss_mlp": 1.01751041, + "epoch": 0.5879877352251548, + "flos": 21868664872320.0, + "grad_norm": 1.4418096988148235, + "language_loss": 0.76725775, + "learning_rate": 1.5315566516510002e-06, + "loss": 0.78949881, + "num_input_tokens_seen": 105462425, + "step": 4890, + "time_per_iteration": 2.6725475788116455 + }, + { + "auxiliary_loss_clip": 0.01177878, + "auxiliary_loss_mlp": 0.01025893, + "balance_loss_clip": 1.05291975, + "balance_loss_mlp": 1.01828122, + "epoch": 0.5881079781157939, + "flos": 17493811989120.0, + "grad_norm": 1.7555598056851252, + "language_loss": 0.67619002, + "learning_rate": 1.5307993840393857e-06, + "loss": 0.69822776, + "num_input_tokens_seen": 105480505, + "step": 4891, + "time_per_iteration": 2.7021844387054443 + }, + { + "auxiliary_loss_clip": 0.01174695, + "auxiliary_loss_mlp": 0.0102649, + "balance_loss_clip": 1.05000997, + "balance_loss_mlp": 1.01895034, + "epoch": 0.588228221006433, + "flos": 22601853285120.0, + "grad_norm": 2.1023521488289805, + "language_loss": 0.80380136, + "learning_rate": 1.530042187605132e-06, + "loss": 0.82581317, + "num_input_tokens_seen": 105499760, + "step": 4892, + "time_per_iteration": 2.6224076747894287 + }, + { + "auxiliary_loss_clip": 0.01185717, + "auxiliary_loss_mlp": 0.0137835, + "balance_loss_clip": 1.05002832, + "balance_loss_mlp": 1.00014091, + "epoch": 0.5883484638970721, + "flos": 26176939896960.0, + "grad_norm": 1.4172563100195699, + "language_loss": 0.83997506, + "learning_rate": 1.5292850624631044e-06, + "loss": 0.86561573, + "num_input_tokens_seen": 105521955, + "step": 4893, + "time_per_iteration": 2.6764559745788574 + }, + { + "auxiliary_loss_clip": 0.01181688, + "auxiliary_loss_mlp": 0.01028983, + "balance_loss_clip": 1.05117702, + "balance_loss_mlp": 1.02137768, + "epoch": 0.5884687067877111, + "flos": 30443737691520.0, + "grad_norm": 3.6489903285640115, + "language_loss": 0.80214965, + "learning_rate": 1.5285280087281593e-06, + "loss": 0.82425636, + "num_input_tokens_seen": 105542685, + "step": 4894, + "time_per_iteration": 2.7156217098236084 + }, + { + "auxiliary_loss_clip": 0.01099804, + "auxiliary_loss_mlp": 0.0100856, + "balance_loss_clip": 1.01625156, + "balance_loss_mlp": 1.00710571, + "epoch": 0.5885889496783503, + "flos": 70507550580480.0, + "grad_norm": 0.6457050545719698, + "language_loss": 0.566661, + "learning_rate": 1.5277710265151398e-06, + "loss": 0.58774465, + "num_input_tokens_seen": 105612165, + "step": 4895, + "time_per_iteration": 3.3958799839019775 + }, + { + "auxiliary_loss_clip": 0.01182282, + "auxiliary_loss_mlp": 0.01031686, + "balance_loss_clip": 1.04840446, + "balance_loss_mlp": 1.02332389, + "epoch": 0.5887091925689893, + "flos": 19098767485440.0, + "grad_norm": 5.625363639929155, + "language_loss": 0.77212441, + "learning_rate": 1.5270141159388803e-06, + "loss": 0.79426408, + "num_input_tokens_seen": 105629185, + "step": 4896, + "time_per_iteration": 2.6420059204101562 + }, + { + "auxiliary_loss_clip": 0.01177385, + "auxiliary_loss_mlp": 0.0102832, + "balance_loss_clip": 1.0512346, + "balance_loss_mlp": 1.02059579, + "epoch": 0.5888294354596284, + "flos": 23294282739840.0, + "grad_norm": 1.9402328019032566, + "language_loss": 0.80402809, + "learning_rate": 1.526257277114203e-06, + "loss": 0.82608515, + "num_input_tokens_seen": 105650260, + "step": 4897, + "time_per_iteration": 2.629640817642212 + }, + { + "auxiliary_loss_clip": 0.01193361, + "auxiliary_loss_mlp": 0.01027911, + "balance_loss_clip": 1.05186093, + "balance_loss_mlp": 1.02053761, + "epoch": 0.5889496783502676, + "flos": 21981532383360.0, + "grad_norm": 1.8454337839092492, + "language_loss": 0.79680955, + "learning_rate": 1.5255005101559201e-06, + "loss": 0.81902224, + "num_input_tokens_seen": 105667870, + "step": 4898, + "time_per_iteration": 3.681647777557373 + }, + { + "auxiliary_loss_clip": 0.0118727, + "auxiliary_loss_mlp": 0.01029176, + "balance_loss_clip": 1.05014777, + "balance_loss_mlp": 1.02172589, + "epoch": 0.5890699212409066, + "flos": 21685233093120.0, + "grad_norm": 2.051714553003266, + "language_loss": 0.76661533, + "learning_rate": 1.524743815178833e-06, + "loss": 0.78877985, + "num_input_tokens_seen": 105685830, + "step": 4899, + "time_per_iteration": 3.5687785148620605 + }, + { + "auxiliary_loss_clip": 0.01195733, + "auxiliary_loss_mlp": 0.01024794, + "balance_loss_clip": 1.04832804, + "balance_loss_mlp": 1.01737309, + "epoch": 0.5891901641315457, + "flos": 19464553635840.0, + "grad_norm": 1.8201472081236096, + "language_loss": 0.80378556, + "learning_rate": 1.5239871922977315e-06, + "loss": 0.82599086, + "num_input_tokens_seen": 105705745, + "step": 4900, + "time_per_iteration": 3.6229820251464844 + }, + { + "auxiliary_loss_clip": 0.01189159, + "auxiliary_loss_mlp": 0.01026676, + "balance_loss_clip": 1.04704785, + "balance_loss_mlp": 1.01889765, + "epoch": 0.5893104070221848, + "flos": 19609884063360.0, + "grad_norm": 2.045084651173101, + "language_loss": 0.89867854, + "learning_rate": 1.523230641627394e-06, + "loss": 0.92083687, + "num_input_tokens_seen": 105724730, + "step": 4901, + "time_per_iteration": 2.718205451965332 + }, + { + "auxiliary_loss_clip": 0.01211787, + "auxiliary_loss_mlp": 0.01031348, + "balance_loss_clip": 1.04318452, + "balance_loss_mlp": 1.02397513, + "epoch": 0.5894306499128239, + "flos": 29060063930880.0, + "grad_norm": 2.0243008161158276, + "language_loss": 0.72757477, + "learning_rate": 1.5224741632825888e-06, + "loss": 0.7500062, + "num_input_tokens_seen": 105744920, + "step": 4902, + "time_per_iteration": 2.8557066917419434 + }, + { + "auxiliary_loss_clip": 0.01179436, + "auxiliary_loss_mlp": 0.01025177, + "balance_loss_clip": 1.0531249, + "balance_loss_mlp": 1.01738691, + "epoch": 0.589550892803463, + "flos": 42298890721920.0, + "grad_norm": 1.7277163448741713, + "language_loss": 0.69215256, + "learning_rate": 1.521717757378074e-06, + "loss": 0.71419871, + "num_input_tokens_seen": 105765465, + "step": 4903, + "time_per_iteration": 3.7000999450683594 + }, + { + "auxiliary_loss_clip": 0.01190805, + "auxiliary_loss_mlp": 0.0102882, + "balance_loss_clip": 1.05115819, + "balance_loss_mlp": 1.02077293, + "epoch": 0.5896711356941021, + "flos": 14137062197760.0, + "grad_norm": 2.4054178159222976, + "language_loss": 0.68952024, + "learning_rate": 1.5209614240285943e-06, + "loss": 0.71171641, + "num_input_tokens_seen": 105783120, + "step": 4904, + "time_per_iteration": 2.690521240234375 + }, + { + "auxiliary_loss_clip": 0.01176831, + "auxiliary_loss_mlp": 0.0137865, + "balance_loss_clip": 1.05087662, + "balance_loss_mlp": 1.0001384, + "epoch": 0.5897913785847412, + "flos": 17201355454080.0, + "grad_norm": 2.3792163693880273, + "language_loss": 0.84814882, + "learning_rate": 1.520205163348887e-06, + "loss": 0.8737036, + "num_input_tokens_seen": 105801055, + "step": 4905, + "time_per_iteration": 2.672571897506714 + }, + { + "auxiliary_loss_clip": 0.01114343, + "auxiliary_loss_mlp": 0.01001668, + "balance_loss_clip": 1.01712489, + "balance_loss_mlp": 1.00023735, + "epoch": 0.5899116214753802, + "flos": 48794164202880.0, + "grad_norm": 0.731274131262861, + "language_loss": 0.56997871, + "learning_rate": 1.519448975453674e-06, + "loss": 0.59113884, + "num_input_tokens_seen": 105856155, + "step": 4906, + "time_per_iteration": 3.15983510017395 + }, + { + "auxiliary_loss_clip": 0.0118677, + "auxiliary_loss_mlp": 0.01378709, + "balance_loss_clip": 1.05337274, + "balance_loss_mlp": 1.00014687, + "epoch": 0.5900318643660194, + "flos": 21103659987840.0, + "grad_norm": 1.9708159692812883, + "language_loss": 0.76182365, + "learning_rate": 1.5186928604576696e-06, + "loss": 0.78747845, + "num_input_tokens_seen": 105873350, + "step": 4907, + "time_per_iteration": 2.635859489440918 + }, + { + "auxiliary_loss_clip": 0.01194334, + "auxiliary_loss_mlp": 0.01026647, + "balance_loss_clip": 1.04806352, + "balance_loss_mlp": 1.01932788, + "epoch": 0.5901521072566585, + "flos": 21178390233600.0, + "grad_norm": 2.4805039860370184, + "language_loss": 0.77229559, + "learning_rate": 1.5179368184755752e-06, + "loss": 0.79450542, + "num_input_tokens_seen": 105891435, + "step": 4908, + "time_per_iteration": 2.622663974761963 + }, + { + "auxiliary_loss_clip": 0.01195545, + "auxiliary_loss_mlp": 0.01028553, + "balance_loss_clip": 1.05037999, + "balance_loss_mlp": 1.02109623, + "epoch": 0.5902723501472975, + "flos": 20225967160320.0, + "grad_norm": 1.4838097798382128, + "language_loss": 0.82575685, + "learning_rate": 1.5171808496220821e-06, + "loss": 0.84799784, + "num_input_tokens_seen": 105910190, + "step": 4909, + "time_per_iteration": 2.697227716445923 + }, + { + "auxiliary_loss_clip": 0.01197827, + "auxiliary_loss_mlp": 0.01023124, + "balance_loss_clip": 1.04835296, + "balance_loss_mlp": 1.01572144, + "epoch": 0.5903925930379367, + "flos": 22964407211520.0, + "grad_norm": 1.6827251867598345, + "language_loss": 0.81294411, + "learning_rate": 1.5164249540118708e-06, + "loss": 0.83515364, + "num_input_tokens_seen": 105929315, + "step": 4910, + "time_per_iteration": 2.65260648727417 + }, + { + "auxiliary_loss_clip": 0.01224245, + "auxiliary_loss_mlp": 0.01024192, + "balance_loss_clip": 1.04361892, + "balance_loss_mlp": 1.01665211, + "epoch": 0.5905128359285757, + "flos": 23367720096000.0, + "grad_norm": 2.0533547894352924, + "language_loss": 0.83204377, + "learning_rate": 1.5156691317596093e-06, + "loss": 0.85452819, + "num_input_tokens_seen": 105950740, + "step": 4911, + "time_per_iteration": 2.8266823291778564 + }, + { + "auxiliary_loss_clip": 0.01187963, + "auxiliary_loss_mlp": 0.01378419, + "balance_loss_clip": 1.05176234, + "balance_loss_mlp": 1.00024402, + "epoch": 0.5906330788192148, + "flos": 28032335994240.0, + "grad_norm": 2.1232674171044166, + "language_loss": 0.66607964, + "learning_rate": 1.5149133829799556e-06, + "loss": 0.69174343, + "num_input_tokens_seen": 105968735, + "step": 4912, + "time_per_iteration": 2.697878837585449 + }, + { + "auxiliary_loss_clip": 0.01201597, + "auxiliary_loss_mlp": 0.01026171, + "balance_loss_clip": 1.04944229, + "balance_loss_mlp": 1.0184288, + "epoch": 0.590753321709854, + "flos": 18477943793280.0, + "grad_norm": 1.9250629534143096, + "language_loss": 0.80793506, + "learning_rate": 1.5141577077875556e-06, + "loss": 0.83021277, + "num_input_tokens_seen": 105986060, + "step": 4913, + "time_per_iteration": 2.6867871284484863 + }, + { + "auxiliary_loss_clip": 0.01186116, + "auxiliary_loss_mlp": 0.01024253, + "balance_loss_clip": 1.04870343, + "balance_loss_mlp": 1.0171783, + "epoch": 0.590873564600493, + "flos": 16873706568960.0, + "grad_norm": 2.506021661640814, + "language_loss": 0.72304398, + "learning_rate": 1.5134021062970451e-06, + "loss": 0.74514771, + "num_input_tokens_seen": 106004440, + "step": 4914, + "time_per_iteration": 2.646012306213379 + }, + { + "auxiliary_loss_clip": 0.01194183, + "auxiliary_loss_mlp": 0.01025106, + "balance_loss_clip": 1.04883242, + "balance_loss_mlp": 1.0174973, + "epoch": 0.5909938074911321, + "flos": 13516166678400.0, + "grad_norm": 1.7247887702536269, + "language_loss": 0.811885, + "learning_rate": 1.5126465786230483e-06, + "loss": 0.83407795, + "num_input_tokens_seen": 106021215, + "step": 4915, + "time_per_iteration": 2.7106754779815674 + }, + { + "auxiliary_loss_clip": 0.01175574, + "auxiliary_loss_mlp": 0.01028616, + "balance_loss_clip": 1.05162334, + "balance_loss_mlp": 1.02112055, + "epoch": 0.5911140503817712, + "flos": 26024067613440.0, + "grad_norm": 1.8536066768682917, + "language_loss": 0.82010758, + "learning_rate": 1.5118911248801787e-06, + "loss": 0.84214944, + "num_input_tokens_seen": 106039225, + "step": 4916, + "time_per_iteration": 2.6112852096557617 + }, + { + "auxiliary_loss_clip": 0.01179843, + "auxiliary_loss_mlp": 0.01027233, + "balance_loss_clip": 1.04915059, + "balance_loss_mlp": 1.01996744, + "epoch": 0.5912342932724103, + "flos": 23258731253760.0, + "grad_norm": 1.8707427955851075, + "language_loss": 0.79714215, + "learning_rate": 1.5111357451830364e-06, + "loss": 0.81921291, + "num_input_tokens_seen": 106057920, + "step": 4917, + "time_per_iteration": 2.679396390914917 + }, + { + "auxiliary_loss_clip": 0.01187302, + "auxiliary_loss_mlp": 0.01021973, + "balance_loss_clip": 1.05167532, + "balance_loss_mlp": 1.01449561, + "epoch": 0.5913545361630493, + "flos": 19573039687680.0, + "grad_norm": 1.8198581783248053, + "language_loss": 0.71079659, + "learning_rate": 1.5103804396462131e-06, + "loss": 0.73288935, + "num_input_tokens_seen": 106077855, + "step": 4918, + "time_per_iteration": 2.602557420730591 + }, + { + "auxiliary_loss_clip": 0.01188034, + "auxiliary_loss_mlp": 0.01025427, + "balance_loss_clip": 1.04811454, + "balance_loss_mlp": 1.01751757, + "epoch": 0.5914747790536885, + "flos": 26213532877440.0, + "grad_norm": 2.085421431243963, + "language_loss": 0.79756176, + "learning_rate": 1.5096252083842877e-06, + "loss": 0.81969631, + "num_input_tokens_seen": 106097065, + "step": 4919, + "time_per_iteration": 2.7491540908813477 + }, + { + "auxiliary_loss_clip": 0.01179896, + "auxiliary_loss_mlp": 0.01025159, + "balance_loss_clip": 1.04664731, + "balance_loss_mlp": 1.01723123, + "epoch": 0.5915950219443276, + "flos": 27417545786880.0, + "grad_norm": 1.8056062335031433, + "language_loss": 0.85323215, + "learning_rate": 1.5088700515118285e-06, + "loss": 0.8752827, + "num_input_tokens_seen": 106116385, + "step": 4920, + "time_per_iteration": 2.7221648693084717 + }, + { + "auxiliary_loss_clip": 0.01195895, + "auxiliary_loss_mlp": 0.01026518, + "balance_loss_clip": 1.0457629, + "balance_loss_mlp": 1.0189364, + "epoch": 0.5917152648349666, + "flos": 21907879545600.0, + "grad_norm": 2.9761244896402532, + "language_loss": 0.6661762, + "learning_rate": 1.508114969143392e-06, + "loss": 0.68840033, + "num_input_tokens_seen": 106136370, + "step": 4921, + "time_per_iteration": 2.6574547290802 + }, + { + "auxiliary_loss_clip": 0.01191669, + "auxiliary_loss_mlp": 0.01025216, + "balance_loss_clip": 1.04552698, + "balance_loss_mlp": 1.01799452, + "epoch": 0.5918355077256057, + "flos": 28109185142400.0, + "grad_norm": 1.5456468858900456, + "language_loss": 0.77510965, + "learning_rate": 1.5073599613935238e-06, + "loss": 0.79727846, + "num_input_tokens_seen": 106158490, + "step": 4922, + "time_per_iteration": 2.7766385078430176 + }, + { + "auxiliary_loss_clip": 0.01195387, + "auxiliary_loss_mlp": 0.01022398, + "balance_loss_clip": 1.04959273, + "balance_loss_mlp": 1.01487947, + "epoch": 0.5919557506162448, + "flos": 28183807647360.0, + "grad_norm": 2.0489057000565425, + "language_loss": 0.57793808, + "learning_rate": 1.5066050283767574e-06, + "loss": 0.6001159, + "num_input_tokens_seen": 106179170, + "step": 4923, + "time_per_iteration": 2.689788341522217 + }, + { + "auxiliary_loss_clip": 0.01187449, + "auxiliary_loss_mlp": 0.01022478, + "balance_loss_clip": 1.04828644, + "balance_loss_mlp": 1.01490784, + "epoch": 0.5920759935068839, + "flos": 12094355652480.0, + "grad_norm": 2.359012502522417, + "language_loss": 0.82661754, + "learning_rate": 1.505850170207616e-06, + "loss": 0.84871674, + "num_input_tokens_seen": 106196035, + "step": 4924, + "time_per_iteration": 3.694469451904297 + }, + { + "auxiliary_loss_clip": 0.01191543, + "auxiliary_loss_mlp": 0.01024079, + "balance_loss_clip": 1.04726267, + "balance_loss_mlp": 1.01627982, + "epoch": 0.592196236397523, + "flos": 29424772673280.0, + "grad_norm": 2.141980585526547, + "language_loss": 0.7792725, + "learning_rate": 1.505095387000611e-06, + "loss": 0.80142868, + "num_input_tokens_seen": 106218335, + "step": 4925, + "time_per_iteration": 3.6661510467529297 + }, + { + "auxiliary_loss_clip": 0.01181858, + "auxiliary_loss_mlp": 0.01028898, + "balance_loss_clip": 1.04820502, + "balance_loss_mlp": 1.02176964, + "epoch": 0.5923164792881621, + "flos": 24384709866240.0, + "grad_norm": 2.503536713159901, + "language_loss": 0.74326533, + "learning_rate": 1.504340678870242e-06, + "loss": 0.76537287, + "num_input_tokens_seen": 106236550, + "step": 4926, + "time_per_iteration": 3.6221871376037598 + }, + { + "auxiliary_loss_clip": 0.01183781, + "auxiliary_loss_mlp": 0.01027784, + "balance_loss_clip": 1.0503118, + "balance_loss_mlp": 1.01934981, + "epoch": 0.5924367221788012, + "flos": 24024238928640.0, + "grad_norm": 1.8559038100490606, + "language_loss": 0.89538884, + "learning_rate": 1.5035860459309989e-06, + "loss": 0.91750455, + "num_input_tokens_seen": 106254265, + "step": 4927, + "time_per_iteration": 2.6506073474884033 + }, + { + "auxiliary_loss_clip": 0.01187171, + "auxiliary_loss_mlp": 0.0103172, + "balance_loss_clip": 1.04660034, + "balance_loss_mlp": 1.02361941, + "epoch": 0.5925569650694402, + "flos": 26870590414080.0, + "grad_norm": 1.9578817457569428, + "language_loss": 0.63521749, + "learning_rate": 1.5028314882973568e-06, + "loss": 0.65740639, + "num_input_tokens_seen": 106274670, + "step": 4928, + "time_per_iteration": 2.7096340656280518 + }, + { + "auxiliary_loss_clip": 0.01196207, + "auxiliary_loss_mlp": 0.01030125, + "balance_loss_clip": 1.04970407, + "balance_loss_mlp": 1.02229333, + "epoch": 0.5926772079600794, + "flos": 22302788647680.0, + "grad_norm": 1.9379983239193972, + "language_loss": 0.84413505, + "learning_rate": 1.502077006083783e-06, + "loss": 0.86639833, + "num_input_tokens_seen": 106293330, + "step": 4929, + "time_per_iteration": 3.6024558544158936 + }, + { + "auxiliary_loss_clip": 0.01187047, + "auxiliary_loss_mlp": 0.01378586, + "balance_loss_clip": 1.04955006, + "balance_loss_mlp": 1.00014472, + "epoch": 0.5927974508507184, + "flos": 19865244827520.0, + "grad_norm": 1.808965904634993, + "language_loss": 0.76590735, + "learning_rate": 1.5013225994047315e-06, + "loss": 0.79156363, + "num_input_tokens_seen": 106310960, + "step": 4930, + "time_per_iteration": 2.6454505920410156 + }, + { + "auxiliary_loss_clip": 0.01186902, + "auxiliary_loss_mlp": 0.01378378, + "balance_loss_clip": 1.05079174, + "balance_loss_mlp": 1.00013328, + "epoch": 0.5929176937413575, + "flos": 15776743167360.0, + "grad_norm": 1.8327825271780886, + "language_loss": 0.80735147, + "learning_rate": 1.5005682683746452e-06, + "loss": 0.8330043, + "num_input_tokens_seen": 106329475, + "step": 4931, + "time_per_iteration": 2.6751868724823 + }, + { + "auxiliary_loss_clip": 0.01185119, + "auxiliary_loss_mlp": 0.01027683, + "balance_loss_clip": 1.05134249, + "balance_loss_mlp": 1.01929069, + "epoch": 0.5930379366319967, + "flos": 17601472028160.0, + "grad_norm": 2.336764277571825, + "language_loss": 0.72853792, + "learning_rate": 1.4998140131079553e-06, + "loss": 0.7506659, + "num_input_tokens_seen": 106345565, + "step": 4932, + "time_per_iteration": 2.613007068634033 + }, + { + "auxiliary_loss_clip": 0.01216232, + "auxiliary_loss_mlp": 0.01378371, + "balance_loss_clip": 1.0431664, + "balance_loss_mlp": 1.00006509, + "epoch": 0.5931581795226357, + "flos": 17704283731200.0, + "grad_norm": 1.7711028938899744, + "language_loss": 0.73536289, + "learning_rate": 1.4990598337190821e-06, + "loss": 0.76130897, + "num_input_tokens_seen": 106361920, + "step": 4933, + "time_per_iteration": 2.7690353393554688 + }, + { + "auxiliary_loss_clip": 0.01177014, + "auxiliary_loss_mlp": 0.01378738, + "balance_loss_clip": 1.05203342, + "balance_loss_mlp": 1.00011992, + "epoch": 0.5932784224132748, + "flos": 24280102483200.0, + "grad_norm": 1.738368904686563, + "language_loss": 0.67919648, + "learning_rate": 1.4983057303224338e-06, + "loss": 0.70475399, + "num_input_tokens_seen": 106381735, + "step": 4934, + "time_per_iteration": 2.680751085281372 + }, + { + "auxiliary_loss_clip": 0.01205369, + "auxiliary_loss_mlp": 0.0102578, + "balance_loss_clip": 1.04368544, + "balance_loss_mlp": 1.01837158, + "epoch": 0.5933986653039139, + "flos": 22926700909440.0, + "grad_norm": 1.6723781393063344, + "language_loss": 0.87643582, + "learning_rate": 1.4975517030324072e-06, + "loss": 0.89874732, + "num_input_tokens_seen": 106399745, + "step": 4935, + "time_per_iteration": 2.7631072998046875 + }, + { + "auxiliary_loss_clip": 0.01068362, + "auxiliary_loss_mlp": 0.01374334, + "balance_loss_clip": 1.01253009, + "balance_loss_mlp": 0.99988014, + "epoch": 0.593518908194553, + "flos": 71121730256640.0, + "grad_norm": 0.78279372825631, + "language_loss": 0.61761278, + "learning_rate": 1.4967977519633882e-06, + "loss": 0.64203972, + "num_input_tokens_seen": 106457205, + "step": 4936, + "time_per_iteration": 3.2491958141326904 + }, + { + "auxiliary_loss_clip": 0.01199285, + "auxiliary_loss_mlp": 0.01028808, + "balance_loss_clip": 1.04697394, + "balance_loss_mlp": 1.02085078, + "epoch": 0.593639151085192, + "flos": 20448649526400.0, + "grad_norm": 3.3751801829811328, + "language_loss": 0.78039372, + "learning_rate": 1.4960438772297494e-06, + "loss": 0.80267465, + "num_input_tokens_seen": 106474250, + "step": 4937, + "time_per_iteration": 2.676424503326416 + }, + { + "auxiliary_loss_clip": 0.01194379, + "auxiliary_loss_mlp": 0.01029021, + "balance_loss_clip": 1.04729867, + "balance_loss_mlp": 1.02081394, + "epoch": 0.5937593939758312, + "flos": 30883428074880.0, + "grad_norm": 1.8831136187573094, + "language_loss": 0.72876388, + "learning_rate": 1.495290078945855e-06, + "loss": 0.75099784, + "num_input_tokens_seen": 106494015, + "step": 4938, + "time_per_iteration": 2.6961188316345215 + }, + { + "auxiliary_loss_clip": 0.01174754, + "auxiliary_loss_mlp": 0.01026425, + "balance_loss_clip": 1.05154204, + "balance_loss_mlp": 1.01877797, + "epoch": 0.5938796368664703, + "flos": 36898069668480.0, + "grad_norm": 2.26264486625646, + "language_loss": 0.74348342, + "learning_rate": 1.4945363572260529e-06, + "loss": 0.76549524, + "num_input_tokens_seen": 106515010, + "step": 4939, + "time_per_iteration": 2.7259767055511475 + }, + { + "auxiliary_loss_clip": 0.01186055, + "auxiliary_loss_mlp": 0.01025159, + "balance_loss_clip": 1.04985154, + "balance_loss_mlp": 1.01756573, + "epoch": 0.5939998797571093, + "flos": 23842926051840.0, + "grad_norm": 2.1380775251193427, + "language_loss": 0.67724371, + "learning_rate": 1.4937827121846845e-06, + "loss": 0.69935584, + "num_input_tokens_seen": 106535265, + "step": 4940, + "time_per_iteration": 2.6376359462738037 + }, + { + "auxiliary_loss_clip": 0.01195721, + "auxiliary_loss_mlp": 0.0102853, + "balance_loss_clip": 1.04708171, + "balance_loss_mlp": 1.02090049, + "epoch": 0.5941201226477485, + "flos": 25191407462400.0, + "grad_norm": 1.581117831516465, + "language_loss": 0.73640811, + "learning_rate": 1.4930291439360755e-06, + "loss": 0.75865066, + "num_input_tokens_seen": 106557830, + "step": 4941, + "time_per_iteration": 2.7309670448303223 + }, + { + "auxiliary_loss_clip": 0.01186252, + "auxiliary_loss_mlp": 0.01030894, + "balance_loss_clip": 1.05113053, + "balance_loss_mlp": 1.02278829, + "epoch": 0.5942403655383875, + "flos": 22418996123520.0, + "grad_norm": 2.0060183565134695, + "language_loss": 0.79081923, + "learning_rate": 1.4922756525945427e-06, + "loss": 0.81299067, + "num_input_tokens_seen": 106577140, + "step": 4942, + "time_per_iteration": 2.6206893920898438 + }, + { + "auxiliary_loss_clip": 0.01082247, + "auxiliary_loss_mlp": 0.01004014, + "balance_loss_clip": 1.01278865, + "balance_loss_mlp": 1.00257158, + "epoch": 0.5943606084290266, + "flos": 67629310796160.0, + "grad_norm": 0.7724082646404062, + "language_loss": 0.59526819, + "learning_rate": 1.4915222382743894e-06, + "loss": 0.61613083, + "num_input_tokens_seen": 106635975, + "step": 4943, + "time_per_iteration": 3.257869243621826 + }, + { + "auxiliary_loss_clip": 0.01187198, + "auxiliary_loss_mlp": 0.01029306, + "balance_loss_clip": 1.05104911, + "balance_loss_mlp": 1.02085388, + "epoch": 0.5944808513196658, + "flos": 18223157646720.0, + "grad_norm": 2.2979591761826015, + "language_loss": 0.72285295, + "learning_rate": 1.4907689010899085e-06, + "loss": 0.74501801, + "num_input_tokens_seen": 106653555, + "step": 4944, + "time_per_iteration": 2.600130558013916 + }, + { + "auxiliary_loss_clip": 0.01196433, + "auxiliary_loss_mlp": 0.01024288, + "balance_loss_clip": 1.04961205, + "balance_loss_mlp": 1.01662016, + "epoch": 0.5946010942103048, + "flos": 24790824011520.0, + "grad_norm": 2.1323163306986417, + "language_loss": 0.6209963, + "learning_rate": 1.4900156411553804e-06, + "loss": 0.6432035, + "num_input_tokens_seen": 106673385, + "step": 4945, + "time_per_iteration": 2.673372507095337 + }, + { + "auxiliary_loss_clip": 0.01196101, + "auxiliary_loss_mlp": 0.01029471, + "balance_loss_clip": 1.05075598, + "balance_loss_mlp": 1.02168655, + "epoch": 0.5947213371009439, + "flos": 15231619388160.0, + "grad_norm": 2.0934670235317365, + "language_loss": 0.85470086, + "learning_rate": 1.4892624585850739e-06, + "loss": 0.87695658, + "num_input_tokens_seen": 106691740, + "step": 4946, + "time_per_iteration": 2.680891752243042 + }, + { + "auxiliary_loss_clip": 0.01179318, + "auxiliary_loss_mlp": 0.0102887, + "balance_loss_clip": 1.05199027, + "balance_loss_mlp": 1.02071643, + "epoch": 0.594841579991583, + "flos": 25848069949440.0, + "grad_norm": 2.0507872259231834, + "language_loss": 0.79228449, + "learning_rate": 1.4885093534932465e-06, + "loss": 0.81436646, + "num_input_tokens_seen": 106709705, + "step": 4947, + "time_per_iteration": 2.66162109375 + }, + { + "auxiliary_loss_clip": 0.01192197, + "auxiliary_loss_mlp": 0.01026688, + "balance_loss_clip": 1.0496099, + "balance_loss_mlp": 1.01840889, + "epoch": 0.5949618228822221, + "flos": 23981109672960.0, + "grad_norm": 1.9521854011519022, + "language_loss": 0.71054214, + "learning_rate": 1.4877563259941433e-06, + "loss": 0.73273098, + "num_input_tokens_seen": 106727560, + "step": 4948, + "time_per_iteration": 2.749681234359741 + }, + { + "auxiliary_loss_clip": 0.01195304, + "auxiliary_loss_mlp": 0.01027337, + "balance_loss_clip": 1.05319428, + "balance_loss_mlp": 1.0191946, + "epoch": 0.5950820657728612, + "flos": 40547491476480.0, + "grad_norm": 2.055152455358312, + "language_loss": 0.67698646, + "learning_rate": 1.4870033762019988e-06, + "loss": 0.69921291, + "num_input_tokens_seen": 106747725, + "step": 4949, + "time_per_iteration": 2.8032073974609375 + }, + { + "auxiliary_loss_clip": 0.01192383, + "auxiliary_loss_mlp": 0.01032518, + "balance_loss_clip": 1.04897773, + "balance_loss_mlp": 1.0244832, + "epoch": 0.5952023086635003, + "flos": 23184467884800.0, + "grad_norm": 1.779198645749974, + "language_loss": 0.73400307, + "learning_rate": 1.4862505042310334e-06, + "loss": 0.75625211, + "num_input_tokens_seen": 106767010, + "step": 4950, + "time_per_iteration": 3.5437233448028564 + }, + { + "auxiliary_loss_clip": 0.01187184, + "auxiliary_loss_mlp": 0.01026741, + "balance_loss_clip": 1.04818833, + "balance_loss_mlp": 1.01911139, + "epoch": 0.5953225515541394, + "flos": 33653289548160.0, + "grad_norm": 1.5944752382819432, + "language_loss": 0.69597536, + "learning_rate": 1.4854977101954587e-06, + "loss": 0.71811461, + "num_input_tokens_seen": 106789230, + "step": 4951, + "time_per_iteration": 2.760876178741455 + }, + { + "auxiliary_loss_clip": 0.01186158, + "auxiliary_loss_mlp": 0.01025886, + "balance_loss_clip": 1.04764116, + "balance_loss_mlp": 1.01820898, + "epoch": 0.5954427944447784, + "flos": 24459619680000.0, + "grad_norm": 1.786852365267225, + "language_loss": 0.86193728, + "learning_rate": 1.4847449942094716e-06, + "loss": 0.88405776, + "num_input_tokens_seen": 106808110, + "step": 4952, + "time_per_iteration": 3.5106866359710693 + }, + { + "auxiliary_loss_clip": 0.01191121, + "auxiliary_loss_mlp": 0.01028126, + "balance_loss_clip": 1.04876757, + "balance_loss_mlp": 1.02045441, + "epoch": 0.5955630373354175, + "flos": 18551848026240.0, + "grad_norm": 2.2562916789638963, + "language_loss": 0.86029422, + "learning_rate": 1.4839923563872598e-06, + "loss": 0.8824867, + "num_input_tokens_seen": 106826650, + "step": 4953, + "time_per_iteration": 2.644876003265381 + }, + { + "auxiliary_loss_clip": 0.01205243, + "auxiliary_loss_mlp": 0.01029216, + "balance_loss_clip": 1.04900765, + "balance_loss_mlp": 1.0216105, + "epoch": 0.5956832802260567, + "flos": 19791699730560.0, + "grad_norm": 1.7874333283782033, + "language_loss": 0.7609573, + "learning_rate": 1.483239796842997e-06, + "loss": 0.78330183, + "num_input_tokens_seen": 106844680, + "step": 4954, + "time_per_iteration": 2.6679441928863525 + }, + { + "auxiliary_loss_clip": 0.0120359, + "auxiliary_loss_mlp": 0.01028765, + "balance_loss_clip": 1.0492475, + "balance_loss_mlp": 1.02091789, + "epoch": 0.5958035231166957, + "flos": 19750868945280.0, + "grad_norm": 1.8491774888263373, + "language_loss": 0.83701026, + "learning_rate": 1.4824873156908462e-06, + "loss": 0.85933381, + "num_input_tokens_seen": 106862605, + "step": 4955, + "time_per_iteration": 3.6189894676208496 + }, + { + "auxiliary_loss_clip": 0.01187103, + "auxiliary_loss_mlp": 0.0137949, + "balance_loss_clip": 1.0514425, + "balance_loss_mlp": 1.00020742, + "epoch": 0.5959237660073348, + "flos": 21652806090240.0, + "grad_norm": 1.5419984196435363, + "language_loss": 0.75614119, + "learning_rate": 1.4817349130449584e-06, + "loss": 0.78180707, + "num_input_tokens_seen": 106882325, + "step": 4956, + "time_per_iteration": 2.6676559448242188 + }, + { + "auxiliary_loss_clip": 0.0118312, + "auxiliary_loss_mlp": 0.01027556, + "balance_loss_clip": 1.05032754, + "balance_loss_mlp": 1.01908565, + "epoch": 0.5960440088979739, + "flos": 21171207513600.0, + "grad_norm": 1.9969604419787508, + "language_loss": 0.83067173, + "learning_rate": 1.4809825890194717e-06, + "loss": 0.85277843, + "num_input_tokens_seen": 106900995, + "step": 4957, + "time_per_iteration": 2.640886068344116 + }, + { + "auxiliary_loss_clip": 0.01186066, + "auxiliary_loss_mlp": 0.0102794, + "balance_loss_clip": 1.04496861, + "balance_loss_mlp": 1.02081132, + "epoch": 0.596164251788613, + "flos": 14757526753920.0, + "grad_norm": 1.792782091736877, + "language_loss": 0.77573669, + "learning_rate": 1.4802303437285139e-06, + "loss": 0.79787672, + "num_input_tokens_seen": 106918265, + "step": 4958, + "time_per_iteration": 2.6674225330352783 + }, + { + "auxiliary_loss_clip": 0.0119133, + "auxiliary_loss_mlp": 0.01027231, + "balance_loss_clip": 1.04708743, + "balance_loss_mlp": 1.01890969, + "epoch": 0.596284494679252, + "flos": 20485924865280.0, + "grad_norm": 2.254089008559132, + "language_loss": 0.80801541, + "learning_rate": 1.4794781772861994e-06, + "loss": 0.83020097, + "num_input_tokens_seen": 106934760, + "step": 4959, + "time_per_iteration": 2.675511598587036 + }, + { + "auxiliary_loss_clip": 0.01191537, + "auxiliary_loss_mlp": 0.0137879, + "balance_loss_clip": 1.04744816, + "balance_loss_mlp": 1.00019956, + "epoch": 0.5964047375698912, + "flos": 31212262108800.0, + "grad_norm": 2.3626037499715613, + "language_loss": 0.66934025, + "learning_rate": 1.4787260898066324e-06, + "loss": 0.6950435, + "num_input_tokens_seen": 106954760, + "step": 4960, + "time_per_iteration": 2.761446475982666 + }, + { + "auxiliary_loss_clip": 0.01174327, + "auxiliary_loss_mlp": 0.01025553, + "balance_loss_clip": 1.05086935, + "balance_loss_mlp": 1.01808786, + "epoch": 0.5965249804605303, + "flos": 27483620855040.0, + "grad_norm": 1.8801936429276753, + "language_loss": 0.85282743, + "learning_rate": 1.4779740814039023e-06, + "loss": 0.87482625, + "num_input_tokens_seen": 106974845, + "step": 4961, + "time_per_iteration": 2.680574417114258 + }, + { + "auxiliary_loss_clip": 0.01176641, + "auxiliary_loss_mlp": 0.01022406, + "balance_loss_clip": 1.05142307, + "balance_loss_mlp": 1.01497293, + "epoch": 0.5966452233511693, + "flos": 30773936442240.0, + "grad_norm": 7.204208544942799, + "language_loss": 0.68579435, + "learning_rate": 1.4772221521920894e-06, + "loss": 0.70778477, + "num_input_tokens_seen": 106994870, + "step": 4962, + "time_per_iteration": 2.6853818893432617 + }, + { + "auxiliary_loss_clip": 0.0119323, + "auxiliary_loss_mlp": 0.01028033, + "balance_loss_clip": 1.04965413, + "balance_loss_mlp": 1.02069533, + "epoch": 0.5967654662418085, + "flos": 25481170477440.0, + "grad_norm": 1.973121553032953, + "language_loss": 0.74209297, + "learning_rate": 1.4764703022852598e-06, + "loss": 0.76430559, + "num_input_tokens_seen": 107015390, + "step": 4963, + "time_per_iteration": 2.747811794281006 + }, + { + "auxiliary_loss_clip": 0.01205859, + "auxiliary_loss_mlp": 0.01019521, + "balance_loss_clip": 1.04192948, + "balance_loss_mlp": 1.01225221, + "epoch": 0.5968857091324475, + "flos": 19099126621440.0, + "grad_norm": 1.7071762940843942, + "language_loss": 0.77287441, + "learning_rate": 1.4757185317974696e-06, + "loss": 0.79512823, + "num_input_tokens_seen": 107033775, + "step": 4964, + "time_per_iteration": 2.777527093887329 + }, + { + "auxiliary_loss_clip": 0.01184805, + "auxiliary_loss_mlp": 0.01028316, + "balance_loss_clip": 1.04746604, + "balance_loss_mlp": 1.02023911, + "epoch": 0.5970059520230866, + "flos": 23692711374720.0, + "grad_norm": 2.7147112018372663, + "language_loss": 0.71626455, + "learning_rate": 1.474966840842761e-06, + "loss": 0.73839581, + "num_input_tokens_seen": 107053355, + "step": 4965, + "time_per_iteration": 2.653012990951538 + }, + { + "auxiliary_loss_clip": 0.01189881, + "auxiliary_loss_mlp": 0.01026252, + "balance_loss_clip": 1.05153918, + "balance_loss_mlp": 1.01880717, + "epoch": 0.5971261949137258, + "flos": 23185545292800.0, + "grad_norm": 1.7505475670241228, + "language_loss": 0.86967933, + "learning_rate": 1.4742152295351655e-06, + "loss": 0.8918407, + "num_input_tokens_seen": 107072510, + "step": 4966, + "time_per_iteration": 2.6377921104431152 + }, + { + "auxiliary_loss_clip": 0.01185075, + "auxiliary_loss_mlp": 0.0137929, + "balance_loss_clip": 1.04888749, + "balance_loss_mlp": 1.00021803, + "epoch": 0.5972464378043648, + "flos": 20557710195840.0, + "grad_norm": 3.0309281636711414, + "language_loss": 0.63798219, + "learning_rate": 1.4734636979887016e-06, + "loss": 0.66362584, + "num_input_tokens_seen": 107089970, + "step": 4967, + "time_per_iteration": 2.60334849357605 + }, + { + "auxiliary_loss_clip": 0.01207768, + "auxiliary_loss_mlp": 0.01025148, + "balance_loss_clip": 1.04698205, + "balance_loss_mlp": 1.01738739, + "epoch": 0.5973666806950039, + "flos": 29387030457600.0, + "grad_norm": 1.9491382794655505, + "language_loss": 0.90175748, + "learning_rate": 1.4727122463173755e-06, + "loss": 0.92408663, + "num_input_tokens_seen": 107108500, + "step": 4968, + "time_per_iteration": 2.8154964447021484 + }, + { + "auxiliary_loss_clip": 0.01194942, + "auxiliary_loss_mlp": 0.01026779, + "balance_loss_clip": 1.04975045, + "balance_loss_mlp": 1.01891685, + "epoch": 0.597486923585643, + "flos": 22273522041600.0, + "grad_norm": 1.7833715721717005, + "language_loss": 0.6429776, + "learning_rate": 1.471960874635183e-06, + "loss": 0.66519475, + "num_input_tokens_seen": 107128060, + "step": 4969, + "time_per_iteration": 2.6523427963256836 + }, + { + "auxiliary_loss_clip": 0.01186266, + "auxiliary_loss_mlp": 0.01023135, + "balance_loss_clip": 1.04592657, + "balance_loss_mlp": 1.01483238, + "epoch": 0.5976071664762821, + "flos": 13772461196160.0, + "grad_norm": 2.2284654188191064, + "language_loss": 0.71477687, + "learning_rate": 1.4712095830561055e-06, + "loss": 0.73687088, + "num_input_tokens_seen": 107146550, + "step": 4970, + "time_per_iteration": 2.673051595687866 + }, + { + "auxiliary_loss_clip": 0.0119131, + "auxiliary_loss_mlp": 0.01022185, + "balance_loss_clip": 1.04689384, + "balance_loss_mlp": 1.01481771, + "epoch": 0.5977274093669211, + "flos": 19098623831040.0, + "grad_norm": 2.658901397018828, + "language_loss": 0.8144924, + "learning_rate": 1.4704583716941147e-06, + "loss": 0.83662736, + "num_input_tokens_seen": 107165415, + "step": 4971, + "time_per_iteration": 2.637568235397339 + }, + { + "auxiliary_loss_clip": 0.01177205, + "auxiliary_loss_mlp": 0.01022226, + "balance_loss_clip": 1.05063379, + "balance_loss_mlp": 1.01467681, + "epoch": 0.5978476522575603, + "flos": 20376002269440.0, + "grad_norm": 1.959990710349794, + "language_loss": 0.72223479, + "learning_rate": 1.4697072406631672e-06, + "loss": 0.74422908, + "num_input_tokens_seen": 107185320, + "step": 4972, + "time_per_iteration": 2.6591880321502686 + }, + { + "auxiliary_loss_clip": 0.01219099, + "auxiliary_loss_mlp": 0.0102921, + "balance_loss_clip": 1.05041087, + "balance_loss_mlp": 1.02082348, + "epoch": 0.5979678951481994, + "flos": 29023147728000.0, + "grad_norm": 1.667379745109581, + "language_loss": 0.73034608, + "learning_rate": 1.4689561900772097e-06, + "loss": 0.75282919, + "num_input_tokens_seen": 107205380, + "step": 4973, + "time_per_iteration": 2.7931947708129883 + }, + { + "auxiliary_loss_clip": 0.011931, + "auxiliary_loss_mlp": 0.01030987, + "balance_loss_clip": 1.04839873, + "balance_loss_mlp": 1.02322006, + "epoch": 0.5980881380388384, + "flos": 17967689141760.0, + "grad_norm": 2.4258351028699714, + "language_loss": 0.72684228, + "learning_rate": 1.4682052200501758e-06, + "loss": 0.74908316, + "num_input_tokens_seen": 107222585, + "step": 4974, + "time_per_iteration": 2.6031227111816406 + }, + { + "auxiliary_loss_clip": 0.0117752, + "auxiliary_loss_mlp": 0.01025505, + "balance_loss_clip": 1.05161607, + "balance_loss_mlp": 1.01730669, + "epoch": 0.5982083809294776, + "flos": 22962827013120.0, + "grad_norm": 1.756469507066638, + "language_loss": 0.80108827, + "learning_rate": 1.4674543306959876e-06, + "loss": 0.82311857, + "num_input_tokens_seen": 107242055, + "step": 4975, + "time_per_iteration": 2.646831750869751 + }, + { + "auxiliary_loss_clip": 0.01196637, + "auxiliary_loss_mlp": 0.0102712, + "balance_loss_clip": 1.04774892, + "balance_loss_mlp": 1.0187993, + "epoch": 0.5983286238201166, + "flos": 20991941712000.0, + "grad_norm": 2.3878095585271604, + "language_loss": 0.84570694, + "learning_rate": 1.4667035221285535e-06, + "loss": 0.86794454, + "num_input_tokens_seen": 107259695, + "step": 4976, + "time_per_iteration": 4.530776023864746 + }, + { + "auxiliary_loss_clip": 0.01181567, + "auxiliary_loss_mlp": 0.01023612, + "balance_loss_clip": 1.04909801, + "balance_loss_mlp": 1.01575637, + "epoch": 0.5984488667107557, + "flos": 28183448511360.0, + "grad_norm": 1.831053754555859, + "language_loss": 0.74013668, + "learning_rate": 1.4659527944617715e-06, + "loss": 0.76218849, + "num_input_tokens_seen": 107279640, + "step": 4977, + "time_per_iteration": 2.682292938232422 + }, + { + "auxiliary_loss_clip": 0.0121213, + "auxiliary_loss_mlp": 0.010258, + "balance_loss_clip": 1.04152012, + "balance_loss_mlp": 1.01813436, + "epoch": 0.5985691096013949, + "flos": 16471794314880.0, + "grad_norm": 1.6254540217437152, + "language_loss": 0.76004171, + "learning_rate": 1.465202147809526e-06, + "loss": 0.78242099, + "num_input_tokens_seen": 107298135, + "step": 4978, + "time_per_iteration": 3.6163442134857178 + }, + { + "auxiliary_loss_clip": 0.01176567, + "auxiliary_loss_mlp": 0.01027765, + "balance_loss_clip": 1.05085814, + "balance_loss_mlp": 1.01992726, + "epoch": 0.5986893524920339, + "flos": 26719046933760.0, + "grad_norm": 2.0257739243944446, + "language_loss": 0.75946361, + "learning_rate": 1.4644515822856888e-06, + "loss": 0.7815069, + "num_input_tokens_seen": 107316570, + "step": 4979, + "time_per_iteration": 2.644519805908203 + }, + { + "auxiliary_loss_clip": 0.01104205, + "auxiliary_loss_mlp": 0.01001649, + "balance_loss_clip": 1.011765, + "balance_loss_mlp": 1.00028443, + "epoch": 0.598809595382673, + "flos": 61608061100160.0, + "grad_norm": 0.7534433514236746, + "language_loss": 0.56576234, + "learning_rate": 1.4637010980041215e-06, + "loss": 0.5868209, + "num_input_tokens_seen": 107378680, + "step": 4980, + "time_per_iteration": 4.192070722579956 + }, + { + "auxiliary_loss_clip": 0.01177872, + "auxiliary_loss_mlp": 0.0102749, + "balance_loss_clip": 1.05223489, + "balance_loss_mlp": 1.01941347, + "epoch": 0.5989298382733121, + "flos": 11801719549440.0, + "grad_norm": 2.1843784107912922, + "language_loss": 0.89679033, + "learning_rate": 1.4629506950786707e-06, + "loss": 0.91884392, + "num_input_tokens_seen": 107394860, + "step": 4981, + "time_per_iteration": 2.594289779663086 + }, + { + "auxiliary_loss_clip": 0.01069, + "auxiliary_loss_mlp": 0.01001084, + "balance_loss_clip": 1.01390791, + "balance_loss_mlp": 0.99988586, + "epoch": 0.5990500811639512, + "flos": 60025800021120.0, + "grad_norm": 0.8142801664145489, + "language_loss": 0.5607599, + "learning_rate": 1.4622003736231733e-06, + "loss": 0.58146071, + "num_input_tokens_seen": 107453850, + "step": 4982, + "time_per_iteration": 3.2590086460113525 + }, + { + "auxiliary_loss_clip": 0.01184273, + "auxiliary_loss_mlp": 0.01026999, + "balance_loss_clip": 1.0510273, + "balance_loss_mlp": 1.01855922, + "epoch": 0.5991703240545903, + "flos": 18222726683520.0, + "grad_norm": 1.9726161565173326, + "language_loss": 0.80786109, + "learning_rate": 1.461450133751451e-06, + "loss": 0.82997382, + "num_input_tokens_seen": 107471920, + "step": 4983, + "time_per_iteration": 2.6681418418884277 + }, + { + "auxiliary_loss_clip": 0.01189424, + "auxiliary_loss_mlp": 0.0102195, + "balance_loss_clip": 1.05095649, + "balance_loss_mlp": 1.0146246, + "epoch": 0.5992905669452293, + "flos": 27709894581120.0, + "grad_norm": 1.693361224000888, + "language_loss": 0.75819403, + "learning_rate": 1.4606999755773153e-06, + "loss": 0.78030777, + "num_input_tokens_seen": 107493125, + "step": 4984, + "time_per_iteration": 2.6367294788360596 + }, + { + "auxiliary_loss_clip": 0.01174546, + "auxiliary_loss_mlp": 0.0102612, + "balance_loss_clip": 1.05058312, + "balance_loss_mlp": 1.01885438, + "epoch": 0.5994108098358685, + "flos": 20449008662400.0, + "grad_norm": 1.6412279029905126, + "language_loss": 0.82310742, + "learning_rate": 1.4599498992145643e-06, + "loss": 0.84511411, + "num_input_tokens_seen": 107513150, + "step": 4985, + "time_per_iteration": 2.598134994506836 + }, + { + "auxiliary_loss_clip": 0.01199437, + "auxiliary_loss_mlp": 0.01378546, + "balance_loss_clip": 1.05040717, + "balance_loss_mlp": 1.0001173, + "epoch": 0.5995310527265075, + "flos": 22269966595200.0, + "grad_norm": 2.062011933841704, + "language_loss": 0.7107085, + "learning_rate": 1.4591999047769846e-06, + "loss": 0.73648834, + "num_input_tokens_seen": 107532005, + "step": 4986, + "time_per_iteration": 2.670072555541992 + }, + { + "auxiliary_loss_clip": 0.01211992, + "auxiliary_loss_mlp": 0.01035134, + "balance_loss_clip": 1.04047513, + "balance_loss_mlp": 1.02742112, + "epoch": 0.5996512956171466, + "flos": 18916951818240.0, + "grad_norm": 1.8800541283070908, + "language_loss": 0.7525211, + "learning_rate": 1.4584499923783486e-06, + "loss": 0.77499229, + "num_input_tokens_seen": 107550585, + "step": 4987, + "time_per_iteration": 2.7835073471069336 + }, + { + "auxiliary_loss_clip": 0.01194447, + "auxiliary_loss_mlp": 0.01023889, + "balance_loss_clip": 1.04920888, + "balance_loss_mlp": 1.01655173, + "epoch": 0.5997715385077858, + "flos": 15370916330880.0, + "grad_norm": 1.8233694956972906, + "language_loss": 0.76138026, + "learning_rate": 1.457700162132419e-06, + "loss": 0.78356361, + "num_input_tokens_seen": 107567575, + "step": 4988, + "time_per_iteration": 2.699075222015381 + }, + { + "auxiliary_loss_clip": 0.01205934, + "auxiliary_loss_mlp": 0.01022508, + "balance_loss_clip": 1.04541099, + "balance_loss_mlp": 1.0152185, + "epoch": 0.5998917813984248, + "flos": 25264844818560.0, + "grad_norm": 2.138544533385474, + "language_loss": 0.72654504, + "learning_rate": 1.4569504141529433e-06, + "loss": 0.74882942, + "num_input_tokens_seen": 107585410, + "step": 4989, + "time_per_iteration": 2.8205173015594482 + }, + { + "auxiliary_loss_clip": 0.01184879, + "auxiliary_loss_mlp": 0.01025548, + "balance_loss_clip": 1.05061054, + "balance_loss_mlp": 1.01719117, + "epoch": 0.6000120242890639, + "flos": 22054502862720.0, + "grad_norm": 6.8397796862381375, + "language_loss": 0.72285521, + "learning_rate": 1.456200748553658e-06, + "loss": 0.74495947, + "num_input_tokens_seen": 107603405, + "step": 4990, + "time_per_iteration": 2.7157278060913086 + }, + { + "auxiliary_loss_clip": 0.01177046, + "auxiliary_loss_mlp": 0.01033543, + "balance_loss_clip": 1.05229688, + "balance_loss_mlp": 1.02534139, + "epoch": 0.600132267179703, + "flos": 29863421562240.0, + "grad_norm": 1.4743090650938189, + "language_loss": 0.78808475, + "learning_rate": 1.455451165448287e-06, + "loss": 0.81019068, + "num_input_tokens_seen": 107626060, + "step": 4991, + "time_per_iteration": 2.724163293838501 + }, + { + "auxiliary_loss_clip": 0.01192683, + "auxiliary_loss_mlp": 0.01026332, + "balance_loss_clip": 1.04918551, + "balance_loss_mlp": 1.01846385, + "epoch": 0.6002525100703421, + "flos": 25045358762880.0, + "grad_norm": 2.3435507352496425, + "language_loss": 0.73307526, + "learning_rate": 1.4547016649505407e-06, + "loss": 0.75526541, + "num_input_tokens_seen": 107644070, + "step": 4992, + "time_per_iteration": 2.765842914581299 + }, + { + "auxiliary_loss_clip": 0.01197773, + "auxiliary_loss_mlp": 0.01028088, + "balance_loss_clip": 1.04363179, + "balance_loss_mlp": 1.02023792, + "epoch": 0.6003727529609811, + "flos": 20849592113280.0, + "grad_norm": 2.1116748378889545, + "language_loss": 0.85000652, + "learning_rate": 1.4539522471741193e-06, + "loss": 0.87226516, + "num_input_tokens_seen": 107661495, + "step": 4993, + "time_per_iteration": 278.9556670188904 + }, + { + "auxiliary_loss_clip": 0.01189575, + "auxiliary_loss_mlp": 0.01033311, + "balance_loss_clip": 1.0501194, + "balance_loss_mlp": 1.02481115, + "epoch": 0.6004929958516203, + "flos": 15594604277760.0, + "grad_norm": 1.9542416436148122, + "language_loss": 0.71049535, + "learning_rate": 1.4532029122327067e-06, + "loss": 0.73272419, + "num_input_tokens_seen": 107678280, + "step": 4994, + "time_per_iteration": 2.7194650173187256 + }, + { + "auxiliary_loss_clip": 0.0119752, + "auxiliary_loss_mlp": 0.01028931, + "balance_loss_clip": 1.05094862, + "balance_loss_mlp": 1.02164686, + "epoch": 0.6006132387422594, + "flos": 21763267390080.0, + "grad_norm": 1.8886393858192396, + "language_loss": 0.7547189, + "learning_rate": 1.4524536602399783e-06, + "loss": 0.77698344, + "num_input_tokens_seen": 107697370, + "step": 4995, + "time_per_iteration": 2.6526713371276855 + }, + { + "auxiliary_loss_clip": 0.01190577, + "auxiliary_loss_mlp": 0.01026926, + "balance_loss_clip": 1.05095506, + "balance_loss_mlp": 1.01920116, + "epoch": 0.6007334816328984, + "flos": 22858542852480.0, + "grad_norm": 3.210277091795197, + "language_loss": 0.77208364, + "learning_rate": 1.4517044913095938e-06, + "loss": 0.79425865, + "num_input_tokens_seen": 107717790, + "step": 4996, + "time_per_iteration": 2.631024122238159 + }, + { + "auxiliary_loss_clip": 0.01182841, + "auxiliary_loss_mlp": 0.01028141, + "balance_loss_clip": 1.04948318, + "balance_loss_mlp": 1.02024984, + "epoch": 0.6008537245235376, + "flos": 28324577047680.0, + "grad_norm": 1.650203704373414, + "language_loss": 0.81581473, + "learning_rate": 1.4509554055552022e-06, + "loss": 0.8379246, + "num_input_tokens_seen": 107738020, + "step": 4997, + "time_per_iteration": 2.6357080936431885 + }, + { + "auxiliary_loss_clip": 0.01187836, + "auxiliary_loss_mlp": 0.01030906, + "balance_loss_clip": 1.04627228, + "balance_loss_mlp": 1.02294016, + "epoch": 0.6009739674141766, + "flos": 20886113266560.0, + "grad_norm": 2.310454651353158, + "language_loss": 0.84087813, + "learning_rate": 1.450206403090439e-06, + "loss": 0.8630656, + "num_input_tokens_seen": 107756215, + "step": 4998, + "time_per_iteration": 2.60292911529541 + }, + { + "auxiliary_loss_clip": 0.01180978, + "auxiliary_loss_mlp": 0.0102027, + "balance_loss_clip": 1.05118227, + "balance_loss_mlp": 1.01334071, + "epoch": 0.6010942103048157, + "flos": 20481004702080.0, + "grad_norm": 2.094770015834136, + "language_loss": 0.86247647, + "learning_rate": 1.4494574840289274e-06, + "loss": 0.88448894, + "num_input_tokens_seen": 107773330, + "step": 4999, + "time_per_iteration": 2.58461332321167 + }, + { + "auxiliary_loss_clip": 0.01188364, + "auxiliary_loss_mlp": 0.01027737, + "balance_loss_clip": 1.04933381, + "balance_loss_mlp": 1.01925576, + "epoch": 0.6012144531954549, + "flos": 23805973935360.0, + "grad_norm": 1.7901054044816458, + "language_loss": 0.74055737, + "learning_rate": 1.4487086484842782e-06, + "loss": 0.76271838, + "num_input_tokens_seen": 107791975, + "step": 5000, + "time_per_iteration": 2.593947172164917 + }, + { + "auxiliary_loss_clip": 0.01176217, + "auxiliary_loss_mlp": 0.01024029, + "balance_loss_clip": 1.05155659, + "balance_loss_mlp": 1.01652479, + "epoch": 0.6013346960860939, + "flos": 18988378012800.0, + "grad_norm": 1.8584328651550805, + "language_loss": 0.60474902, + "learning_rate": 1.4479598965700878e-06, + "loss": 0.62675148, + "num_input_tokens_seen": 107809240, + "step": 5001, + "time_per_iteration": 3.472097635269165 + }, + { + "auxiliary_loss_clip": 0.01197666, + "auxiliary_loss_mlp": 0.01025535, + "balance_loss_clip": 1.04547286, + "balance_loss_mlp": 1.01752985, + "epoch": 0.601454938976733, + "flos": 24025316336640.0, + "grad_norm": 2.377773997444708, + "language_loss": 0.68957388, + "learning_rate": 1.4472112283999427e-06, + "loss": 0.71180588, + "num_input_tokens_seen": 107827895, + "step": 5002, + "time_per_iteration": 3.6353039741516113 + }, + { + "auxiliary_loss_clip": 0.01178602, + "auxiliary_loss_mlp": 0.01025611, + "balance_loss_clip": 1.04894948, + "balance_loss_mlp": 1.01826739, + "epoch": 0.6015751818673721, + "flos": 26427129102720.0, + "grad_norm": 1.9234792047571303, + "language_loss": 0.69085443, + "learning_rate": 1.4464626440874143e-06, + "loss": 0.71289659, + "num_input_tokens_seen": 107847010, + "step": 5003, + "time_per_iteration": 3.498868703842163 + }, + { + "auxiliary_loss_clip": 0.01215952, + "auxiliary_loss_mlp": 0.01028243, + "balance_loss_clip": 1.04254961, + "balance_loss_mlp": 1.02005887, + "epoch": 0.6016954247580112, + "flos": 13115260005120.0, + "grad_norm": 2.8458608533574177, + "language_loss": 0.74272943, + "learning_rate": 1.4457141437460636e-06, + "loss": 0.76517141, + "num_input_tokens_seen": 107864235, + "step": 5004, + "time_per_iteration": 2.6700122356414795 + }, + { + "auxiliary_loss_clip": 0.01193905, + "auxiliary_loss_mlp": 0.01030498, + "balance_loss_clip": 1.04842269, + "balance_loss_mlp": 1.02218342, + "epoch": 0.6018156676486502, + "flos": 23768447201280.0, + "grad_norm": 1.8421989814914366, + "language_loss": 0.73403627, + "learning_rate": 1.444965727489436e-06, + "loss": 0.7562803, + "num_input_tokens_seen": 107883680, + "step": 5005, + "time_per_iteration": 2.6201157569885254 + }, + { + "auxiliary_loss_clip": 0.01195689, + "auxiliary_loss_mlp": 0.01023672, + "balance_loss_clip": 1.0432291, + "balance_loss_mlp": 1.01640916, + "epoch": 0.6019359105392894, + "flos": 26469360518400.0, + "grad_norm": 1.7975295969816631, + "language_loss": 0.63169003, + "learning_rate": 1.444217395431066e-06, + "loss": 0.6538837, + "num_input_tokens_seen": 107906220, + "step": 5006, + "time_per_iteration": 3.5711910724639893 + }, + { + "auxiliary_loss_clip": 0.01108939, + "auxiliary_loss_mlp": 0.01002721, + "balance_loss_clip": 1.02031636, + "balance_loss_mlp": 1.00158298, + "epoch": 0.6020561534299285, + "flos": 69190849728000.0, + "grad_norm": 0.8519951561309418, + "language_loss": 0.55796051, + "learning_rate": 1.4434691476844755e-06, + "loss": 0.57907712, + "num_input_tokens_seen": 107967195, + "step": 5007, + "time_per_iteration": 3.235978126525879 + }, + { + "auxiliary_loss_clip": 0.01192253, + "auxiliary_loss_mlp": 0.01024369, + "balance_loss_clip": 1.0504868, + "balance_loss_mlp": 1.01773453, + "epoch": 0.6021763963205675, + "flos": 21835304115840.0, + "grad_norm": 2.0152564651581084, + "language_loss": 0.66984671, + "learning_rate": 1.4427209843631729e-06, + "loss": 0.69201291, + "num_input_tokens_seen": 107984245, + "step": 5008, + "time_per_iteration": 2.670264959335327 + }, + { + "auxiliary_loss_clip": 0.01171998, + "auxiliary_loss_mlp": 0.01378565, + "balance_loss_clip": 1.04989767, + "balance_loss_mlp": 1.00021374, + "epoch": 0.6022966392112067, + "flos": 26578636669440.0, + "grad_norm": 1.7305350653769673, + "language_loss": 0.81313449, + "learning_rate": 1.4419729055806534e-06, + "loss": 0.83864015, + "num_input_tokens_seen": 108003680, + "step": 5009, + "time_per_iteration": 2.647916078567505 + }, + { + "auxiliary_loss_clip": 0.0118835, + "auxiliary_loss_mlp": 0.01378602, + "balance_loss_clip": 1.04973447, + "balance_loss_mlp": 1.0001626, + "epoch": 0.6024168821018457, + "flos": 20703722981760.0, + "grad_norm": 2.289899656900874, + "language_loss": 0.82156283, + "learning_rate": 1.441224911450401e-06, + "loss": 0.84723234, + "num_input_tokens_seen": 108019635, + "step": 5010, + "time_per_iteration": 2.6436808109283447 + }, + { + "auxiliary_loss_clip": 0.01185784, + "auxiliary_loss_mlp": 0.01025814, + "balance_loss_clip": 1.04908645, + "balance_loss_mlp": 1.01839364, + "epoch": 0.6025371249924848, + "flos": 24680973242880.0, + "grad_norm": 2.0636820343112507, + "language_loss": 0.8215487, + "learning_rate": 1.4404770020858851e-06, + "loss": 0.84366465, + "num_input_tokens_seen": 108039120, + "step": 5011, + "time_per_iteration": 2.6895675659179688 + }, + { + "auxiliary_loss_clip": 0.01177205, + "auxiliary_loss_mlp": 0.01027068, + "balance_loss_clip": 1.04820538, + "balance_loss_mlp": 1.01967096, + "epoch": 0.602657367883124, + "flos": 25955801815680.0, + "grad_norm": 1.809296358535974, + "language_loss": 0.86109871, + "learning_rate": 1.439729177600563e-06, + "loss": 0.8831414, + "num_input_tokens_seen": 108059615, + "step": 5012, + "time_per_iteration": 2.695675849914551 + }, + { + "auxiliary_loss_clip": 0.01180319, + "auxiliary_loss_mlp": 0.01028731, + "balance_loss_clip": 1.04893899, + "balance_loss_mlp": 1.02117062, + "epoch": 0.602777610773763, + "flos": 16690633925760.0, + "grad_norm": 2.1851002884885116, + "language_loss": 0.73159254, + "learning_rate": 1.4389814381078793e-06, + "loss": 0.75368297, + "num_input_tokens_seen": 108078855, + "step": 5013, + "time_per_iteration": 2.5987110137939453 + }, + { + "auxiliary_loss_clip": 0.01244595, + "auxiliary_loss_mlp": 0.01024745, + "balance_loss_clip": 1.04050422, + "balance_loss_mlp": 1.01725292, + "epoch": 0.6028978536644021, + "flos": 13334243270400.0, + "grad_norm": 1.9001742351339244, + "language_loss": 0.80029887, + "learning_rate": 1.438233783721265e-06, + "loss": 0.82299221, + "num_input_tokens_seen": 108095020, + "step": 5014, + "time_per_iteration": 3.2404139041900635 + }, + { + "auxiliary_loss_clip": 0.01186861, + "auxiliary_loss_mlp": 0.01024234, + "balance_loss_clip": 1.05050862, + "balance_loss_mlp": 1.01688457, + "epoch": 0.6030180965550412, + "flos": 19644825018240.0, + "grad_norm": 2.0315641732441447, + "language_loss": 0.77840042, + "learning_rate": 1.43748621455414e-06, + "loss": 0.80051136, + "num_input_tokens_seen": 108111455, + "step": 5015, + "time_per_iteration": 2.8461592197418213 + }, + { + "auxiliary_loss_clip": 0.01190162, + "auxiliary_loss_mlp": 0.01024401, + "balance_loss_clip": 1.04820001, + "balance_loss_mlp": 1.0158062, + "epoch": 0.6031383394456803, + "flos": 14458390289280.0, + "grad_norm": 3.4984680431524064, + "language_loss": 0.80443555, + "learning_rate": 1.4367387307199082e-06, + "loss": 0.82658124, + "num_input_tokens_seen": 108128305, + "step": 5016, + "time_per_iteration": 2.7211220264434814 + }, + { + "auxiliary_loss_clip": 0.01175912, + "auxiliary_loss_mlp": 0.01027763, + "balance_loss_clip": 1.04668355, + "balance_loss_mlp": 1.02036011, + "epoch": 0.6032585823363193, + "flos": 13917791623680.0, + "grad_norm": 1.733449069920537, + "language_loss": 0.82623637, + "learning_rate": 1.4359913323319632e-06, + "loss": 0.84827316, + "num_input_tokens_seen": 108145475, + "step": 5017, + "time_per_iteration": 2.6385879516601562 + }, + { + "auxiliary_loss_clip": 0.01204633, + "auxiliary_loss_mlp": 0.01024659, + "balance_loss_clip": 1.04091895, + "balance_loss_mlp": 1.01629055, + "epoch": 0.6033788252269584, + "flos": 24353252530560.0, + "grad_norm": 1.6917507700374184, + "language_loss": 0.77813107, + "learning_rate": 1.4352440195036847e-06, + "loss": 0.80042398, + "num_input_tokens_seen": 108165650, + "step": 5018, + "time_per_iteration": 2.8069300651550293 + }, + { + "auxiliary_loss_clip": 0.01229137, + "auxiliary_loss_mlp": 0.01025041, + "balance_loss_clip": 1.03919411, + "balance_loss_mlp": 1.01790047, + "epoch": 0.6034990681175976, + "flos": 25521247077120.0, + "grad_norm": 1.8606410479642317, + "language_loss": 0.79749084, + "learning_rate": 1.4344967923484395e-06, + "loss": 0.8200326, + "num_input_tokens_seen": 108187620, + "step": 5019, + "time_per_iteration": 2.789881944656372 + }, + { + "auxiliary_loss_clip": 0.01180332, + "auxiliary_loss_mlp": 0.01026662, + "balance_loss_clip": 1.04965138, + "balance_loss_mlp": 1.01894951, + "epoch": 0.6036193110082366, + "flos": 25958387594880.0, + "grad_norm": 1.9943698453956604, + "language_loss": 0.72275233, + "learning_rate": 1.433749650979581e-06, + "loss": 0.74482226, + "num_input_tokens_seen": 108207605, + "step": 5020, + "time_per_iteration": 2.887808322906494 + }, + { + "auxiliary_loss_clip": 0.01205352, + "auxiliary_loss_mlp": 0.01031795, + "balance_loss_clip": 1.04607749, + "balance_loss_mlp": 1.02421331, + "epoch": 0.6037395538988757, + "flos": 25593427457280.0, + "grad_norm": 1.6900778875060796, + "language_loss": 0.6793679, + "learning_rate": 1.433002595510451e-06, + "loss": 0.70173937, + "num_input_tokens_seen": 108226385, + "step": 5021, + "time_per_iteration": 3.5044820308685303 + }, + { + "auxiliary_loss_clip": 0.01185568, + "auxiliary_loss_mlp": 0.01378859, + "balance_loss_clip": 1.04526055, + "balance_loss_mlp": 1.00017786, + "epoch": 0.6038597967895148, + "flos": 17816253402240.0, + "grad_norm": 1.736943521791554, + "language_loss": 0.71671075, + "learning_rate": 1.4322556260543757e-06, + "loss": 0.74235499, + "num_input_tokens_seen": 108242960, + "step": 5022, + "time_per_iteration": 2.658573627471924 + }, + { + "auxiliary_loss_clip": 0.01104977, + "auxiliary_loss_mlp": 0.01000278, + "balance_loss_clip": 1.01133084, + "balance_loss_mlp": 0.9990446, + "epoch": 0.6039800396801539, + "flos": 65169213235200.0, + "grad_norm": 0.894153701548628, + "language_loss": 0.62752819, + "learning_rate": 1.4315087427246703e-06, + "loss": 0.64858073, + "num_input_tokens_seen": 108296785, + "step": 5023, + "time_per_iteration": 3.1860246658325195 + }, + { + "auxiliary_loss_clip": 0.01066228, + "auxiliary_loss_mlp": 0.01001851, + "balance_loss_clip": 1.01137984, + "balance_loss_mlp": 1.00067651, + "epoch": 0.604100282570793, + "flos": 67386409073280.0, + "grad_norm": 0.87281248856669, + "language_loss": 0.58485025, + "learning_rate": 1.4307619456346372e-06, + "loss": 0.60553104, + "num_input_tokens_seen": 108341090, + "step": 5024, + "time_per_iteration": 2.879485607147217 + }, + { + "auxiliary_loss_clip": 0.01181541, + "auxiliary_loss_mlp": 0.01022583, + "balance_loss_clip": 1.04531646, + "balance_loss_mlp": 1.01497126, + "epoch": 0.6042205254614321, + "flos": 35297495631360.0, + "grad_norm": 1.8564567993084808, + "language_loss": 0.74177516, + "learning_rate": 1.430015234897564e-06, + "loss": 0.76381636, + "num_input_tokens_seen": 108364370, + "step": 5025, + "time_per_iteration": 2.7953710556030273 + }, + { + "auxiliary_loss_clip": 0.01172055, + "auxiliary_loss_mlp": 0.01378517, + "balance_loss_clip": 1.04886436, + "balance_loss_mlp": 1.00017381, + "epoch": 0.6043407683520712, + "flos": 45658262206080.0, + "grad_norm": 1.539970554244968, + "language_loss": 0.66578662, + "learning_rate": 1.4292686106267274e-06, + "loss": 0.69129241, + "num_input_tokens_seen": 108387220, + "step": 5026, + "time_per_iteration": 2.750547409057617 + }, + { + "auxiliary_loss_clip": 0.01184925, + "auxiliary_loss_mlp": 0.01030725, + "balance_loss_clip": 1.0485779, + "balance_loss_mlp": 1.02341771, + "epoch": 0.6044610112427102, + "flos": 16180020138240.0, + "grad_norm": 1.5512020129153021, + "language_loss": 0.77426076, + "learning_rate": 1.4285220729353876e-06, + "loss": 0.79641736, + "num_input_tokens_seen": 108405760, + "step": 5027, + "time_per_iteration": 3.3895082473754883 + }, + { + "auxiliary_loss_clip": 0.01187601, + "auxiliary_loss_mlp": 0.01023278, + "balance_loss_clip": 1.04477227, + "balance_loss_mlp": 1.01576519, + "epoch": 0.6045812541333494, + "flos": 13804062186240.0, + "grad_norm": 1.8086359331506427, + "language_loss": 0.77865499, + "learning_rate": 1.4277756219367957e-06, + "loss": 0.80076373, + "num_input_tokens_seen": 108422785, + "step": 5028, + "time_per_iteration": 3.700751543045044 + }, + { + "auxiliary_loss_clip": 0.01205726, + "auxiliary_loss_mlp": 0.01030573, + "balance_loss_clip": 1.04618418, + "balance_loss_mlp": 1.02258658, + "epoch": 0.6047014970239885, + "flos": 19975059682560.0, + "grad_norm": 1.8174060986292806, + "language_loss": 0.79718423, + "learning_rate": 1.4270292577441864e-06, + "loss": 0.81954724, + "num_input_tokens_seen": 108442290, + "step": 5029, + "time_per_iteration": 2.6915500164031982 + }, + { + "auxiliary_loss_clip": 0.01183767, + "auxiliary_loss_mlp": 0.01022902, + "balance_loss_clip": 1.04642725, + "balance_loss_mlp": 1.01545763, + "epoch": 0.6048217399146275, + "flos": 25337097025920.0, + "grad_norm": 1.5635158360415142, + "language_loss": 0.720052, + "learning_rate": 1.4262829804707836e-06, + "loss": 0.7421186, + "num_input_tokens_seen": 108464280, + "step": 5030, + "time_per_iteration": 3.5695228576660156 + }, + { + "auxiliary_loss_clip": 0.01182214, + "auxiliary_loss_mlp": 0.01024804, + "balance_loss_clip": 1.04637492, + "balance_loss_mlp": 1.01677561, + "epoch": 0.6049419828052667, + "flos": 26030819370240.0, + "grad_norm": 1.4250032210874128, + "language_loss": 0.6978668, + "learning_rate": 1.4255367902297958e-06, + "loss": 0.71993697, + "num_input_tokens_seen": 108485610, + "step": 5031, + "time_per_iteration": 2.704991579055786 + }, + { + "auxiliary_loss_clip": 0.01172066, + "auxiliary_loss_mlp": 0.01031569, + "balance_loss_clip": 1.04958296, + "balance_loss_mlp": 1.02409124, + "epoch": 0.6050622256959057, + "flos": 14648106948480.0, + "grad_norm": 2.10374596958434, + "language_loss": 0.78658462, + "learning_rate": 1.4247906871344215e-06, + "loss": 0.80862099, + "num_input_tokens_seen": 108501005, + "step": 5032, + "time_per_iteration": 3.4191696643829346 + }, + { + "auxiliary_loss_clip": 0.01183912, + "auxiliary_loss_mlp": 0.01024141, + "balance_loss_clip": 1.04451871, + "balance_loss_mlp": 1.01660061, + "epoch": 0.6051824685865448, + "flos": 23331450337920.0, + "grad_norm": 2.178546680394733, + "language_loss": 0.75269401, + "learning_rate": 1.4240446712978415e-06, + "loss": 0.77477449, + "num_input_tokens_seen": 108519990, + "step": 5033, + "time_per_iteration": 2.695950984954834 + }, + { + "auxiliary_loss_clip": 0.01184793, + "auxiliary_loss_mlp": 0.01027126, + "balance_loss_clip": 1.0477556, + "balance_loss_mlp": 1.01889467, + "epoch": 0.605302711477184, + "flos": 27563307177600.0, + "grad_norm": 1.8115677155084506, + "language_loss": 0.74081063, + "learning_rate": 1.423298742833227e-06, + "loss": 0.7629298, + "num_input_tokens_seen": 108538650, + "step": 5034, + "time_per_iteration": 2.63480281829834 + }, + { + "auxiliary_loss_clip": 0.01203684, + "auxiliary_loss_mlp": 0.01026639, + "balance_loss_clip": 1.04445779, + "balance_loss_mlp": 1.01872349, + "epoch": 0.605422954367823, + "flos": 15154698412800.0, + "grad_norm": 1.7876856808666606, + "language_loss": 0.71574175, + "learning_rate": 1.4225529018537352e-06, + "loss": 0.73804492, + "num_input_tokens_seen": 108554155, + "step": 5035, + "time_per_iteration": 2.708824872970581 + }, + { + "auxiliary_loss_clip": 0.01172393, + "auxiliary_loss_mlp": 0.01025171, + "balance_loss_clip": 1.05016708, + "balance_loss_mlp": 1.01744032, + "epoch": 0.6055431972584621, + "flos": 27673912131840.0, + "grad_norm": 1.739510168718603, + "language_loss": 0.77966535, + "learning_rate": 1.4218071484725082e-06, + "loss": 0.80164099, + "num_input_tokens_seen": 108576275, + "step": 5036, + "time_per_iteration": 2.6675803661346436 + }, + { + "auxiliary_loss_clip": 0.01188429, + "auxiliary_loss_mlp": 0.01028718, + "balance_loss_clip": 1.04952693, + "balance_loss_mlp": 1.0211122, + "epoch": 0.6056634401491012, + "flos": 19387489006080.0, + "grad_norm": 1.9652691033172112, + "language_loss": 0.76399577, + "learning_rate": 1.4210614828026786e-06, + "loss": 0.78616726, + "num_input_tokens_seen": 108594125, + "step": 5037, + "time_per_iteration": 2.6519978046417236 + }, + { + "auxiliary_loss_clip": 0.0117117, + "auxiliary_loss_mlp": 0.01017845, + "balance_loss_clip": 1.04836869, + "balance_loss_mlp": 1.01061535, + "epoch": 0.6057836830397403, + "flos": 24789459294720.0, + "grad_norm": 1.624955341742406, + "language_loss": 0.74402583, + "learning_rate": 1.4203159049573605e-06, + "loss": 0.76591599, + "num_input_tokens_seen": 108615360, + "step": 5038, + "time_per_iteration": 2.599740505218506 + }, + { + "auxiliary_loss_clip": 0.01194956, + "auxiliary_loss_mlp": 0.01029138, + "balance_loss_clip": 1.04519844, + "balance_loss_mlp": 1.02152085, + "epoch": 0.6059039259303793, + "flos": 20558248899840.0, + "grad_norm": 1.982037460669727, + "language_loss": 0.87194753, + "learning_rate": 1.4195704150496593e-06, + "loss": 0.8941884, + "num_input_tokens_seen": 108633075, + "step": 5039, + "time_per_iteration": 2.6819283962249756 + }, + { + "auxiliary_loss_clip": 0.01191202, + "auxiliary_loss_mlp": 0.01025606, + "balance_loss_clip": 1.04798067, + "balance_loss_mlp": 1.01753545, + "epoch": 0.6060241688210185, + "flos": 21069724613760.0, + "grad_norm": 1.6281269075682039, + "language_loss": 0.73870277, + "learning_rate": 1.4188250131926639e-06, + "loss": 0.76087087, + "num_input_tokens_seen": 108651875, + "step": 5040, + "time_per_iteration": 2.67334246635437 + }, + { + "auxiliary_loss_clip": 0.01190548, + "auxiliary_loss_mlp": 0.01027555, + "balance_loss_clip": 1.04670942, + "balance_loss_mlp": 1.01980615, + "epoch": 0.6061444117116576, + "flos": 16361081619840.0, + "grad_norm": 1.795902818309764, + "language_loss": 0.80714798, + "learning_rate": 1.4180796994994525e-06, + "loss": 0.82932901, + "num_input_tokens_seen": 108669290, + "step": 5041, + "time_per_iteration": 2.6400911808013916 + }, + { + "auxiliary_loss_clip": 0.01187015, + "auxiliary_loss_mlp": 0.01022385, + "balance_loss_clip": 1.04527473, + "balance_loss_mlp": 1.01547337, + "epoch": 0.6062646546022966, + "flos": 21507296094720.0, + "grad_norm": 4.351895094201186, + "language_loss": 0.72577298, + "learning_rate": 1.4173344740830877e-06, + "loss": 0.74786693, + "num_input_tokens_seen": 108688420, + "step": 5042, + "time_per_iteration": 2.678586483001709 + }, + { + "auxiliary_loss_clip": 0.0118362, + "auxiliary_loss_mlp": 0.01024844, + "balance_loss_clip": 1.04798603, + "balance_loss_mlp": 1.0174408, + "epoch": 0.6063848974929358, + "flos": 38983151283840.0, + "grad_norm": 1.5408031462275746, + "language_loss": 0.7044484, + "learning_rate": 1.4165893370566206e-06, + "loss": 0.72653306, + "num_input_tokens_seen": 108712175, + "step": 5043, + "time_per_iteration": 2.8405823707580566 + }, + { + "auxiliary_loss_clip": 0.01176266, + "auxiliary_loss_mlp": 0.01026195, + "balance_loss_clip": 1.04666495, + "balance_loss_mlp": 1.01867247, + "epoch": 0.6065051403835748, + "flos": 19646584784640.0, + "grad_norm": 1.6829704463065933, + "language_loss": 0.77632737, + "learning_rate": 1.4158442885330865e-06, + "loss": 0.798352, + "num_input_tokens_seen": 108730745, + "step": 5044, + "time_per_iteration": 2.6199798583984375 + }, + { + "auxiliary_loss_clip": 0.01174969, + "auxiliary_loss_mlp": 0.0102436, + "balance_loss_clip": 1.04594755, + "balance_loss_mlp": 1.01694548, + "epoch": 0.6066253832742139, + "flos": 23513086437120.0, + "grad_norm": 2.2017105935050116, + "language_loss": 0.79035324, + "learning_rate": 1.4150993286255094e-06, + "loss": 0.81234652, + "num_input_tokens_seen": 108749995, + "step": 5045, + "time_per_iteration": 2.6393744945526123 + }, + { + "auxiliary_loss_clip": 0.01169075, + "auxiliary_loss_mlp": 0.01026035, + "balance_loss_clip": 1.04687536, + "balance_loss_mlp": 1.01912665, + "epoch": 0.6067456261648531, + "flos": 19133708440320.0, + "grad_norm": 1.889658728308776, + "language_loss": 0.79865587, + "learning_rate": 1.4143544574468993e-06, + "loss": 0.82060695, + "num_input_tokens_seen": 108768355, + "step": 5046, + "time_per_iteration": 2.6335508823394775 + }, + { + "auxiliary_loss_clip": 0.0117847, + "auxiliary_loss_mlp": 0.01023558, + "balance_loss_clip": 1.04799318, + "balance_loss_mlp": 1.01649177, + "epoch": 0.6068658690554921, + "flos": 20520614424960.0, + "grad_norm": 1.7522357919018865, + "language_loss": 0.82489032, + "learning_rate": 1.4136096751102523e-06, + "loss": 0.8469106, + "num_input_tokens_seen": 108786685, + "step": 5047, + "time_per_iteration": 2.637357473373413 + }, + { + "auxiliary_loss_clip": 0.01195126, + "auxiliary_loss_mlp": 0.01029355, + "balance_loss_clip": 1.04887748, + "balance_loss_mlp": 1.02238166, + "epoch": 0.6069861119461312, + "flos": 27374560185600.0, + "grad_norm": 1.968996300401948, + "language_loss": 0.83096051, + "learning_rate": 1.4128649817285516e-06, + "loss": 0.85320532, + "num_input_tokens_seen": 108804820, + "step": 5048, + "time_per_iteration": 2.737483263015747 + }, + { + "auxiliary_loss_clip": 0.01189293, + "auxiliary_loss_mlp": 0.01029982, + "balance_loss_clip": 1.04370475, + "balance_loss_mlp": 1.02276695, + "epoch": 0.6071063548367702, + "flos": 25626500904960.0, + "grad_norm": 1.9055241248954158, + "language_loss": 0.62896669, + "learning_rate": 1.412120377414766e-06, + "loss": 0.65115947, + "num_input_tokens_seen": 108825010, + "step": 5049, + "time_per_iteration": 2.7001259326934814 + }, + { + "auxiliary_loss_clip": 0.01171843, + "auxiliary_loss_mlp": 0.01024268, + "balance_loss_clip": 1.05039692, + "balance_loss_mlp": 1.01710927, + "epoch": 0.6072265977274094, + "flos": 24460517520000.0, + "grad_norm": 1.7660009028426797, + "language_loss": 0.71232373, + "learning_rate": 1.4113758622818522e-06, + "loss": 0.73428482, + "num_input_tokens_seen": 108845075, + "step": 5050, + "time_per_iteration": 2.6957361698150635 + }, + { + "auxiliary_loss_clip": 0.01196243, + "auxiliary_loss_mlp": 0.01378064, + "balance_loss_clip": 1.04930198, + "balance_loss_mlp": 1.00025117, + "epoch": 0.6073468406180484, + "flos": 18149253413760.0, + "grad_norm": 2.073710166160785, + "language_loss": 0.83096743, + "learning_rate": 1.410631436442751e-06, + "loss": 0.85671055, + "num_input_tokens_seen": 108863870, + "step": 5051, + "time_per_iteration": 2.720871686935425 + }, + { + "auxiliary_loss_clip": 0.01185755, + "auxiliary_loss_mlp": 0.01022081, + "balance_loss_clip": 1.04794455, + "balance_loss_mlp": 1.01478589, + "epoch": 0.6074670835086875, + "flos": 20697617669760.0, + "grad_norm": 2.107018526237109, + "language_loss": 0.8649981, + "learning_rate": 1.4098871000103936e-06, + "loss": 0.88707644, + "num_input_tokens_seen": 108882470, + "step": 5052, + "time_per_iteration": 2.6497995853424072 + }, + { + "auxiliary_loss_clip": 0.01189787, + "auxiliary_loss_mlp": 0.01024839, + "balance_loss_clip": 1.0459553, + "balance_loss_mlp": 1.01715064, + "epoch": 0.6075873263993267, + "flos": 23769955572480.0, + "grad_norm": 1.7147129482831582, + "language_loss": 0.82722437, + "learning_rate": 1.409142853097693e-06, + "loss": 0.8493706, + "num_input_tokens_seen": 108902710, + "step": 5053, + "time_per_iteration": 3.5723376274108887 + }, + { + "auxiliary_loss_clip": 0.0119131, + "auxiliary_loss_mlp": 0.01026649, + "balance_loss_clip": 1.04717422, + "balance_loss_mlp": 1.01873922, + "epoch": 0.6077075692899657, + "flos": 24454484035200.0, + "grad_norm": 1.9441844022616672, + "language_loss": 0.79443502, + "learning_rate": 1.408398695817553e-06, + "loss": 0.81661451, + "num_input_tokens_seen": 108919935, + "step": 5054, + "time_per_iteration": 3.5411176681518555 + }, + { + "auxiliary_loss_clip": 0.01187016, + "auxiliary_loss_mlp": 0.01023588, + "balance_loss_clip": 1.04509008, + "balance_loss_mlp": 1.01558924, + "epoch": 0.6078278121806048, + "flos": 27382102041600.0, + "grad_norm": 9.34394413028431, + "language_loss": 0.70168692, + "learning_rate": 1.4076546282828593e-06, + "loss": 0.72379291, + "num_input_tokens_seen": 108942790, + "step": 5055, + "time_per_iteration": 2.7430145740509033 + }, + { + "auxiliary_loss_clip": 0.01190736, + "auxiliary_loss_mlp": 0.01028479, + "balance_loss_clip": 1.04251766, + "balance_loss_mlp": 1.02103448, + "epoch": 0.6079480550712439, + "flos": 38436447306240.0, + "grad_norm": 3.1261300037881643, + "language_loss": 0.65868759, + "learning_rate": 1.4069106506064874e-06, + "loss": 0.68087971, + "num_input_tokens_seen": 108964215, + "step": 5056, + "time_per_iteration": 3.59738826751709 + }, + { + "auxiliary_loss_clip": 0.0118899, + "auxiliary_loss_mlp": 0.01026107, + "balance_loss_clip": 1.05000913, + "balance_loss_mlp": 1.01919615, + "epoch": 0.608068297961883, + "flos": 25336271013120.0, + "grad_norm": 1.8006266288280157, + "language_loss": 0.78641605, + "learning_rate": 1.4061667629012989e-06, + "loss": 0.80856699, + "num_input_tokens_seen": 108984885, + "step": 5057, + "time_per_iteration": 2.634763717651367 + }, + { + "auxiliary_loss_clip": 0.01176361, + "auxiliary_loss_mlp": 0.0102785, + "balance_loss_clip": 1.04470575, + "balance_loss_mlp": 1.02040815, + "epoch": 0.608188540852522, + "flos": 24202463235840.0, + "grad_norm": 1.7176498110391982, + "language_loss": 0.83394992, + "learning_rate": 1.40542296528014e-06, + "loss": 0.85599202, + "num_input_tokens_seen": 109004545, + "step": 5058, + "time_per_iteration": 3.6182427406311035 + }, + { + "auxiliary_loss_clip": 0.0118094, + "auxiliary_loss_mlp": 0.01027242, + "balance_loss_clip": 1.04804111, + "balance_loss_mlp": 1.01974034, + "epoch": 0.6083087837431612, + "flos": 21284146851840.0, + "grad_norm": 2.0976651442245244, + "language_loss": 0.75918335, + "learning_rate": 1.4046792578558452e-06, + "loss": 0.78126514, + "num_input_tokens_seen": 109022440, + "step": 5059, + "time_per_iteration": 2.68562650680542 + }, + { + "auxiliary_loss_clip": 0.0118818, + "auxiliary_loss_mlp": 0.01027428, + "balance_loss_clip": 1.04759383, + "balance_loss_mlp": 1.02013826, + "epoch": 0.6084290266338003, + "flos": 16471435178880.0, + "grad_norm": 2.141049919182063, + "language_loss": 0.75815749, + "learning_rate": 1.4039356407412325e-06, + "loss": 0.78031355, + "num_input_tokens_seen": 109035680, + "step": 5060, + "time_per_iteration": 2.776806116104126 + }, + { + "auxiliary_loss_clip": 0.01079679, + "auxiliary_loss_mlp": 0.01002836, + "balance_loss_clip": 1.01150107, + "balance_loss_mlp": 1.00159645, + "epoch": 0.6085492695244393, + "flos": 66443574931200.0, + "grad_norm": 0.787410382652146, + "language_loss": 0.57146621, + "learning_rate": 1.40319211404911e-06, + "loss": 0.59229136, + "num_input_tokens_seen": 109090680, + "step": 5061, + "time_per_iteration": 3.1962053775787354 + }, + { + "auxiliary_loss_clip": 0.0117545, + "auxiliary_loss_mlp": 0.01027885, + "balance_loss_clip": 1.05131721, + "balance_loss_mlp": 1.02086973, + "epoch": 0.6086695124150785, + "flos": 23618986709760.0, + "grad_norm": 2.3182288595184715, + "language_loss": 0.90557986, + "learning_rate": 1.4024486778922691e-06, + "loss": 0.9276132, + "num_input_tokens_seen": 109108995, + "step": 5062, + "time_per_iteration": 2.570935010910034 + }, + { + "auxiliary_loss_clip": 0.01194058, + "auxiliary_loss_mlp": 0.01025354, + "balance_loss_clip": 1.04451561, + "balance_loss_mlp": 1.01793647, + "epoch": 0.6087897553057176, + "flos": 20157054917760.0, + "grad_norm": 7.646880112292274, + "language_loss": 0.77516758, + "learning_rate": 1.4017053323834884e-06, + "loss": 0.79736173, + "num_input_tokens_seen": 109128825, + "step": 5063, + "time_per_iteration": 2.6905572414398193 + }, + { + "auxiliary_loss_clip": 0.01192948, + "auxiliary_loss_mlp": 0.01023586, + "balance_loss_clip": 1.04560685, + "balance_loss_mlp": 1.01658821, + "epoch": 0.6089099981963566, + "flos": 25482535194240.0, + "grad_norm": 2.0244269463825177, + "language_loss": 0.76019371, + "learning_rate": 1.4009620776355333e-06, + "loss": 0.78235906, + "num_input_tokens_seen": 109150425, + "step": 5064, + "time_per_iteration": 2.6679346561431885 + }, + { + "auxiliary_loss_clip": 0.01177238, + "auxiliary_loss_mlp": 0.01030198, + "balance_loss_clip": 1.0460813, + "balance_loss_mlp": 1.02229476, + "epoch": 0.6090302410869958, + "flos": 25332895134720.0, + "grad_norm": 2.0367212735906963, + "language_loss": 0.79354304, + "learning_rate": 1.4002189137611553e-06, + "loss": 0.81561744, + "num_input_tokens_seen": 109169765, + "step": 5065, + "time_per_iteration": 2.6926538944244385 + }, + { + "auxiliary_loss_clip": 0.01177567, + "auxiliary_loss_mlp": 0.0102427, + "balance_loss_clip": 1.04657197, + "balance_loss_mlp": 1.01736438, + "epoch": 0.6091504839776348, + "flos": 23987358639360.0, + "grad_norm": 1.8811615582721448, + "language_loss": 0.697191, + "learning_rate": 1.3994758408730901e-06, + "loss": 0.71920931, + "num_input_tokens_seen": 109188950, + "step": 5066, + "time_per_iteration": 2.6695470809936523 + }, + { + "auxiliary_loss_clip": 0.01195823, + "auxiliary_loss_mlp": 0.01030655, + "balance_loss_clip": 1.05104434, + "balance_loss_mlp": 1.02334785, + "epoch": 0.6092707268682739, + "flos": 29643037666560.0, + "grad_norm": 2.002795448093154, + "language_loss": 0.7696979, + "learning_rate": 1.3987328590840629e-06, + "loss": 0.79196274, + "num_input_tokens_seen": 109209895, + "step": 5067, + "time_per_iteration": 2.6862685680389404 + }, + { + "auxiliary_loss_clip": 0.01175673, + "auxiliary_loss_mlp": 0.0102608, + "balance_loss_clip": 1.04455316, + "balance_loss_mlp": 1.01859331, + "epoch": 0.609390969758913, + "flos": 24024957200640.0, + "grad_norm": 1.8521387339581064, + "language_loss": 0.86203206, + "learning_rate": 1.397989968506783e-06, + "loss": 0.88404959, + "num_input_tokens_seen": 109228905, + "step": 5068, + "time_per_iteration": 2.6183454990386963 + }, + { + "auxiliary_loss_clip": 0.01178291, + "auxiliary_loss_mlp": 0.01024501, + "balance_loss_clip": 1.05242419, + "balance_loss_mlp": 1.01711631, + "epoch": 0.6095112126495521, + "flos": 11102143288320.0, + "grad_norm": 9.356131169102184, + "language_loss": 0.72417325, + "learning_rate": 1.3972471692539458e-06, + "loss": 0.74620116, + "num_input_tokens_seen": 109243620, + "step": 5069, + "time_per_iteration": 2.582779884338379 + }, + { + "auxiliary_loss_clip": 0.01183878, + "auxiliary_loss_mlp": 0.01026581, + "balance_loss_clip": 1.04566193, + "balance_loss_mlp": 1.01936626, + "epoch": 0.6096314555401912, + "flos": 17265491187840.0, + "grad_norm": 2.10475790760658, + "language_loss": 0.75329113, + "learning_rate": 1.3965044614382348e-06, + "loss": 0.77539575, + "num_input_tokens_seen": 109259070, + "step": 5070, + "time_per_iteration": 2.591376304626465 + }, + { + "auxiliary_loss_clip": 0.01175309, + "auxiliary_loss_mlp": 0.01027506, + "balance_loss_clip": 1.05104053, + "balance_loss_mlp": 1.01949549, + "epoch": 0.6097516984308303, + "flos": 21645910679040.0, + "grad_norm": 2.555740339584095, + "language_loss": 0.75499809, + "learning_rate": 1.3957618451723162e-06, + "loss": 0.77702624, + "num_input_tokens_seen": 109275100, + "step": 5071, + "time_per_iteration": 2.6054434776306152 + }, + { + "auxiliary_loss_clip": 0.01191951, + "auxiliary_loss_mlp": 0.01028105, + "balance_loss_clip": 1.0467844, + "balance_loss_mlp": 1.02013612, + "epoch": 0.6098719413214694, + "flos": 27199208966400.0, + "grad_norm": 1.828499877216809, + "language_loss": 0.714782, + "learning_rate": 1.3950193205688457e-06, + "loss": 0.73698252, + "num_input_tokens_seen": 109294825, + "step": 5072, + "time_per_iteration": 2.717057943344116 + }, + { + "auxiliary_loss_clip": 0.01189113, + "auxiliary_loss_mlp": 0.01023107, + "balance_loss_clip": 1.04732871, + "balance_loss_mlp": 1.01645184, + "epoch": 0.6099921842121084, + "flos": 20412954385920.0, + "grad_norm": 1.7635717098872892, + "language_loss": 0.83535302, + "learning_rate": 1.3942768877404627e-06, + "loss": 0.85747516, + "num_input_tokens_seen": 109313790, + "step": 5073, + "time_per_iteration": 2.6553397178649902 + }, + { + "auxiliary_loss_clip": 0.01172473, + "auxiliary_loss_mlp": 0.0102076, + "balance_loss_clip": 1.04878831, + "balance_loss_mlp": 1.01373267, + "epoch": 0.6101124271027476, + "flos": 23366139897600.0, + "grad_norm": 1.4962684191886504, + "language_loss": 0.73858792, + "learning_rate": 1.393534546799795e-06, + "loss": 0.76052022, + "num_input_tokens_seen": 109333490, + "step": 5074, + "time_per_iteration": 2.618119239807129 + }, + { + "auxiliary_loss_clip": 0.01183615, + "auxiliary_loss_mlp": 0.01026462, + "balance_loss_clip": 1.04852414, + "balance_loss_mlp": 1.01900589, + "epoch": 0.6102326699933867, + "flos": 26687840993280.0, + "grad_norm": 1.8093183996689899, + "language_loss": 0.67931223, + "learning_rate": 1.3927922978594536e-06, + "loss": 0.70141304, + "num_input_tokens_seen": 109354575, + "step": 5075, + "time_per_iteration": 2.6739728450775146 + }, + { + "auxiliary_loss_clip": 0.01076324, + "auxiliary_loss_mlp": 0.01001986, + "balance_loss_clip": 1.01254725, + "balance_loss_mlp": 1.00078821, + "epoch": 0.6103529128840257, + "flos": 60644612551680.0, + "grad_norm": 0.8173677543081892, + "language_loss": 0.5747155, + "learning_rate": 1.3920501410320387e-06, + "loss": 0.59549856, + "num_input_tokens_seen": 109410690, + "step": 5076, + "time_per_iteration": 3.143334150314331 + }, + { + "auxiliary_loss_clip": 0.0119152, + "auxiliary_loss_mlp": 0.01028707, + "balance_loss_clip": 1.04722846, + "balance_loss_mlp": 1.0211612, + "epoch": 0.6104731557746649, + "flos": 19021307806080.0, + "grad_norm": 2.485369227272328, + "language_loss": 0.76058835, + "learning_rate": 1.3913080764301333e-06, + "loss": 0.78279066, + "num_input_tokens_seen": 109427650, + "step": 5077, + "time_per_iteration": 2.6071135997772217 + }, + { + "auxiliary_loss_clip": 0.01214192, + "auxiliary_loss_mlp": 0.01025245, + "balance_loss_clip": 1.0413543, + "balance_loss_mlp": 1.01762748, + "epoch": 0.6105933986653039, + "flos": 23366894083200.0, + "grad_norm": 1.6940330425036256, + "language_loss": 0.71565199, + "learning_rate": 1.3905661041663085e-06, + "loss": 0.73804629, + "num_input_tokens_seen": 109448835, + "step": 5078, + "time_per_iteration": 2.8196568489074707 + }, + { + "auxiliary_loss_clip": 0.0118495, + "auxiliary_loss_mlp": 0.01023829, + "balance_loss_clip": 1.05014527, + "balance_loss_mlp": 1.01594353, + "epoch": 0.610713641555943, + "flos": 34637565006720.0, + "grad_norm": 2.2922764496078165, + "language_loss": 0.65140957, + "learning_rate": 1.389824224353122e-06, + "loss": 0.67349732, + "num_input_tokens_seen": 109470425, + "step": 5079, + "time_per_iteration": 3.6541993618011475 + }, + { + "auxiliary_loss_clip": 0.01179602, + "auxiliary_loss_mlp": 0.01024475, + "balance_loss_clip": 1.04935098, + "balance_loss_mlp": 1.01734638, + "epoch": 0.610833884446582, + "flos": 26646471504000.0, + "grad_norm": 1.4751536701301249, + "language_loss": 0.77104181, + "learning_rate": 1.389082437103115e-06, + "loss": 0.79308259, + "num_input_tokens_seen": 109489695, + "step": 5080, + "time_per_iteration": 3.508945941925049 + }, + { + "auxiliary_loss_clip": 0.01196449, + "auxiliary_loss_mlp": 0.01025394, + "balance_loss_clip": 1.04407072, + "balance_loss_mlp": 1.01783633, + "epoch": 0.6109541273372212, + "flos": 21215126868480.0, + "grad_norm": 1.9935177281664116, + "language_loss": 0.78032601, + "learning_rate": 1.3883407425288172e-06, + "loss": 0.80254447, + "num_input_tokens_seen": 109510030, + "step": 5081, + "time_per_iteration": 2.6616432666778564 + }, + { + "auxiliary_loss_clip": 0.01186384, + "auxiliary_loss_mlp": 0.01025443, + "balance_loss_clip": 1.04501474, + "balance_loss_mlp": 1.01730061, + "epoch": 0.6110743702278603, + "flos": 20084084438400.0, + "grad_norm": 1.9791369060020063, + "language_loss": 0.79759127, + "learning_rate": 1.3875991407427417e-06, + "loss": 0.81970954, + "num_input_tokens_seen": 109528255, + "step": 5082, + "time_per_iteration": 3.6140544414520264 + }, + { + "auxiliary_loss_clip": 0.01106051, + "auxiliary_loss_mlp": 0.01002741, + "balance_loss_clip": 1.01095271, + "balance_loss_mlp": 1.00140595, + "epoch": 0.6111946131184993, + "flos": 68302957438080.0, + "grad_norm": 0.7671254139770796, + "language_loss": 0.5822165, + "learning_rate": 1.38685763185739e-06, + "loss": 0.60330439, + "num_input_tokens_seen": 109581915, + "step": 5083, + "time_per_iteration": 3.243497371673584 + }, + { + "auxiliary_loss_clip": 0.01170989, + "auxiliary_loss_mlp": 0.01028821, + "balance_loss_clip": 1.04817688, + "balance_loss_mlp": 1.02111745, + "epoch": 0.6113148560091385, + "flos": 19937676602880.0, + "grad_norm": 4.393137236577229, + "language_loss": 0.67776847, + "learning_rate": 1.3861162159852476e-06, + "loss": 0.69976664, + "num_input_tokens_seen": 109600050, + "step": 5084, + "time_per_iteration": 3.5655317306518555 + }, + { + "auxiliary_loss_clip": 0.01193747, + "auxiliary_loss_mlp": 0.01024308, + "balance_loss_clip": 1.04707837, + "balance_loss_mlp": 1.01665473, + "epoch": 0.6114350988997775, + "flos": 23731854220800.0, + "grad_norm": 2.031675352004249, + "language_loss": 0.79985571, + "learning_rate": 1.3853748932387875e-06, + "loss": 0.82203627, + "num_input_tokens_seen": 109620690, + "step": 5085, + "time_per_iteration": 2.6523375511169434 + }, + { + "auxiliary_loss_clip": 0.01176336, + "auxiliary_loss_mlp": 0.01026284, + "balance_loss_clip": 1.04342294, + "balance_loss_mlp": 1.01856816, + "epoch": 0.6115553417904166, + "flos": 24023700224640.0, + "grad_norm": 9.754352953506595, + "language_loss": 0.74890554, + "learning_rate": 1.3846336637304671e-06, + "loss": 0.77093184, + "num_input_tokens_seen": 109638960, + "step": 5086, + "time_per_iteration": 2.683394432067871 + }, + { + "auxiliary_loss_clip": 0.01179958, + "auxiliary_loss_mlp": 0.01028094, + "balance_loss_clip": 1.04442263, + "balance_loss_mlp": 1.02041125, + "epoch": 0.6116755846810558, + "flos": 23733542160000.0, + "grad_norm": 2.314824566978744, + "language_loss": 0.83334392, + "learning_rate": 1.3838925275727316e-06, + "loss": 0.8554244, + "num_input_tokens_seen": 109659700, + "step": 5087, + "time_per_iteration": 2.679874897003174 + }, + { + "auxiliary_loss_clip": 0.01176562, + "auxiliary_loss_mlp": 0.01021631, + "balance_loss_clip": 1.05218029, + "balance_loss_mlp": 1.01394749, + "epoch": 0.6117958275716948, + "flos": 18661626967680.0, + "grad_norm": 1.7962126688568691, + "language_loss": 0.79365194, + "learning_rate": 1.3831514848780089e-06, + "loss": 0.81563389, + "num_input_tokens_seen": 109679275, + "step": 5088, + "time_per_iteration": 2.5830085277557373 + }, + { + "auxiliary_loss_clip": 0.01172122, + "auxiliary_loss_mlp": 0.01024073, + "balance_loss_clip": 1.04643321, + "balance_loss_mlp": 1.01657462, + "epoch": 0.6119160704623339, + "flos": 16471183783680.0, + "grad_norm": 2.274642833193415, + "language_loss": 0.9178437, + "learning_rate": 1.3824105357587152e-06, + "loss": 0.93980563, + "num_input_tokens_seen": 109696380, + "step": 5089, + "time_per_iteration": 2.713649034500122 + }, + { + "auxiliary_loss_clip": 0.01183278, + "auxiliary_loss_mlp": 0.01025489, + "balance_loss_clip": 1.04390955, + "balance_loss_mlp": 1.0189352, + "epoch": 0.612036313352973, + "flos": 23915465568000.0, + "grad_norm": 1.5000013504114686, + "language_loss": 0.82451457, + "learning_rate": 1.381669680327253e-06, + "loss": 0.84660226, + "num_input_tokens_seen": 109718060, + "step": 5090, + "time_per_iteration": 2.642984628677368 + }, + { + "auxiliary_loss_clip": 0.01181691, + "auxiliary_loss_mlp": 0.0102294, + "balance_loss_clip": 1.04697752, + "balance_loss_mlp": 1.01533484, + "epoch": 0.6121565562436121, + "flos": 26974766833920.0, + "grad_norm": 1.8719749671390475, + "language_loss": 0.70926392, + "learning_rate": 1.380928918696008e-06, + "loss": 0.73131019, + "num_input_tokens_seen": 109736830, + "step": 5091, + "time_per_iteration": 2.7297470569610596 + }, + { + "auxiliary_loss_clip": 0.01181492, + "auxiliary_loss_mlp": 0.01026742, + "balance_loss_clip": 1.04710603, + "balance_loss_mlp": 1.01877308, + "epoch": 0.6122767991342511, + "flos": 15668867646720.0, + "grad_norm": 2.9303043160459272, + "language_loss": 0.7179873, + "learning_rate": 1.3801882509773548e-06, + "loss": 0.74006969, + "num_input_tokens_seen": 109754690, + "step": 5092, + "time_per_iteration": 2.5594265460968018 + }, + { + "auxiliary_loss_clip": 0.0117635, + "auxiliary_loss_mlp": 0.0102522, + "balance_loss_clip": 1.04663396, + "balance_loss_mlp": 1.01726246, + "epoch": 0.6123970420248903, + "flos": 27964321591680.0, + "grad_norm": 2.997958315054078, + "language_loss": 0.81749403, + "learning_rate": 1.3794476772836503e-06, + "loss": 0.83950973, + "num_input_tokens_seen": 109775790, + "step": 5093, + "time_per_iteration": 2.7277045249938965 + }, + { + "auxiliary_loss_clip": 0.01189713, + "auxiliary_loss_mlp": 0.01025624, + "balance_loss_clip": 1.04456365, + "balance_loss_mlp": 1.0180366, + "epoch": 0.6125172849155294, + "flos": 21468727866240.0, + "grad_norm": 1.6302250960682663, + "language_loss": 0.84438765, + "learning_rate": 1.3787071977272402e-06, + "loss": 0.86654103, + "num_input_tokens_seen": 109795050, + "step": 5094, + "time_per_iteration": 2.680312156677246 + }, + { + "auxiliary_loss_clip": 0.0120681, + "auxiliary_loss_mlp": 0.01027589, + "balance_loss_clip": 1.04766202, + "balance_loss_mlp": 1.01984644, + "epoch": 0.6126375278061684, + "flos": 16248321849600.0, + "grad_norm": 3.112352388427994, + "language_loss": 0.72220224, + "learning_rate": 1.3779668124204535e-06, + "loss": 0.74454623, + "num_input_tokens_seen": 109811465, + "step": 5095, + "time_per_iteration": 2.7014338970184326 + }, + { + "auxiliary_loss_clip": 0.01185308, + "auxiliary_loss_mlp": 0.01024991, + "balance_loss_clip": 1.04805064, + "balance_loss_mlp": 1.01737404, + "epoch": 0.6127577706968076, + "flos": 20448865008000.0, + "grad_norm": 1.6995139642606978, + "language_loss": 0.80895829, + "learning_rate": 1.3772265214756074e-06, + "loss": 0.8310613, + "num_input_tokens_seen": 109831225, + "step": 5096, + "time_per_iteration": 2.8021023273468018 + }, + { + "auxiliary_loss_clip": 0.01183185, + "auxiliary_loss_mlp": 0.01031719, + "balance_loss_clip": 1.04673839, + "balance_loss_mlp": 1.02384567, + "epoch": 0.6128780135874466, + "flos": 18260397072000.0, + "grad_norm": 1.736963717411517, + "language_loss": 0.75153065, + "learning_rate": 1.3764863250050025e-06, + "loss": 0.77367961, + "num_input_tokens_seen": 109849465, + "step": 5097, + "time_per_iteration": 2.686950922012329 + }, + { + "auxiliary_loss_clip": 0.01197337, + "auxiliary_loss_mlp": 0.01026259, + "balance_loss_clip": 1.04356682, + "balance_loss_mlp": 1.01913595, + "epoch": 0.6129982564780857, + "flos": 24937088192640.0, + "grad_norm": 1.7666660196838777, + "language_loss": 0.80423307, + "learning_rate": 1.3757462231209272e-06, + "loss": 0.82646906, + "num_input_tokens_seen": 109869770, + "step": 5098, + "time_per_iteration": 2.7291462421417236 + }, + { + "auxiliary_loss_clip": 0.01181999, + "auxiliary_loss_mlp": 0.01026339, + "balance_loss_clip": 1.04512, + "balance_loss_mlp": 1.01847124, + "epoch": 0.6131184993687249, + "flos": 22492038430080.0, + "grad_norm": 1.9033874671508946, + "language_loss": 0.88775229, + "learning_rate": 1.3750062159356525e-06, + "loss": 0.90983564, + "num_input_tokens_seen": 109889120, + "step": 5099, + "time_per_iteration": 2.694331407546997 + }, + { + "auxiliary_loss_clip": 0.0118888, + "auxiliary_loss_mlp": 0.01026863, + "balance_loss_clip": 1.04579425, + "balance_loss_mlp": 1.01948369, + "epoch": 0.6132387422593639, + "flos": 15885839750400.0, + "grad_norm": 2.083194676091527, + "language_loss": 0.83066964, + "learning_rate": 1.3742663035614382e-06, + "loss": 0.85282707, + "num_input_tokens_seen": 109906490, + "step": 5100, + "time_per_iteration": 2.6084392070770264 + }, + { + "auxiliary_loss_clip": 0.01176335, + "auxiliary_loss_mlp": 0.01028682, + "balance_loss_clip": 1.05204535, + "balance_loss_mlp": 1.02111793, + "epoch": 0.613358985150003, + "flos": 25411539962880.0, + "grad_norm": 1.721489670504918, + "language_loss": 0.80024052, + "learning_rate": 1.3735264861105283e-06, + "loss": 0.82229078, + "num_input_tokens_seen": 109927130, + "step": 5101, + "time_per_iteration": 2.6757237911224365 + }, + { + "auxiliary_loss_clip": 0.01201423, + "auxiliary_loss_mlp": 0.01022993, + "balance_loss_clip": 1.04477859, + "balance_loss_mlp": 1.01597786, + "epoch": 0.6134792280406421, + "flos": 21361283308800.0, + "grad_norm": 1.9940898341591027, + "language_loss": 0.7848472, + "learning_rate": 1.372786763695152e-06, + "loss": 0.80709136, + "num_input_tokens_seen": 109945890, + "step": 5102, + "time_per_iteration": 2.6960439682006836 + }, + { + "auxiliary_loss_clip": 0.01183768, + "auxiliary_loss_mlp": 0.01024315, + "balance_loss_clip": 1.04846549, + "balance_loss_mlp": 1.0166024, + "epoch": 0.6135994709312812, + "flos": 21211248199680.0, + "grad_norm": 1.7653529651390456, + "language_loss": 0.77331781, + "learning_rate": 1.3720471364275257e-06, + "loss": 0.79539865, + "num_input_tokens_seen": 109965535, + "step": 5103, + "time_per_iteration": 2.6634511947631836 + }, + { + "auxiliary_loss_clip": 0.01192643, + "auxiliary_loss_mlp": 0.01378901, + "balance_loss_clip": 1.04327559, + "balance_loss_mlp": 1.00016952, + "epoch": 0.6137197138219203, + "flos": 14794047907200.0, + "grad_norm": 2.1801956929216737, + "language_loss": 0.77733994, + "learning_rate": 1.3713076044198486e-06, + "loss": 0.80305541, + "num_input_tokens_seen": 109982345, + "step": 5104, + "time_per_iteration": 2.7406342029571533 + }, + { + "auxiliary_loss_clip": 0.01182532, + "auxiliary_loss_mlp": 0.01026732, + "balance_loss_clip": 1.04371738, + "balance_loss_mlp": 1.01937437, + "epoch": 0.6138399567125594, + "flos": 20084515401600.0, + "grad_norm": 3.0826928601640757, + "language_loss": 0.81650156, + "learning_rate": 1.3705681677843086e-06, + "loss": 0.8385942, + "num_input_tokens_seen": 110000940, + "step": 5105, + "time_per_iteration": 3.4416747093200684 + }, + { + "auxiliary_loss_clip": 0.01066587, + "auxiliary_loss_mlp": 0.01001492, + "balance_loss_clip": 1.01208782, + "balance_loss_mlp": 1.00031173, + "epoch": 0.6139601996031985, + "flos": 60123838193280.0, + "grad_norm": 0.8032689566612465, + "language_loss": 0.60619599, + "learning_rate": 1.3698288266330768e-06, + "loss": 0.62687677, + "num_input_tokens_seen": 110061565, + "step": 5106, + "time_per_iteration": 4.198812484741211 + }, + { + "auxiliary_loss_clip": 0.01186573, + "auxiliary_loss_mlp": 0.01023369, + "balance_loss_clip": 1.04912055, + "balance_loss_mlp": 1.01629376, + "epoch": 0.6140804424938375, + "flos": 23586703361280.0, + "grad_norm": 2.485637216129516, + "language_loss": 0.72526371, + "learning_rate": 1.3690895810783113e-06, + "loss": 0.74736315, + "num_input_tokens_seen": 110080360, + "step": 5107, + "time_per_iteration": 2.6487507820129395 + }, + { + "auxiliary_loss_clip": 0.01223885, + "auxiliary_loss_mlp": 0.01378797, + "balance_loss_clip": 1.04242408, + "balance_loss_mlp": 1.00019884, + "epoch": 0.6142006853844767, + "flos": 21398199511680.0, + "grad_norm": 1.8420416125770123, + "language_loss": 0.71945, + "learning_rate": 1.3683504312321543e-06, + "loss": 0.74547678, + "num_input_tokens_seen": 110100695, + "step": 5108, + "time_per_iteration": 3.72360897064209 + }, + { + "auxiliary_loss_clip": 0.0118807, + "auxiliary_loss_mlp": 0.01028668, + "balance_loss_clip": 1.05031347, + "balance_loss_mlp": 1.02129769, + "epoch": 0.6143209282751158, + "flos": 12057367622400.0, + "grad_norm": 3.4797204275669085, + "language_loss": 0.80285501, + "learning_rate": 1.3676113772067355e-06, + "loss": 0.82502246, + "num_input_tokens_seen": 110117750, + "step": 5109, + "time_per_iteration": 2.812711715698242 + }, + { + "auxiliary_loss_clip": 0.01213049, + "auxiliary_loss_mlp": 0.0102092, + "balance_loss_clip": 1.04402304, + "balance_loss_mlp": 1.01352334, + "epoch": 0.6144411711657548, + "flos": 25082274965760.0, + "grad_norm": 1.8276451986147315, + "language_loss": 0.72623062, + "learning_rate": 1.3668724191141671e-06, + "loss": 0.74857032, + "num_input_tokens_seen": 110137020, + "step": 5110, + "time_per_iteration": 3.697143077850342 + }, + { + "auxiliary_loss_clip": 0.01199568, + "auxiliary_loss_mlp": 0.01029682, + "balance_loss_clip": 1.05224109, + "balance_loss_mlp": 1.02231824, + "epoch": 0.6145614140563939, + "flos": 20114069316480.0, + "grad_norm": 2.2985379089717997, + "language_loss": 0.66701394, + "learning_rate": 1.3661335570665493e-06, + "loss": 0.68930644, + "num_input_tokens_seen": 110154930, + "step": 5111, + "time_per_iteration": 2.6650357246398926 + }, + { + "auxiliary_loss_clip": 0.011955, + "auxiliary_loss_mlp": 0.01025346, + "balance_loss_clip": 1.04995394, + "balance_loss_mlp": 1.01767516, + "epoch": 0.614681656947033, + "flos": 16800376953600.0, + "grad_norm": 2.4103317310754306, + "language_loss": 0.69864202, + "learning_rate": 1.3653947911759676e-06, + "loss": 0.72085047, + "num_input_tokens_seen": 110172480, + "step": 5112, + "time_per_iteration": 2.624075174331665 + }, + { + "auxiliary_loss_clip": 0.01198549, + "auxiliary_loss_mlp": 0.01035187, + "balance_loss_clip": 1.04340613, + "balance_loss_mlp": 1.02699757, + "epoch": 0.6148018998376721, + "flos": 38801587011840.0, + "grad_norm": 1.8822209725306036, + "language_loss": 0.74515641, + "learning_rate": 1.3646561215544904e-06, + "loss": 0.76749372, + "num_input_tokens_seen": 110197120, + "step": 5113, + "time_per_iteration": 2.8511171340942383 + }, + { + "auxiliary_loss_clip": 0.01182983, + "auxiliary_loss_mlp": 0.01026738, + "balance_loss_clip": 1.04969501, + "balance_loss_mlp": 1.01885188, + "epoch": 0.6149221427283111, + "flos": 23327032965120.0, + "grad_norm": 1.909668272476531, + "language_loss": 0.7924937, + "learning_rate": 1.363917548314176e-06, + "loss": 0.81459093, + "num_input_tokens_seen": 110216385, + "step": 5114, + "time_per_iteration": 2.640557289123535 + }, + { + "auxiliary_loss_clip": 0.01188551, + "auxiliary_loss_mlp": 0.01025365, + "balance_loss_clip": 1.04898512, + "balance_loss_mlp": 1.01695514, + "epoch": 0.6150423856189503, + "flos": 22379494141440.0, + "grad_norm": 1.6356287624445862, + "language_loss": 0.73402876, + "learning_rate": 1.3631790715670626e-06, + "loss": 0.75616789, + "num_input_tokens_seen": 110234790, + "step": 5115, + "time_per_iteration": 2.5900397300720215 + }, + { + "auxiliary_loss_clip": 0.012142, + "auxiliary_loss_mlp": 0.01026335, + "balance_loss_clip": 1.04294419, + "balance_loss_mlp": 1.01924467, + "epoch": 0.6151626285095894, + "flos": 18692078722560.0, + "grad_norm": 2.20372280679702, + "language_loss": 0.85766423, + "learning_rate": 1.3624406914251783e-06, + "loss": 0.88006949, + "num_input_tokens_seen": 110251910, + "step": 5116, + "time_per_iteration": 3.096787691116333 + }, + { + "auxiliary_loss_clip": 0.01183705, + "auxiliary_loss_mlp": 0.01025387, + "balance_loss_clip": 1.0492152, + "balance_loss_mlp": 1.01763225, + "epoch": 0.6152828714002284, + "flos": 15851688894720.0, + "grad_norm": 2.4199471268077057, + "language_loss": 0.88508463, + "learning_rate": 1.3617024080005335e-06, + "loss": 0.90717554, + "num_input_tokens_seen": 110268810, + "step": 5117, + "time_per_iteration": 2.8918299674987793 + }, + { + "auxiliary_loss_clip": 0.0119686, + "auxiliary_loss_mlp": 0.01378729, + "balance_loss_clip": 1.04752922, + "balance_loss_mlp": 1.00018561, + "epoch": 0.6154031142908676, + "flos": 24869792062080.0, + "grad_norm": 1.5497942969039815, + "language_loss": 0.74336016, + "learning_rate": 1.3609642214051266e-06, + "loss": 0.76911604, + "num_input_tokens_seen": 110293035, + "step": 5118, + "time_per_iteration": 2.738150119781494 + }, + { + "auxiliary_loss_clip": 0.01183769, + "auxiliary_loss_mlp": 0.01036892, + "balance_loss_clip": 1.04709935, + "balance_loss_mlp": 1.02918541, + "epoch": 0.6155233571815066, + "flos": 19244744357760.0, + "grad_norm": 2.0867622984444965, + "language_loss": 0.6591289, + "learning_rate": 1.3602261317509385e-06, + "loss": 0.68133551, + "num_input_tokens_seen": 110309695, + "step": 5119, + "time_per_iteration": 2.659801959991455 + }, + { + "auxiliary_loss_clip": 0.01186322, + "auxiliary_loss_mlp": 0.01028259, + "balance_loss_clip": 1.04925919, + "balance_loss_mlp": 1.01990271, + "epoch": 0.6156436000721457, + "flos": 18770077105920.0, + "grad_norm": 2.2272540922537014, + "language_loss": 0.82824814, + "learning_rate": 1.3594881391499387e-06, + "loss": 0.85039395, + "num_input_tokens_seen": 110328610, + "step": 5120, + "time_per_iteration": 2.6043102741241455 + }, + { + "auxiliary_loss_clip": 0.01193015, + "auxiliary_loss_mlp": 0.01024253, + "balance_loss_clip": 1.04754686, + "balance_loss_mlp": 1.0172143, + "epoch": 0.6157638429627849, + "flos": 18041198325120.0, + "grad_norm": 1.6806959095108358, + "language_loss": 0.79018843, + "learning_rate": 1.3587502437140778e-06, + "loss": 0.81236112, + "num_input_tokens_seen": 110346775, + "step": 5121, + "time_per_iteration": 2.6658923625946045 + }, + { + "auxiliary_loss_clip": 0.01195394, + "auxiliary_loss_mlp": 0.01023658, + "balance_loss_clip": 1.04583597, + "balance_loss_mlp": 1.01568925, + "epoch": 0.6158840858534239, + "flos": 25556726736000.0, + "grad_norm": 2.1177290381414386, + "language_loss": 0.85193717, + "learning_rate": 1.3580124455552952e-06, + "loss": 0.87412769, + "num_input_tokens_seen": 110366140, + "step": 5122, + "time_per_iteration": 2.6720995903015137 + }, + { + "auxiliary_loss_clip": 0.01183133, + "auxiliary_loss_mlp": 0.01378718, + "balance_loss_clip": 1.04884291, + "balance_loss_mlp": 1.00017774, + "epoch": 0.616004328744063, + "flos": 24640788902400.0, + "grad_norm": 1.6399556342585868, + "language_loss": 0.87457585, + "learning_rate": 1.3572747447855148e-06, + "loss": 0.90019441, + "num_input_tokens_seen": 110386550, + "step": 5123, + "time_per_iteration": 2.6479170322418213 + }, + { + "auxiliary_loss_clip": 0.01174837, + "auxiliary_loss_mlp": 0.01025683, + "balance_loss_clip": 1.05010653, + "balance_loss_mlp": 1.01769042, + "epoch": 0.6161245716347021, + "flos": 21689686379520.0, + "grad_norm": 1.8110081004635579, + "language_loss": 0.69493949, + "learning_rate": 1.356537141516644e-06, + "loss": 0.71694469, + "num_input_tokens_seen": 110403970, + "step": 5124, + "time_per_iteration": 2.558889627456665 + }, + { + "auxiliary_loss_clip": 0.01183193, + "auxiliary_loss_mlp": 0.01026708, + "balance_loss_clip": 1.05181837, + "balance_loss_mlp": 1.01962709, + "epoch": 0.6162448145253412, + "flos": 35189225061120.0, + "grad_norm": 2.2945202972308456, + "language_loss": 0.61984873, + "learning_rate": 1.3557996358605775e-06, + "loss": 0.64194769, + "num_input_tokens_seen": 110423890, + "step": 5125, + "time_per_iteration": 2.7832791805267334 + }, + { + "auxiliary_loss_clip": 0.01183916, + "auxiliary_loss_mlp": 0.01025539, + "balance_loss_clip": 1.04956472, + "balance_loss_mlp": 1.01807666, + "epoch": 0.6163650574159802, + "flos": 21615279356160.0, + "grad_norm": 4.899384271194756, + "language_loss": 0.70037705, + "learning_rate": 1.3550622279291941e-06, + "loss": 0.72247159, + "num_input_tokens_seen": 110442035, + "step": 5126, + "time_per_iteration": 2.605290174484253 + }, + { + "auxiliary_loss_clip": 0.01194006, + "auxiliary_loss_mlp": 0.01028257, + "balance_loss_clip": 1.04027212, + "balance_loss_mlp": 1.02059782, + "epoch": 0.6164853003066194, + "flos": 24572163968640.0, + "grad_norm": 1.4291043908170555, + "language_loss": 0.83260775, + "learning_rate": 1.354324917834358e-06, + "loss": 0.85483038, + "num_input_tokens_seen": 110463280, + "step": 5127, + "time_per_iteration": 2.7777106761932373 + }, + { + "auxiliary_loss_clip": 0.01209293, + "auxiliary_loss_mlp": 0.01378461, + "balance_loss_clip": 1.040272, + "balance_loss_mlp": 1.00022411, + "epoch": 0.6166055431972585, + "flos": 21835986474240.0, + "grad_norm": 1.8107311463881968, + "language_loss": 0.76944762, + "learning_rate": 1.353587705687918e-06, + "loss": 0.79532516, + "num_input_tokens_seen": 110481455, + "step": 5128, + "time_per_iteration": 2.7619171142578125 + }, + { + "auxiliary_loss_clip": 0.01195599, + "auxiliary_loss_mlp": 0.01028805, + "balance_loss_clip": 1.04921675, + "balance_loss_mlp": 1.02031732, + "epoch": 0.6167257860878975, + "flos": 17785262943360.0, + "grad_norm": 2.3664029386672047, + "language_loss": 0.72562253, + "learning_rate": 1.3528505916017096e-06, + "loss": 0.74786657, + "num_input_tokens_seen": 110499155, + "step": 5129, + "time_per_iteration": 2.6443023681640625 + }, + { + "auxiliary_loss_clip": 0.01181212, + "auxiliary_loss_mlp": 0.01030788, + "balance_loss_clip": 1.0479455, + "balance_loss_mlp": 1.02354646, + "epoch": 0.6168460289785367, + "flos": 23214811898880.0, + "grad_norm": 2.361397011889762, + "language_loss": 0.88605243, + "learning_rate": 1.3521135756875514e-06, + "loss": 0.90817249, + "num_input_tokens_seen": 110515470, + "step": 5130, + "time_per_iteration": 3.6771655082702637 + }, + { + "auxiliary_loss_clip": 0.01206604, + "auxiliary_loss_mlp": 0.01024011, + "balance_loss_clip": 1.04261732, + "balance_loss_mlp": 1.01724613, + "epoch": 0.6169662718691757, + "flos": 26213281482240.0, + "grad_norm": 1.5162086505347896, + "language_loss": 0.86273819, + "learning_rate": 1.3513766580572496e-06, + "loss": 0.8850444, + "num_input_tokens_seen": 110538290, + "step": 5131, + "time_per_iteration": 3.9729108810424805 + }, + { + "auxiliary_loss_clip": 0.01181949, + "auxiliary_loss_mlp": 0.01025324, + "balance_loss_clip": 1.04919302, + "balance_loss_mlp": 1.01792097, + "epoch": 0.6170865147598148, + "flos": 19026120228480.0, + "grad_norm": 2.021510014672573, + "language_loss": 0.7734825, + "learning_rate": 1.3506398388225924e-06, + "loss": 0.79555523, + "num_input_tokens_seen": 110555610, + "step": 5132, + "time_per_iteration": 2.645951509475708 + }, + { + "auxiliary_loss_clip": 0.01175886, + "auxiliary_loss_mlp": 0.01025782, + "balance_loss_clip": 1.0534687, + "balance_loss_mlp": 1.01831698, + "epoch": 0.617206757650454, + "flos": 18260361158400.0, + "grad_norm": 1.7819829899374586, + "language_loss": 0.71778715, + "learning_rate": 1.349903118095355e-06, + "loss": 0.73980385, + "num_input_tokens_seen": 110574745, + "step": 5133, + "time_per_iteration": 3.9083869457244873 + }, + { + "auxiliary_loss_clip": 0.01186122, + "auxiliary_loss_mlp": 0.0102442, + "balance_loss_clip": 1.04883623, + "balance_loss_mlp": 1.01681137, + "epoch": 0.617327000541093, + "flos": 18186959715840.0, + "grad_norm": 1.706705704206341, + "language_loss": 0.73153609, + "learning_rate": 1.349166495987298e-06, + "loss": 0.75364149, + "num_input_tokens_seen": 110593310, + "step": 5134, + "time_per_iteration": 2.6564929485321045 + }, + { + "auxiliary_loss_clip": 0.01093892, + "auxiliary_loss_mlp": 0.01000578, + "balance_loss_clip": 1.01944447, + "balance_loss_mlp": 0.99959415, + "epoch": 0.6174472434317321, + "flos": 61833796122240.0, + "grad_norm": 0.823059213632666, + "language_loss": 0.6091435, + "learning_rate": 1.348429972610166e-06, + "loss": 0.63008821, + "num_input_tokens_seen": 110657615, + "step": 5135, + "time_per_iteration": 3.3525495529174805 + }, + { + "auxiliary_loss_clip": 0.01123614, + "auxiliary_loss_mlp": 0.01001629, + "balance_loss_clip": 1.02712321, + "balance_loss_mlp": 1.00055063, + "epoch": 0.6175674863223712, + "flos": 71230970494080.0, + "grad_norm": 0.8426185508015155, + "language_loss": 0.57787406, + "learning_rate": 1.3476935480756897e-06, + "loss": 0.59912646, + "num_input_tokens_seen": 110714365, + "step": 5136, + "time_per_iteration": 4.045832395553589 + }, + { + "auxiliary_loss_clip": 0.01187469, + "auxiliary_loss_mlp": 0.01022583, + "balance_loss_clip": 1.04332304, + "balance_loss_mlp": 1.01477516, + "epoch": 0.6176877292130103, + "flos": 21835447770240.0, + "grad_norm": 2.40913327992124, + "language_loss": 0.75473905, + "learning_rate": 1.346957222495583e-06, + "loss": 0.77683961, + "num_input_tokens_seen": 110732160, + "step": 5137, + "time_per_iteration": 2.676870346069336 + }, + { + "auxiliary_loss_clip": 0.01199487, + "auxiliary_loss_mlp": 0.01379, + "balance_loss_clip": 1.05066347, + "balance_loss_mlp": 1.00022733, + "epoch": 0.6178079721036493, + "flos": 17741738638080.0, + "grad_norm": 2.7030913390240814, + "language_loss": 0.71048039, + "learning_rate": 1.3462209959815466e-06, + "loss": 0.7362653, + "num_input_tokens_seen": 110746900, + "step": 5138, + "time_per_iteration": 2.5954928398132324 + }, + { + "auxiliary_loss_clip": 0.01193426, + "auxiliary_loss_mlp": 0.01029509, + "balance_loss_clip": 1.04933, + "balance_loss_mlp": 1.02235031, + "epoch": 0.6179282149942885, + "flos": 22633131052800.0, + "grad_norm": 2.2283209920860156, + "language_loss": 0.74572527, + "learning_rate": 1.345484868645265e-06, + "loss": 0.76795459, + "num_input_tokens_seen": 110765710, + "step": 5139, + "time_per_iteration": 2.6524603366851807 + }, + { + "auxiliary_loss_clip": 0.01210381, + "auxiliary_loss_mlp": 0.01022287, + "balance_loss_clip": 1.04815245, + "balance_loss_mlp": 1.01477623, + "epoch": 0.6180484578849276, + "flos": 22310330503680.0, + "grad_norm": 2.017629815552318, + "language_loss": 0.78634953, + "learning_rate": 1.3447488405984088e-06, + "loss": 0.80867624, + "num_input_tokens_seen": 110783970, + "step": 5140, + "time_per_iteration": 2.7033097743988037 + }, + { + "auxiliary_loss_clip": 0.01191592, + "auxiliary_loss_mlp": 0.01027052, + "balance_loss_clip": 1.04808617, + "balance_loss_mlp": 1.01871955, + "epoch": 0.6181687007755666, + "flos": 35225458905600.0, + "grad_norm": 2.248293122738211, + "language_loss": 0.70183671, + "learning_rate": 1.3440129119526322e-06, + "loss": 0.7240231, + "num_input_tokens_seen": 110806395, + "step": 5141, + "time_per_iteration": 2.769543409347534 + }, + { + "auxiliary_loss_clip": 0.01067573, + "auxiliary_loss_mlp": 0.01000016, + "balance_loss_clip": 1.01352811, + "balance_loss_mlp": 0.99881756, + "epoch": 0.6182889436662057, + "flos": 61547370094080.0, + "grad_norm": 0.8227258310719788, + "language_loss": 0.51213521, + "learning_rate": 1.3432770828195762e-06, + "loss": 0.53281111, + "num_input_tokens_seen": 110867380, + "step": 5142, + "time_per_iteration": 3.3066136837005615 + }, + { + "auxiliary_loss_clip": 0.0118758, + "auxiliary_loss_mlp": 0.01031435, + "balance_loss_clip": 1.0430665, + "balance_loss_mlp": 1.02383518, + "epoch": 0.6184091865568448, + "flos": 19609991804160.0, + "grad_norm": 4.713924142381766, + "language_loss": 0.70317042, + "learning_rate": 1.3425413533108635e-06, + "loss": 0.72536063, + "num_input_tokens_seen": 110885980, + "step": 5143, + "time_per_iteration": 2.708279609680176 + }, + { + "auxiliary_loss_clip": 0.01209281, + "auxiliary_loss_mlp": 0.01027046, + "balance_loss_clip": 1.04683518, + "balance_loss_mlp": 1.01951766, + "epoch": 0.6185294294474839, + "flos": 23586882929280.0, + "grad_norm": 3.166117295054762, + "language_loss": 0.70824796, + "learning_rate": 1.341805723538105e-06, + "loss": 0.73061121, + "num_input_tokens_seen": 110906085, + "step": 5144, + "time_per_iteration": 2.739205837249756 + }, + { + "auxiliary_loss_clip": 0.01200849, + "auxiliary_loss_mlp": 0.01030241, + "balance_loss_clip": 1.05024672, + "balance_loss_mlp": 1.02304399, + "epoch": 0.618649672338123, + "flos": 26762032535040.0, + "grad_norm": 1.5211366705450422, + "language_loss": 0.77755296, + "learning_rate": 1.3410701936128948e-06, + "loss": 0.79986393, + "num_input_tokens_seen": 110928865, + "step": 5145, + "time_per_iteration": 2.982651472091675 + }, + { + "auxiliary_loss_clip": 0.01182527, + "auxiliary_loss_mlp": 0.01024856, + "balance_loss_clip": 1.05084515, + "balance_loss_mlp": 1.01747131, + "epoch": 0.6187699152287621, + "flos": 14456630522880.0, + "grad_norm": 2.4659922669281262, + "language_loss": 0.84959018, + "learning_rate": 1.340334763646812e-06, + "loss": 0.87166399, + "num_input_tokens_seen": 110943000, + "step": 5146, + "time_per_iteration": 2.591214656829834 + }, + { + "auxiliary_loss_clip": 0.01175882, + "auxiliary_loss_mlp": 0.01027067, + "balance_loss_clip": 1.05078149, + "balance_loss_mlp": 1.01921678, + "epoch": 0.6188901581194012, + "flos": 20084766796800.0, + "grad_norm": 1.6762856864615612, + "language_loss": 0.74346697, + "learning_rate": 1.3395994337514218e-06, + "loss": 0.76549643, + "num_input_tokens_seen": 110963170, + "step": 5147, + "time_per_iteration": 2.602576494216919 + }, + { + "auxiliary_loss_clip": 0.01174102, + "auxiliary_loss_mlp": 0.01024196, + "balance_loss_clip": 1.04650533, + "balance_loss_mlp": 1.0168463, + "epoch": 0.6190104010100402, + "flos": 25700728360320.0, + "grad_norm": 1.4695195519074824, + "language_loss": 0.78674662, + "learning_rate": 1.3388642040382725e-06, + "loss": 0.80872965, + "num_input_tokens_seen": 110983595, + "step": 5148, + "time_per_iteration": 2.6578121185302734 + }, + { + "auxiliary_loss_clip": 0.0120125, + "auxiliary_loss_mlp": 0.01026983, + "balance_loss_clip": 1.04354715, + "balance_loss_mlp": 1.01893687, + "epoch": 0.6191306439006794, + "flos": 30442372974720.0, + "grad_norm": 1.7272305114690225, + "language_loss": 0.8407461, + "learning_rate": 1.3381290746188975e-06, + "loss": 0.86302847, + "num_input_tokens_seen": 111002965, + "step": 5149, + "time_per_iteration": 2.7598092555999756 + }, + { + "auxiliary_loss_clip": 0.01186567, + "auxiliary_loss_mlp": 0.01028983, + "balance_loss_clip": 1.05399191, + "balance_loss_mlp": 1.02147841, + "epoch": 0.6192508867913185, + "flos": 26685793918080.0, + "grad_norm": 1.7466585883977308, + "language_loss": 0.6753599, + "learning_rate": 1.3373940456048152e-06, + "loss": 0.69751543, + "num_input_tokens_seen": 111022990, + "step": 5150, + "time_per_iteration": 2.7393763065338135 + }, + { + "auxiliary_loss_clip": 0.01174408, + "auxiliary_loss_mlp": 0.0102665, + "balance_loss_clip": 1.05140138, + "balance_loss_mlp": 1.01922023, + "epoch": 0.6193711296819575, + "flos": 36722036090880.0, + "grad_norm": 1.6344703864940175, + "language_loss": 0.58785444, + "learning_rate": 1.3366591171075299e-06, + "loss": 0.60986507, + "num_input_tokens_seen": 111046495, + "step": 5151, + "time_per_iteration": 2.6933345794677734 + }, + { + "auxiliary_loss_clip": 0.01192424, + "auxiliary_loss_mlp": 0.0102285, + "balance_loss_clip": 1.04845548, + "balance_loss_mlp": 1.01503563, + "epoch": 0.6194913725725967, + "flos": 25192556697600.0, + "grad_norm": 1.873256592268319, + "language_loss": 0.90886879, + "learning_rate": 1.335924289238529e-06, + "loss": 0.93102151, + "num_input_tokens_seen": 111065705, + "step": 5152, + "time_per_iteration": 2.722119092941284 + }, + { + "auxiliary_loss_clip": 0.01183082, + "auxiliary_loss_mlp": 0.01378737, + "balance_loss_clip": 1.05117917, + "balance_loss_mlp": 1.0001359, + "epoch": 0.6196116154632357, + "flos": 21178821196800.0, + "grad_norm": 1.7591285928085951, + "language_loss": 0.76943398, + "learning_rate": 1.3351895621092859e-06, + "loss": 0.79505217, + "num_input_tokens_seen": 111086050, + "step": 5153, + "time_per_iteration": 2.6320583820343018 + }, + { + "auxiliary_loss_clip": 0.01218142, + "auxiliary_loss_mlp": 0.01021889, + "balance_loss_clip": 1.03419352, + "balance_loss_mlp": 1.01490951, + "epoch": 0.6197318583538748, + "flos": 16253744803200.0, + "grad_norm": 1.8313105254160977, + "language_loss": 0.76173019, + "learning_rate": 1.3344549358312567e-06, + "loss": 0.78413051, + "num_input_tokens_seen": 111104450, + "step": 5154, + "time_per_iteration": 3.093158721923828 + }, + { + "auxiliary_loss_clip": 0.01187607, + "auxiliary_loss_mlp": 0.01029085, + "balance_loss_clip": 1.04987204, + "balance_loss_mlp": 1.02124131, + "epoch": 0.619852101244514, + "flos": 24425612478720.0, + "grad_norm": 1.857187501535294, + "language_loss": 0.77948499, + "learning_rate": 1.3337204105158852e-06, + "loss": 0.80165195, + "num_input_tokens_seen": 111123320, + "step": 5155, + "time_per_iteration": 3.27028489112854 + }, + { + "auxiliary_loss_clip": 0.01179333, + "auxiliary_loss_mlp": 0.010215, + "balance_loss_clip": 1.03657222, + "balance_loss_mlp": 1.01401925, + "epoch": 0.619972344135153, + "flos": 16727298733440.0, + "grad_norm": 2.3906227007750567, + "language_loss": 0.73070669, + "learning_rate": 1.332985986274597e-06, + "loss": 0.75271499, + "num_input_tokens_seen": 111140950, + "step": 5156, + "time_per_iteration": 3.6972100734710693 + }, + { + "auxiliary_loss_clip": 0.01209803, + "auxiliary_loss_mlp": 0.01378241, + "balance_loss_clip": 1.04413402, + "balance_loss_mlp": 1.00015473, + "epoch": 0.6200925870257921, + "flos": 12495190498560.0, + "grad_norm": 2.053960528551534, + "language_loss": 0.75364745, + "learning_rate": 1.3322516632188047e-06, + "loss": 0.7795279, + "num_input_tokens_seen": 111157845, + "step": 5157, + "time_per_iteration": 3.922154188156128 + }, + { + "auxiliary_loss_clip": 0.01197501, + "auxiliary_loss_mlp": 0.0102687, + "balance_loss_clip": 1.04558218, + "balance_loss_mlp": 1.01906204, + "epoch": 0.6202128299164312, + "flos": 26539350168960.0, + "grad_norm": 1.7170512008698409, + "language_loss": 0.67091703, + "learning_rate": 1.3315174414599045e-06, + "loss": 0.69316071, + "num_input_tokens_seen": 111179165, + "step": 5158, + "time_per_iteration": 2.8378539085388184 + }, + { + "auxiliary_loss_clip": 0.0117546, + "auxiliary_loss_mlp": 0.01025464, + "balance_loss_clip": 1.04557014, + "balance_loss_mlp": 1.01751304, + "epoch": 0.6203330728070703, + "flos": 18770508069120.0, + "grad_norm": 2.1327102046170885, + "language_loss": 0.75260139, + "learning_rate": 1.3307833211092768e-06, + "loss": 0.77461064, + "num_input_tokens_seen": 111197830, + "step": 5159, + "time_per_iteration": 3.532038450241089 + }, + { + "auxiliary_loss_clip": 0.01178551, + "auxiliary_loss_mlp": 0.01029812, + "balance_loss_clip": 1.05437493, + "balance_loss_mlp": 1.02231646, + "epoch": 0.6204533156977093, + "flos": 20629782835200.0, + "grad_norm": 1.606107493120828, + "language_loss": 0.7534138, + "learning_rate": 1.3300493022782873e-06, + "loss": 0.7754975, + "num_input_tokens_seen": 111218400, + "step": 5160, + "time_per_iteration": 2.656130313873291 + }, + { + "auxiliary_loss_clip": 0.01200157, + "auxiliary_loss_mlp": 0.01379021, + "balance_loss_clip": 1.0446316, + "balance_loss_mlp": 1.00017333, + "epoch": 0.6205735585883485, + "flos": 17348050598400.0, + "grad_norm": 1.7482775670459496, + "language_loss": 0.72476745, + "learning_rate": 1.3293153850782855e-06, + "loss": 0.75055921, + "num_input_tokens_seen": 111236720, + "step": 5161, + "time_per_iteration": 2.683941602706909 + }, + { + "auxiliary_loss_clip": 0.01189932, + "auxiliary_loss_mlp": 0.01034588, + "balance_loss_clip": 1.04538894, + "balance_loss_mlp": 1.02539706, + "epoch": 0.6206938014789876, + "flos": 22965017742720.0, + "grad_norm": 1.871623836354955, + "language_loss": 0.716039, + "learning_rate": 1.3285815696206069e-06, + "loss": 0.73828423, + "num_input_tokens_seen": 111258265, + "step": 5162, + "time_per_iteration": 3.624229907989502 + }, + { + "auxiliary_loss_clip": 0.01203156, + "auxiliary_loss_mlp": 0.01029716, + "balance_loss_clip": 1.04465699, + "balance_loss_mlp": 1.02172875, + "epoch": 0.6208140443696266, + "flos": 23983192661760.0, + "grad_norm": 1.9810709675658753, + "language_loss": 0.77086061, + "learning_rate": 1.32784785601657e-06, + "loss": 0.79318935, + "num_input_tokens_seen": 111277675, + "step": 5163, + "time_per_iteration": 2.721118927001953 + }, + { + "auxiliary_loss_clip": 0.01192453, + "auxiliary_loss_mlp": 0.01023874, + "balance_loss_clip": 1.04518294, + "balance_loss_mlp": 1.01620817, + "epoch": 0.6209342872602658, + "flos": 35077291303680.0, + "grad_norm": 1.6445764700620409, + "language_loss": 0.73640537, + "learning_rate": 1.3271142443774798e-06, + "loss": 0.75856864, + "num_input_tokens_seen": 111299910, + "step": 5164, + "time_per_iteration": 2.7459115982055664 + }, + { + "auxiliary_loss_clip": 0.01190551, + "auxiliary_loss_mlp": 0.01029867, + "balance_loss_clip": 1.04888606, + "balance_loss_mlp": 1.02274442, + "epoch": 0.6210545301509048, + "flos": 26979327861120.0, + "grad_norm": 1.7754650799714409, + "language_loss": 0.81347287, + "learning_rate": 1.3263807348146228e-06, + "loss": 0.83567703, + "num_input_tokens_seen": 111319765, + "step": 5165, + "time_per_iteration": 2.6507978439331055 + }, + { + "auxiliary_loss_clip": 0.01186997, + "auxiliary_loss_mlp": 0.01027378, + "balance_loss_clip": 1.044402, + "balance_loss_mlp": 1.01944411, + "epoch": 0.6211747730415439, + "flos": 33618240852480.0, + "grad_norm": 2.246210434146957, + "language_loss": 0.73284334, + "learning_rate": 1.3256473274392733e-06, + "loss": 0.75498712, + "num_input_tokens_seen": 111341110, + "step": 5166, + "time_per_iteration": 2.775635004043579 + }, + { + "auxiliary_loss_clip": 0.01172637, + "auxiliary_loss_mlp": 0.01027741, + "balance_loss_clip": 1.05037594, + "balance_loss_mlp": 1.01976633, + "epoch": 0.6212950159321831, + "flos": 34167099646080.0, + "grad_norm": 1.7690717092257633, + "language_loss": 0.70183599, + "learning_rate": 1.3249140223626873e-06, + "loss": 0.72383976, + "num_input_tokens_seen": 111362730, + "step": 5167, + "time_per_iteration": 2.7139203548431396 + }, + { + "auxiliary_loss_clip": 0.01181559, + "auxiliary_loss_mlp": 0.01025169, + "balance_loss_clip": 1.05069375, + "balance_loss_mlp": 1.01753712, + "epoch": 0.6214152588228221, + "flos": 27965758135680.0, + "grad_norm": 14.189811161771084, + "language_loss": 0.75460219, + "learning_rate": 1.3241808196961077e-06, + "loss": 0.77666944, + "num_input_tokens_seen": 111383855, + "step": 5168, + "time_per_iteration": 2.641932964324951 + }, + { + "auxiliary_loss_clip": 0.01175226, + "auxiliary_loss_mlp": 0.0102021, + "balance_loss_clip": 1.04381752, + "balance_loss_mlp": 1.01290822, + "epoch": 0.6215355017134612, + "flos": 20230204965120.0, + "grad_norm": 1.7335162285496482, + "language_loss": 0.70615673, + "learning_rate": 1.3234477195507608e-06, + "loss": 0.72811103, + "num_input_tokens_seen": 111402685, + "step": 5169, + "time_per_iteration": 2.7155826091766357 + }, + { + "auxiliary_loss_clip": 0.01199086, + "auxiliary_loss_mlp": 0.01028219, + "balance_loss_clip": 1.04601896, + "balance_loss_mlp": 1.02049994, + "epoch": 0.6216557446041003, + "flos": 41428129219200.0, + "grad_norm": 1.9112270000440486, + "language_loss": 0.62692326, + "learning_rate": 1.322714722037857e-06, + "loss": 0.64919627, + "num_input_tokens_seen": 111424130, + "step": 5170, + "time_per_iteration": 2.8688580989837646 + }, + { + "auxiliary_loss_clip": 0.01207077, + "auxiliary_loss_mlp": 0.01026324, + "balance_loss_clip": 1.04547358, + "balance_loss_mlp": 1.01822925, + "epoch": 0.6217759874947394, + "flos": 27928770105600.0, + "grad_norm": 2.177324950345139, + "language_loss": 0.77388656, + "learning_rate": 1.321981827268591e-06, + "loss": 0.79622054, + "num_input_tokens_seen": 111444785, + "step": 5171, + "time_per_iteration": 2.8007876873016357 + }, + { + "auxiliary_loss_clip": 0.01194714, + "auxiliary_loss_mlp": 0.01031608, + "balance_loss_clip": 1.04769838, + "balance_loss_mlp": 1.02379417, + "epoch": 0.6218962303853784, + "flos": 21765673601280.0, + "grad_norm": 2.05100687237423, + "language_loss": 0.81313097, + "learning_rate": 1.3212490353541426e-06, + "loss": 0.8353942, + "num_input_tokens_seen": 111467045, + "step": 5172, + "time_per_iteration": 2.716766119003296 + }, + { + "auxiliary_loss_clip": 0.01175151, + "auxiliary_loss_mlp": 0.01023962, + "balance_loss_clip": 1.04966474, + "balance_loss_mlp": 1.0157547, + "epoch": 0.6220164732760175, + "flos": 21246260981760.0, + "grad_norm": 2.2553390646870306, + "language_loss": 0.80363452, + "learning_rate": 1.3205163464056762e-06, + "loss": 0.82562566, + "num_input_tokens_seen": 111483650, + "step": 5173, + "time_per_iteration": 2.6251816749572754 + }, + { + "auxiliary_loss_clip": 0.01182079, + "auxiliary_loss_mlp": 0.01025501, + "balance_loss_clip": 1.04955494, + "balance_loss_mlp": 1.01776385, + "epoch": 0.6221367161666567, + "flos": 26136360506880.0, + "grad_norm": 1.799223625794307, + "language_loss": 0.72932655, + "learning_rate": 1.319783760534339e-06, + "loss": 0.75140238, + "num_input_tokens_seen": 111502895, + "step": 5174, + "time_per_iteration": 2.733595848083496 + }, + { + "auxiliary_loss_clip": 0.0118531, + "auxiliary_loss_mlp": 0.010264, + "balance_loss_clip": 1.0507524, + "balance_loss_mlp": 1.01909876, + "epoch": 0.6222569590572957, + "flos": 16284196558080.0, + "grad_norm": 2.0977272967125553, + "language_loss": 0.75665259, + "learning_rate": 1.319051277851266e-06, + "loss": 0.77876967, + "num_input_tokens_seen": 111519180, + "step": 5175, + "time_per_iteration": 2.5563876628875732 + }, + { + "auxiliary_loss_clip": 0.01185664, + "auxiliary_loss_mlp": 0.01025376, + "balance_loss_clip": 1.04924834, + "balance_loss_mlp": 1.01806867, + "epoch": 0.6223772019479348, + "flos": 18223840005120.0, + "grad_norm": 1.931753746510616, + "language_loss": 0.84059846, + "learning_rate": 1.3183188984675716e-06, + "loss": 0.86270893, + "num_input_tokens_seen": 111537545, + "step": 5176, + "time_per_iteration": 2.7044737339019775 + }, + { + "auxiliary_loss_clip": 0.0119309, + "auxiliary_loss_mlp": 0.0102706, + "balance_loss_clip": 1.05030107, + "balance_loss_mlp": 1.01933527, + "epoch": 0.6224974448385739, + "flos": 27489797994240.0, + "grad_norm": 2.274730646338194, + "language_loss": 0.7163294, + "learning_rate": 1.3175866224943586e-06, + "loss": 0.73853087, + "num_input_tokens_seen": 111556265, + "step": 5177, + "time_per_iteration": 2.7190818786621094 + }, + { + "auxiliary_loss_clip": 0.01198101, + "auxiliary_loss_mlp": 0.01020959, + "balance_loss_clip": 1.04938269, + "balance_loss_mlp": 1.01315033, + "epoch": 0.622617687729213, + "flos": 19791951125760.0, + "grad_norm": 2.3423198459202483, + "language_loss": 0.73369193, + "learning_rate": 1.316854450042712e-06, + "loss": 0.75588256, + "num_input_tokens_seen": 111574205, + "step": 5178, + "time_per_iteration": 2.7750589847564697 + }, + { + "auxiliary_loss_clip": 0.01188876, + "auxiliary_loss_mlp": 0.01025939, + "balance_loss_clip": 1.05075741, + "balance_loss_mlp": 1.01834512, + "epoch": 0.622737930619852, + "flos": 23038886062080.0, + "grad_norm": 1.8655015478362436, + "language_loss": 0.74369597, + "learning_rate": 1.3161223812237024e-06, + "loss": 0.76584411, + "num_input_tokens_seen": 111593560, + "step": 5179, + "time_per_iteration": 2.641402244567871 + }, + { + "auxiliary_loss_clip": 0.01176557, + "auxiliary_loss_mlp": 0.01026862, + "balance_loss_clip": 1.05145407, + "balance_loss_mlp": 1.01925039, + "epoch": 0.6228581735104912, + "flos": 12634271959680.0, + "grad_norm": 2.717351238897899, + "language_loss": 0.85055584, + "learning_rate": 1.3153904161483842e-06, + "loss": 0.87259007, + "num_input_tokens_seen": 111608860, + "step": 5180, + "time_per_iteration": 2.603398561477661 + }, + { + "auxiliary_loss_clip": 0.0119375, + "auxiliary_loss_mlp": 0.01031454, + "balance_loss_clip": 1.04445195, + "balance_loss_mlp": 1.02345526, + "epoch": 0.6229784164011303, + "flos": 23802813538560.0, + "grad_norm": 2.1550715086965626, + "language_loss": 0.85427809, + "learning_rate": 1.3146585549277953e-06, + "loss": 0.87653011, + "num_input_tokens_seen": 111627500, + "step": 5181, + "time_per_iteration": 2.6604533195495605 + }, + { + "auxiliary_loss_clip": 0.01204973, + "auxiliary_loss_mlp": 0.01031619, + "balance_loss_clip": 1.05243194, + "balance_loss_mlp": 1.02398944, + "epoch": 0.6230986592917693, + "flos": 22414219614720.0, + "grad_norm": 17.822400485630958, + "language_loss": 0.78122318, + "learning_rate": 1.3139267976729591e-06, + "loss": 0.80358911, + "num_input_tokens_seen": 111647690, + "step": 5182, + "time_per_iteration": 3.6542258262634277 + }, + { + "auxiliary_loss_clip": 0.01185941, + "auxiliary_loss_mlp": 0.01029884, + "balance_loss_clip": 1.04966617, + "balance_loss_mlp": 1.02230263, + "epoch": 0.6232189021824085, + "flos": 34528217028480.0, + "grad_norm": 2.1047327354044354, + "language_loss": 0.72056055, + "learning_rate": 1.3131951444948815e-06, + "loss": 0.74271882, + "num_input_tokens_seen": 111667090, + "step": 5183, + "time_per_iteration": 3.6328723430633545 + }, + { + "auxiliary_loss_clip": 0.01200132, + "auxiliary_loss_mlp": 0.01027381, + "balance_loss_clip": 1.05231929, + "balance_loss_mlp": 1.01970959, + "epoch": 0.6233391450730476, + "flos": 22237000888320.0, + "grad_norm": 1.7986483899863934, + "language_loss": 0.76188529, + "learning_rate": 1.3124635955045546e-06, + "loss": 0.78416038, + "num_input_tokens_seen": 111686905, + "step": 5184, + "time_per_iteration": 2.602656126022339 + }, + { + "auxiliary_loss_clip": 0.01194384, + "auxiliary_loss_mlp": 0.01378342, + "balance_loss_clip": 1.04065526, + "balance_loss_mlp": 1.00018692, + "epoch": 0.6234593879636866, + "flos": 20332693445760.0, + "grad_norm": 2.7685266920510916, + "language_loss": 0.84358108, + "learning_rate": 1.3117321508129537e-06, + "loss": 0.86930829, + "num_input_tokens_seen": 111704985, + "step": 5185, + "time_per_iteration": 3.5603275299072266 + }, + { + "auxiliary_loss_clip": 0.01193786, + "auxiliary_loss_mlp": 0.01026878, + "balance_loss_clip": 1.04833984, + "balance_loss_mlp": 1.01906967, + "epoch": 0.6235796308543258, + "flos": 20664903358080.0, + "grad_norm": 1.5460819619228225, + "language_loss": 0.76508683, + "learning_rate": 1.3110008105310388e-06, + "loss": 0.78729349, + "num_input_tokens_seen": 111724805, + "step": 5186, + "time_per_iteration": 2.6464807987213135 + }, + { + "auxiliary_loss_clip": 0.01175365, + "auxiliary_loss_mlp": 0.01027489, + "balance_loss_clip": 1.04893923, + "balance_loss_mlp": 1.0196147, + "epoch": 0.6236998737449648, + "flos": 26618641441920.0, + "grad_norm": 1.6218849120885381, + "language_loss": 0.78060567, + "learning_rate": 1.3102695747697526e-06, + "loss": 0.80263424, + "num_input_tokens_seen": 111747675, + "step": 5187, + "time_per_iteration": 2.570676803588867 + }, + { + "auxiliary_loss_clip": 0.01224286, + "auxiliary_loss_mlp": 0.01024131, + "balance_loss_clip": 1.04968584, + "balance_loss_mlp": 1.01633501, + "epoch": 0.6238201166356039, + "flos": 12674599954560.0, + "grad_norm": 2.237321086382213, + "language_loss": 0.90538764, + "learning_rate": 1.3095384436400237e-06, + "loss": 0.92787182, + "num_input_tokens_seen": 111759205, + "step": 5188, + "time_per_iteration": 3.5215935707092285 + }, + { + "auxiliary_loss_clip": 0.01197792, + "auxiliary_loss_mlp": 0.01026844, + "balance_loss_clip": 1.04695964, + "balance_loss_mlp": 1.01930368, + "epoch": 0.623940359526243, + "flos": 10452160730880.0, + "grad_norm": 3.5238443679630764, + "language_loss": 0.82209003, + "learning_rate": 1.3088074172527633e-06, + "loss": 0.84433639, + "num_input_tokens_seen": 111776335, + "step": 5189, + "time_per_iteration": 2.580531120300293 + }, + { + "auxiliary_loss_clip": 0.01199, + "auxiliary_loss_mlp": 0.01022363, + "balance_loss_clip": 1.04784381, + "balance_loss_mlp": 1.01411963, + "epoch": 0.6240606024168821, + "flos": 29059525226880.0, + "grad_norm": 1.859160386219143, + "language_loss": 0.71456736, + "learning_rate": 1.3080764957188684e-06, + "loss": 0.73678094, + "num_input_tokens_seen": 111796580, + "step": 5190, + "time_per_iteration": 2.676900863647461 + }, + { + "auxiliary_loss_clip": 0.01210852, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 1.04462934, + "balance_loss_mlp": 1.01999617, + "epoch": 0.6241808453075212, + "flos": 22018089450240.0, + "grad_norm": 1.7933605074613987, + "language_loss": 0.70936644, + "learning_rate": 1.3073456791492192e-06, + "loss": 0.73175341, + "num_input_tokens_seen": 111816290, + "step": 5191, + "time_per_iteration": 2.66233491897583 + }, + { + "auxiliary_loss_clip": 0.01195736, + "auxiliary_loss_mlp": 0.01023097, + "balance_loss_clip": 1.04792356, + "balance_loss_mlp": 1.01563191, + "epoch": 0.6243010881981603, + "flos": 21138708683520.0, + "grad_norm": 1.787011191256065, + "language_loss": 0.78299779, + "learning_rate": 1.3066149676546801e-06, + "loss": 0.80518615, + "num_input_tokens_seen": 111834470, + "step": 5192, + "time_per_iteration": 2.6589744091033936 + }, + { + "auxiliary_loss_clip": 0.01190414, + "auxiliary_loss_mlp": 0.01032573, + "balance_loss_clip": 1.05000281, + "balance_loss_mlp": 1.02535796, + "epoch": 0.6244213310887994, + "flos": 22344948236160.0, + "grad_norm": 1.9453009224763758, + "language_loss": 0.66107112, + "learning_rate": 1.3058843613460985e-06, + "loss": 0.68330097, + "num_input_tokens_seen": 111852410, + "step": 5193, + "time_per_iteration": 2.6189026832580566 + }, + { + "auxiliary_loss_clip": 0.01210876, + "auxiliary_loss_mlp": 0.01026154, + "balance_loss_clip": 1.04737639, + "balance_loss_mlp": 1.01834321, + "epoch": 0.6245415739794384, + "flos": 15231978524160.0, + "grad_norm": 1.9334768949370842, + "language_loss": 0.74269152, + "learning_rate": 1.3051538603343075e-06, + "loss": 0.7650618, + "num_input_tokens_seen": 111870340, + "step": 5194, + "time_per_iteration": 2.6424880027770996 + }, + { + "auxiliary_loss_clip": 0.01184181, + "auxiliary_loss_mlp": 0.01030759, + "balance_loss_clip": 1.05226433, + "balance_loss_mlp": 1.02349281, + "epoch": 0.6246618168700776, + "flos": 18879891960960.0, + "grad_norm": 1.9523630052985472, + "language_loss": 0.68003649, + "learning_rate": 1.3044234647301235e-06, + "loss": 0.70218587, + "num_input_tokens_seen": 111888365, + "step": 5195, + "time_per_iteration": 2.6482582092285156 + }, + { + "auxiliary_loss_clip": 0.01178951, + "auxiliary_loss_mlp": 0.01024313, + "balance_loss_clip": 1.04938483, + "balance_loss_mlp": 1.01696646, + "epoch": 0.6247820597607167, + "flos": 14319201087360.0, + "grad_norm": 1.8420048679204608, + "language_loss": 0.7254191, + "learning_rate": 1.303693174644347e-06, + "loss": 0.74745178, + "num_input_tokens_seen": 111905840, + "step": 5196, + "time_per_iteration": 2.6276485919952393 + }, + { + "auxiliary_loss_clip": 0.01186019, + "auxiliary_loss_mlp": 0.01024923, + "balance_loss_clip": 1.046525, + "balance_loss_mlp": 1.01705527, + "epoch": 0.6249023026513557, + "flos": 22637979388800.0, + "grad_norm": 1.7603601711883163, + "language_loss": 0.79999119, + "learning_rate": 1.3029629901877625e-06, + "loss": 0.82210058, + "num_input_tokens_seen": 111925215, + "step": 5197, + "time_per_iteration": 2.7743136882781982 + }, + { + "auxiliary_loss_clip": 0.0119417, + "auxiliary_loss_mlp": 0.01024615, + "balance_loss_clip": 1.05329609, + "balance_loss_mlp": 1.01742685, + "epoch": 0.6250225455419949, + "flos": 20266690204800.0, + "grad_norm": 2.677956067413912, + "language_loss": 0.77425593, + "learning_rate": 1.3022329114711376e-06, + "loss": 0.79644382, + "num_input_tokens_seen": 111943925, + "step": 5198, + "time_per_iteration": 2.6022119522094727 + }, + { + "auxiliary_loss_clip": 0.01188468, + "auxiliary_loss_mlp": 0.01027721, + "balance_loss_clip": 1.04908204, + "balance_loss_mlp": 1.02043104, + "epoch": 0.6251427884326339, + "flos": 23437853400960.0, + "grad_norm": 2.2302355584691487, + "language_loss": 0.69640309, + "learning_rate": 1.3015029386052256e-06, + "loss": 0.71856499, + "num_input_tokens_seen": 111964095, + "step": 5199, + "time_per_iteration": 2.7267584800720215 + }, + { + "auxiliary_loss_clip": 0.01211824, + "auxiliary_loss_mlp": 0.01031489, + "balance_loss_clip": 1.04810405, + "balance_loss_mlp": 1.02426791, + "epoch": 0.625263031323273, + "flos": 31723055464320.0, + "grad_norm": 1.8547933402021912, + "language_loss": 0.72554135, + "learning_rate": 1.3007730717007622e-06, + "loss": 0.74797451, + "num_input_tokens_seen": 111984910, + "step": 5200, + "time_per_iteration": 2.7580761909484863 + }, + { + "auxiliary_loss_clip": 0.0117779, + "auxiliary_loss_mlp": 0.01030065, + "balance_loss_clip": 1.05242145, + "balance_loss_mlp": 1.02238822, + "epoch": 0.6253832742139122, + "flos": 24134341092480.0, + "grad_norm": 1.7187213176409764, + "language_loss": 0.75735426, + "learning_rate": 1.3000433108684676e-06, + "loss": 0.77943277, + "num_input_tokens_seen": 112005410, + "step": 5201, + "time_per_iteration": 2.6829733848571777 + }, + { + "auxiliary_loss_clip": 0.01183581, + "auxiliary_loss_mlp": 0.01024806, + "balance_loss_clip": 1.05144346, + "balance_loss_mlp": 1.0168364, + "epoch": 0.6255035171045512, + "flos": 27668812400640.0, + "grad_norm": 2.2985251975666325, + "language_loss": 0.80076706, + "learning_rate": 1.2993136562190467e-06, + "loss": 0.82285094, + "num_input_tokens_seen": 112024530, + "step": 5202, + "time_per_iteration": 2.7327799797058105 + }, + { + "auxiliary_loss_clip": 0.01194635, + "auxiliary_loss_mlp": 0.01026734, + "balance_loss_clip": 1.04819882, + "balance_loss_mlp": 1.01904488, + "epoch": 0.6256237599951903, + "flos": 20227798753920.0, + "grad_norm": 2.2973649722354557, + "language_loss": 0.70094371, + "learning_rate": 1.2985841078631871e-06, + "loss": 0.72315741, + "num_input_tokens_seen": 112043850, + "step": 5203, + "time_per_iteration": 2.6752755641937256 + }, + { + "auxiliary_loss_clip": 0.01214558, + "auxiliary_loss_mlp": 0.01028932, + "balance_loss_clip": 1.04062879, + "balance_loss_mlp": 1.02058744, + "epoch": 0.6257440028858293, + "flos": 24170574936960.0, + "grad_norm": 1.7556020973165924, + "language_loss": 0.781057, + "learning_rate": 1.2978546659115608e-06, + "loss": 0.80349195, + "num_input_tokens_seen": 112061930, + "step": 5204, + "time_per_iteration": 2.8483471870422363 + }, + { + "auxiliary_loss_clip": 0.01195533, + "auxiliary_loss_mlp": 0.01034097, + "balance_loss_clip": 1.04852903, + "balance_loss_mlp": 1.02656317, + "epoch": 0.6258642457764685, + "flos": 15851940289920.0, + "grad_norm": 2.2293772114026744, + "language_loss": 0.8529973, + "learning_rate": 1.2971253304748228e-06, + "loss": 0.87529361, + "num_input_tokens_seen": 112079645, + "step": 5205, + "time_per_iteration": 2.888129711151123 + }, + { + "auxiliary_loss_clip": 0.01190338, + "auxiliary_loss_mlp": 0.01028062, + "balance_loss_clip": 1.05316925, + "balance_loss_mlp": 1.0198667, + "epoch": 0.6259844886671075, + "flos": 11911354836480.0, + "grad_norm": 1.7848197169600022, + "language_loss": 0.74884683, + "learning_rate": 1.296396101663614e-06, + "loss": 0.77103084, + "num_input_tokens_seen": 112096205, + "step": 5206, + "time_per_iteration": 2.5765061378479004 + }, + { + "auxiliary_loss_clip": 0.01187573, + "auxiliary_loss_mlp": 0.01029572, + "balance_loss_clip": 1.05134749, + "balance_loss_mlp": 1.02152562, + "epoch": 0.6261047315577466, + "flos": 15887958652800.0, + "grad_norm": 2.6245928539674988, + "language_loss": 0.8397162, + "learning_rate": 1.2956669795885565e-06, + "loss": 0.86188763, + "num_input_tokens_seen": 112112835, + "step": 5207, + "time_per_iteration": 2.6679093837738037 + }, + { + "auxiliary_loss_clip": 0.01196041, + "auxiliary_loss_mlp": 0.01031685, + "balance_loss_clip": 1.0484457, + "balance_loss_mlp": 1.0235846, + "epoch": 0.6262249744483858, + "flos": 31248926916480.0, + "grad_norm": 3.16115317425259, + "language_loss": 0.68343937, + "learning_rate": 1.294937964360259e-06, + "loss": 0.70571667, + "num_input_tokens_seen": 112133105, + "step": 5208, + "time_per_iteration": 3.6471636295318604 + }, + { + "auxiliary_loss_clip": 0.01198255, + "auxiliary_loss_mlp": 0.01022848, + "balance_loss_clip": 1.04716349, + "balance_loss_mlp": 1.01441979, + "epoch": 0.6263452173390248, + "flos": 27198598435200.0, + "grad_norm": 9.594449243970365, + "language_loss": 0.71669996, + "learning_rate": 1.2942090560893108e-06, + "loss": 0.73891097, + "num_input_tokens_seen": 112152510, + "step": 5209, + "time_per_iteration": 3.773629665374756 + }, + { + "auxiliary_loss_clip": 0.01174403, + "auxiliary_loss_mlp": 0.01025565, + "balance_loss_clip": 1.0507462, + "balance_loss_mlp": 1.01776886, + "epoch": 0.6264654602296639, + "flos": 37342069683840.0, + "grad_norm": 1.804119281523766, + "language_loss": 0.60095525, + "learning_rate": 1.2934802548862882e-06, + "loss": 0.62295491, + "num_input_tokens_seen": 112175295, + "step": 5210, + "time_per_iteration": 2.7485411167144775 + }, + { + "auxiliary_loss_clip": 0.01189164, + "auxiliary_loss_mlp": 0.01024766, + "balance_loss_clip": 1.0463903, + "balance_loss_mlp": 1.01683831, + "epoch": 0.626585703120303, + "flos": 14756952136320.0, + "grad_norm": 2.714803492539972, + "language_loss": 0.82322013, + "learning_rate": 1.292751560861749e-06, + "loss": 0.84535944, + "num_input_tokens_seen": 112190200, + "step": 5211, + "time_per_iteration": 3.6306817531585693 + }, + { + "auxiliary_loss_clip": 0.01177932, + "auxiliary_loss_mlp": 0.0102782, + "balance_loss_clip": 1.05149674, + "balance_loss_mlp": 1.02016115, + "epoch": 0.6267059460109421, + "flos": 22347318533760.0, + "grad_norm": 1.8845679625306966, + "language_loss": 0.79570138, + "learning_rate": 1.2920229741262354e-06, + "loss": 0.81775892, + "num_input_tokens_seen": 112208205, + "step": 5212, + "time_per_iteration": 2.6618897914886475 + }, + { + "auxiliary_loss_clip": 0.0119196, + "auxiliary_loss_mlp": 0.01026118, + "balance_loss_clip": 1.04713237, + "balance_loss_mlp": 1.01800036, + "epoch": 0.6268261889015811, + "flos": 17748813617280.0, + "grad_norm": 1.9664641541270476, + "language_loss": 0.75089943, + "learning_rate": 1.2912944947902739e-06, + "loss": 0.77308017, + "num_input_tokens_seen": 112224690, + "step": 5213, + "time_per_iteration": 2.637284278869629 + }, + { + "auxiliary_loss_clip": 0.01199272, + "auxiliary_loss_mlp": 0.01023972, + "balance_loss_clip": 1.04864931, + "balance_loss_mlp": 1.01625896, + "epoch": 0.6269464317922203, + "flos": 32846484211200.0, + "grad_norm": 2.423772201567775, + "language_loss": 0.71645081, + "learning_rate": 1.2905661229643742e-06, + "loss": 0.73868322, + "num_input_tokens_seen": 112244450, + "step": 5214, + "time_per_iteration": 3.668069362640381 + }, + { + "auxiliary_loss_clip": 0.01174676, + "auxiliary_loss_mlp": 0.01026638, + "balance_loss_clip": 1.05062973, + "balance_loss_mlp": 1.01900864, + "epoch": 0.6270666746828594, + "flos": 17929192740480.0, + "grad_norm": 2.1319582147196967, + "language_loss": 0.84274971, + "learning_rate": 1.2898378587590299e-06, + "loss": 0.86476284, + "num_input_tokens_seen": 112261050, + "step": 5215, + "time_per_iteration": 2.5558433532714844 + }, + { + "auxiliary_loss_clip": 0.01183443, + "auxiliary_loss_mlp": 0.01026942, + "balance_loss_clip": 1.05195498, + "balance_loss_mlp": 1.01966166, + "epoch": 0.6271869175734984, + "flos": 17457326749440.0, + "grad_norm": 2.319534103326048, + "language_loss": 0.87284625, + "learning_rate": 1.2891097022847173e-06, + "loss": 0.89495003, + "num_input_tokens_seen": 112278395, + "step": 5216, + "time_per_iteration": 2.5950443744659424 + }, + { + "auxiliary_loss_clip": 0.01191773, + "auxiliary_loss_mlp": 0.01027601, + "balance_loss_clip": 1.04829979, + "balance_loss_mlp": 1.01943541, + "epoch": 0.6273071604641376, + "flos": 26868615166080.0, + "grad_norm": 4.780721583652035, + "language_loss": 0.6647585, + "learning_rate": 1.2883816536518978e-06, + "loss": 0.68695223, + "num_input_tokens_seen": 112299535, + "step": 5217, + "time_per_iteration": 2.718841314315796 + }, + { + "auxiliary_loss_clip": 0.01181951, + "auxiliary_loss_mlp": 0.01026681, + "balance_loss_clip": 1.04878342, + "balance_loss_mlp": 1.01905179, + "epoch": 0.6274274033547766, + "flos": 26062384446720.0, + "grad_norm": 1.7913910980914034, + "language_loss": 0.81637353, + "learning_rate": 1.2876537129710155e-06, + "loss": 0.83845985, + "num_input_tokens_seen": 112317265, + "step": 5218, + "time_per_iteration": 2.671882390975952 + }, + { + "auxiliary_loss_clip": 0.01189638, + "auxiliary_loss_mlp": 0.01030143, + "balance_loss_clip": 1.05079174, + "balance_loss_mlp": 1.02167344, + "epoch": 0.6275476462454157, + "flos": 20266259241600.0, + "grad_norm": 3.5565950519794254, + "language_loss": 0.75822079, + "learning_rate": 1.286925880352499e-06, + "loss": 0.78041863, + "num_input_tokens_seen": 112336125, + "step": 5219, + "time_per_iteration": 2.6275365352630615 + }, + { + "auxiliary_loss_clip": 0.01187227, + "auxiliary_loss_mlp": 0.01029503, + "balance_loss_clip": 1.04638481, + "balance_loss_mlp": 1.02174568, + "epoch": 0.6276678891360549, + "flos": 26320402817280.0, + "grad_norm": 1.8083750231953937, + "language_loss": 0.71250379, + "learning_rate": 1.2861981559067592e-06, + "loss": 0.73467106, + "num_input_tokens_seen": 112356730, + "step": 5220, + "time_per_iteration": 2.671520709991455 + }, + { + "auxiliary_loss_clip": 0.01200077, + "auxiliary_loss_mlp": 0.01028048, + "balance_loss_clip": 1.04391587, + "balance_loss_mlp": 1.02031755, + "epoch": 0.6277881320266939, + "flos": 13912512324480.0, + "grad_norm": 1.9776296439537944, + "language_loss": 0.80180728, + "learning_rate": 1.2854705397441917e-06, + "loss": 0.82408857, + "num_input_tokens_seen": 112372270, + "step": 5221, + "time_per_iteration": 2.6732237339019775 + }, + { + "auxiliary_loss_clip": 0.01193315, + "auxiliary_loss_mlp": 0.0102986, + "balance_loss_clip": 1.04338717, + "balance_loss_mlp": 1.02203989, + "epoch": 0.627908374917333, + "flos": 27048922462080.0, + "grad_norm": 2.0206576685329978, + "language_loss": 0.7764473, + "learning_rate": 1.2847430319751747e-06, + "loss": 0.79867899, + "num_input_tokens_seen": 112390365, + "step": 5222, + "time_per_iteration": 2.7684319019317627 + }, + { + "auxiliary_loss_clip": 0.01180856, + "auxiliary_loss_mlp": 0.0102549, + "balance_loss_clip": 1.05174387, + "balance_loss_mlp": 1.01838505, + "epoch": 0.6280286178079721, + "flos": 23769201386880.0, + "grad_norm": 3.267504995476446, + "language_loss": 0.67120755, + "learning_rate": 1.2840156327100712e-06, + "loss": 0.69327104, + "num_input_tokens_seen": 112407490, + "step": 5223, + "time_per_iteration": 2.585726261138916 + }, + { + "auxiliary_loss_clip": 0.0117481, + "auxiliary_loss_mlp": 0.01023279, + "balance_loss_clip": 1.05207133, + "balance_loss_mlp": 1.01576853, + "epoch": 0.6281488606986112, + "flos": 26359150613760.0, + "grad_norm": 1.84927735936292, + "language_loss": 0.72329426, + "learning_rate": 1.2832883420592272e-06, + "loss": 0.74527514, + "num_input_tokens_seen": 112426385, + "step": 5224, + "time_per_iteration": 2.6814112663269043 + }, + { + "auxiliary_loss_clip": 0.01186497, + "auxiliary_loss_mlp": 0.01020041, + "balance_loss_clip": 1.04705739, + "balance_loss_mlp": 1.01241183, + "epoch": 0.6282691035892503, + "flos": 36137194848000.0, + "grad_norm": 2.873557242121126, + "language_loss": 0.64185631, + "learning_rate": 1.282561160132972e-06, + "loss": 0.66392171, + "num_input_tokens_seen": 112446905, + "step": 5225, + "time_per_iteration": 2.838843584060669 + }, + { + "auxiliary_loss_clip": 0.01196668, + "auxiliary_loss_mlp": 0.01025757, + "balance_loss_clip": 1.0458535, + "balance_loss_mlp": 1.01748991, + "epoch": 0.6283893464798894, + "flos": 26537231266560.0, + "grad_norm": 1.5351272938603788, + "language_loss": 0.80726308, + "learning_rate": 1.2818340870416186e-06, + "loss": 0.82948732, + "num_input_tokens_seen": 112468040, + "step": 5226, + "time_per_iteration": 2.6960091590881348 + }, + { + "auxiliary_loss_clip": 0.01206683, + "auxiliary_loss_mlp": 0.01025222, + "balance_loss_clip": 1.04453111, + "balance_loss_mlp": 1.0171994, + "epoch": 0.6285095893705285, + "flos": 22237216369920.0, + "grad_norm": 3.1699929443424195, + "language_loss": 0.75851351, + "learning_rate": 1.2811071228954626e-06, + "loss": 0.78083259, + "num_input_tokens_seen": 112486675, + "step": 5227, + "time_per_iteration": 2.6798179149627686 + }, + { + "auxiliary_loss_clip": 0.01194983, + "auxiliary_loss_mlp": 0.01027021, + "balance_loss_clip": 1.05142117, + "balance_loss_mlp": 1.01916194, + "epoch": 0.6286298322611675, + "flos": 26542259170560.0, + "grad_norm": 1.7346392453074337, + "language_loss": 0.80684215, + "learning_rate": 1.2803802678047846e-06, + "loss": 0.82906216, + "num_input_tokens_seen": 112506825, + "step": 5228, + "time_per_iteration": 2.7128710746765137 + }, + { + "auxiliary_loss_clip": 0.01199905, + "auxiliary_loss_mlp": 0.0102415, + "balance_loss_clip": 1.05092406, + "balance_loss_mlp": 1.01523292, + "epoch": 0.6287500751518067, + "flos": 21795227516160.0, + "grad_norm": 2.3819512473310405, + "language_loss": 0.74326026, + "learning_rate": 1.279653521879848e-06, + "loss": 0.76550078, + "num_input_tokens_seen": 112526890, + "step": 5229, + "time_per_iteration": 2.711638927459717 + }, + { + "auxiliary_loss_clip": 0.01215899, + "auxiliary_loss_mlp": 0.01027877, + "balance_loss_clip": 1.03936076, + "balance_loss_mlp": 1.01987791, + "epoch": 0.6288703180424458, + "flos": 20009605587840.0, + "grad_norm": 2.491449873046776, + "language_loss": 0.83759749, + "learning_rate": 1.2789268852308997e-06, + "loss": 0.8600353, + "num_input_tokens_seen": 112542100, + "step": 5230, + "time_per_iteration": 2.789515733718872 + }, + { + "auxiliary_loss_clip": 0.01174347, + "auxiliary_loss_mlp": 0.01032596, + "balance_loss_clip": 1.04839206, + "balance_loss_mlp": 1.0249728, + "epoch": 0.6289905609330848, + "flos": 22124923476480.0, + "grad_norm": 3.102059553412775, + "language_loss": 0.70540017, + "learning_rate": 1.2782003579681688e-06, + "loss": 0.72746962, + "num_input_tokens_seen": 112561630, + "step": 5231, + "time_per_iteration": 2.6834094524383545 + }, + { + "auxiliary_loss_clip": 0.01175954, + "auxiliary_loss_mlp": 0.01028526, + "balance_loss_clip": 1.05126405, + "balance_loss_mlp": 1.02095056, + "epoch": 0.629110803823724, + "flos": 25518481729920.0, + "grad_norm": 2.239492700438655, + "language_loss": 0.74351835, + "learning_rate": 1.2774739402018701e-06, + "loss": 0.76556313, + "num_input_tokens_seen": 112582465, + "step": 5232, + "time_per_iteration": 2.630434513092041 + }, + { + "auxiliary_loss_clip": 0.01182376, + "auxiliary_loss_mlp": 0.01031859, + "balance_loss_clip": 1.05079699, + "balance_loss_mlp": 1.02368093, + "epoch": 0.629231046714363, + "flos": 20886616056960.0, + "grad_norm": 1.9652417044297072, + "language_loss": 0.73021257, + "learning_rate": 1.2767476320422002e-06, + "loss": 0.75235492, + "num_input_tokens_seen": 112602390, + "step": 5233, + "time_per_iteration": 2.6282200813293457 + }, + { + "auxiliary_loss_clip": 0.01110328, + "auxiliary_loss_mlp": 0.00999397, + "balance_loss_clip": 1.01531541, + "balance_loss_mlp": 0.99830067, + "epoch": 0.6293512896050021, + "flos": 65050027908480.0, + "grad_norm": 0.6796573117933139, + "language_loss": 0.57214415, + "learning_rate": 1.2760214335993392e-06, + "loss": 0.59324133, + "num_input_tokens_seen": 112669035, + "step": 5234, + "time_per_iteration": 3.5691967010498047 + }, + { + "auxiliary_loss_clip": 0.01175631, + "auxiliary_loss_mlp": 0.01024866, + "balance_loss_clip": 1.04811561, + "balance_loss_mlp": 1.01771343, + "epoch": 0.6294715324956413, + "flos": 34677857088000.0, + "grad_norm": 1.941558897594646, + "language_loss": 0.5868625, + "learning_rate": 1.2752953449834514e-06, + "loss": 0.60886741, + "num_input_tokens_seen": 112691485, + "step": 5235, + "time_per_iteration": 4.5291852951049805 + }, + { + "auxiliary_loss_clip": 0.0117529, + "auxiliary_loss_mlp": 0.01024349, + "balance_loss_clip": 1.05172896, + "balance_loss_mlp": 1.01728272, + "epoch": 0.6295917753862803, + "flos": 22784207656320.0, + "grad_norm": 1.6353051584327638, + "language_loss": 0.8015188, + "learning_rate": 1.2745693663046836e-06, + "loss": 0.82351518, + "num_input_tokens_seen": 112710555, + "step": 5236, + "time_per_iteration": 2.8477895259857178 + }, + { + "auxiliary_loss_clip": 0.01178391, + "auxiliary_loss_mlp": 0.01026578, + "balance_loss_clip": 1.04824233, + "balance_loss_mlp": 1.01914549, + "epoch": 0.6297120182769194, + "flos": 20850454039680.0, + "grad_norm": 1.7504889033561648, + "language_loss": 0.80701262, + "learning_rate": 1.2738434976731662e-06, + "loss": 0.82906234, + "num_input_tokens_seen": 112728740, + "step": 5237, + "time_per_iteration": 3.5197744369506836 + }, + { + "auxiliary_loss_clip": 0.01195837, + "auxiliary_loss_mlp": 0.01024633, + "balance_loss_clip": 1.05125952, + "balance_loss_mlp": 1.01747191, + "epoch": 0.6298322611675584, + "flos": 19497662997120.0, + "grad_norm": 1.7083109952854607, + "language_loss": 0.75011986, + "learning_rate": 1.2731177391990125e-06, + "loss": 0.77232456, + "num_input_tokens_seen": 112748665, + "step": 5238, + "time_per_iteration": 2.7034072875976562 + }, + { + "auxiliary_loss_clip": 0.01194337, + "auxiliary_loss_mlp": 0.01025807, + "balance_loss_clip": 1.04759741, + "balance_loss_mlp": 1.01796556, + "epoch": 0.6299525040581976, + "flos": 12604466649600.0, + "grad_norm": 1.9009694331324856, + "language_loss": 0.81891727, + "learning_rate": 1.2723920909923203e-06, + "loss": 0.84111869, + "num_input_tokens_seen": 112764410, + "step": 5239, + "time_per_iteration": 2.6733415126800537 + }, + { + "auxiliary_loss_clip": 0.01068363, + "auxiliary_loss_mlp": 0.01000628, + "balance_loss_clip": 1.01398683, + "balance_loss_mlp": 0.99950153, + "epoch": 0.6300727469488366, + "flos": 57725685636480.0, + "grad_norm": 0.8700967936053101, + "language_loss": 0.6041742, + "learning_rate": 1.2716665531631688e-06, + "loss": 0.62486416, + "num_input_tokens_seen": 112818695, + "step": 5240, + "time_per_iteration": 3.99774432182312 + }, + { + "auxiliary_loss_clip": 0.01188758, + "auxiliary_loss_mlp": 0.01027236, + "balance_loss_clip": 1.04853678, + "balance_loss_mlp": 1.01964855, + "epoch": 0.6301929898394757, + "flos": 22527302607360.0, + "grad_norm": 1.6603753100139327, + "language_loss": 0.77469206, + "learning_rate": 1.270941125821623e-06, + "loss": 0.79685199, + "num_input_tokens_seen": 112839120, + "step": 5241, + "time_per_iteration": 2.7057433128356934 + }, + { + "auxiliary_loss_clip": 0.01180239, + "auxiliary_loss_mlp": 0.01023225, + "balance_loss_clip": 1.0497067, + "balance_loss_mlp": 1.01572073, + "epoch": 0.6303132327301149, + "flos": 28293550675200.0, + "grad_norm": 1.6933089570470707, + "language_loss": 0.75557345, + "learning_rate": 1.2702158090777278e-06, + "loss": 0.7776081, + "num_input_tokens_seen": 112860210, + "step": 5242, + "time_per_iteration": 2.7533934116363525 + }, + { + "auxiliary_loss_clip": 0.01195677, + "auxiliary_loss_mlp": 0.01024216, + "balance_loss_clip": 1.045331, + "balance_loss_mlp": 1.01655686, + "epoch": 0.6304334756207539, + "flos": 25264521596160.0, + "grad_norm": 1.832292071353379, + "language_loss": 0.75058043, + "learning_rate": 1.2694906030415148e-06, + "loss": 0.77277935, + "num_input_tokens_seen": 112877955, + "step": 5243, + "time_per_iteration": 2.7125484943389893 + }, + { + "auxiliary_loss_clip": 0.01199128, + "auxiliary_loss_mlp": 0.01026432, + "balance_loss_clip": 1.0472976, + "balance_loss_mlp": 1.01876652, + "epoch": 0.630553718511393, + "flos": 18033548728320.0, + "grad_norm": 4.812230994766702, + "language_loss": 0.82612967, + "learning_rate": 1.2687655078229958e-06, + "loss": 0.84838533, + "num_input_tokens_seen": 112892285, + "step": 5244, + "time_per_iteration": 2.697072744369507 + }, + { + "auxiliary_loss_clip": 0.01190694, + "auxiliary_loss_mlp": 0.0102832, + "balance_loss_clip": 1.05006123, + "balance_loss_mlp": 1.02070892, + "epoch": 0.6306739614020321, + "flos": 27304103658240.0, + "grad_norm": 2.946990116592498, + "language_loss": 0.69303536, + "learning_rate": 1.2680405235321678e-06, + "loss": 0.71522546, + "num_input_tokens_seen": 112913620, + "step": 5245, + "time_per_iteration": 2.7466259002685547 + }, + { + "auxiliary_loss_clip": 0.0119476, + "auxiliary_loss_mlp": 0.01379099, + "balance_loss_clip": 1.04967093, + "balance_loss_mlp": 1.00035024, + "epoch": 0.6307942042926712, + "flos": 15341434243200.0, + "grad_norm": 2.2635775823616506, + "language_loss": 0.78926164, + "learning_rate": 1.267315650279011e-06, + "loss": 0.81500024, + "num_input_tokens_seen": 112932090, + "step": 5246, + "time_per_iteration": 2.6566109657287598 + }, + { + "auxiliary_loss_clip": 0.01198841, + "auxiliary_loss_mlp": 0.01032029, + "balance_loss_clip": 1.05173302, + "balance_loss_mlp": 1.02338076, + "epoch": 0.6309144471833102, + "flos": 19606400444160.0, + "grad_norm": 1.7773836245317374, + "language_loss": 0.73722935, + "learning_rate": 1.2665908881734874e-06, + "loss": 0.75953805, + "num_input_tokens_seen": 112950925, + "step": 5247, + "time_per_iteration": 2.68058705329895 + }, + { + "auxiliary_loss_clip": 0.0118556, + "auxiliary_loss_mlp": 0.01025437, + "balance_loss_clip": 1.05120444, + "balance_loss_mlp": 1.01792336, + "epoch": 0.6310346900739494, + "flos": 17493345112320.0, + "grad_norm": 2.4404995291581, + "language_loss": 0.84848762, + "learning_rate": 1.2658662373255432e-06, + "loss": 0.87059754, + "num_input_tokens_seen": 112969315, + "step": 5248, + "time_per_iteration": 2.619981288909912 + }, + { + "auxiliary_loss_clip": 0.01092679, + "auxiliary_loss_mlp": 0.01003066, + "balance_loss_clip": 1.01297379, + "balance_loss_mlp": 1.00189805, + "epoch": 0.6311549329645885, + "flos": 55070164131840.0, + "grad_norm": 0.7108715051783111, + "language_loss": 0.52261031, + "learning_rate": 1.2651416978451063e-06, + "loss": 0.54356778, + "num_input_tokens_seen": 113034700, + "step": 5249, + "time_per_iteration": 3.3262863159179688 + }, + { + "auxiliary_loss_clip": 0.01178351, + "auxiliary_loss_mlp": 0.01025881, + "balance_loss_clip": 1.05306745, + "balance_loss_mlp": 1.01767361, + "epoch": 0.6312751758552275, + "flos": 41902545075840.0, + "grad_norm": 1.7439119545012343, + "language_loss": 0.64994943, + "learning_rate": 1.2644172698420903e-06, + "loss": 0.67199177, + "num_input_tokens_seen": 113056805, + "step": 5250, + "time_per_iteration": 2.757988929748535 + }, + { + "auxiliary_loss_clip": 0.01202699, + "auxiliary_loss_mlp": 0.01020641, + "balance_loss_clip": 1.04718578, + "balance_loss_mlp": 1.01328266, + "epoch": 0.6313954187458667, + "flos": 19646800266240.0, + "grad_norm": 1.896831373979289, + "language_loss": 0.84618258, + "learning_rate": 1.2636929534263892e-06, + "loss": 0.86841595, + "num_input_tokens_seen": 113075790, + "step": 5251, + "time_per_iteration": 2.684751033782959 + }, + { + "auxiliary_loss_clip": 0.01201182, + "auxiliary_loss_mlp": 0.01022045, + "balance_loss_clip": 1.04337549, + "balance_loss_mlp": 1.01435566, + "epoch": 0.6315156616365057, + "flos": 22894273906560.0, + "grad_norm": 2.2826256815939128, + "language_loss": 0.77557421, + "learning_rate": 1.2629687487078821e-06, + "loss": 0.79780644, + "num_input_tokens_seen": 113094600, + "step": 5252, + "time_per_iteration": 2.643815279006958 + }, + { + "auxiliary_loss_clip": 0.01185889, + "auxiliary_loss_mlp": 0.01026406, + "balance_loss_clip": 1.047876, + "balance_loss_mlp": 1.01853216, + "epoch": 0.6316359045271448, + "flos": 23726251699200.0, + "grad_norm": 1.9697329358185358, + "language_loss": 0.76259315, + "learning_rate": 1.2622446557964293e-06, + "loss": 0.78471613, + "num_input_tokens_seen": 113112605, + "step": 5253, + "time_per_iteration": 2.6635849475860596 + }, + { + "auxiliary_loss_clip": 0.01187102, + "auxiliary_loss_mlp": 0.01026271, + "balance_loss_clip": 1.04218197, + "balance_loss_mlp": 1.01876712, + "epoch": 0.631756147417784, + "flos": 33108417164160.0, + "grad_norm": 2.026198022500904, + "language_loss": 0.7132163, + "learning_rate": 1.261520674801876e-06, + "loss": 0.73535007, + "num_input_tokens_seen": 113133200, + "step": 5254, + "time_per_iteration": 2.76004695892334 + }, + { + "auxiliary_loss_clip": 0.01191297, + "auxiliary_loss_mlp": 0.01030961, + "balance_loss_clip": 1.04982948, + "balance_loss_mlp": 1.02241671, + "epoch": 0.631876390308423, + "flos": 31248424126080.0, + "grad_norm": 1.915763955561821, + "language_loss": 0.72041833, + "learning_rate": 1.2607968058340488e-06, + "loss": 0.74264097, + "num_input_tokens_seen": 113152895, + "step": 5255, + "time_per_iteration": 2.736048698425293 + }, + { + "auxiliary_loss_clip": 0.0118949, + "auxiliary_loss_mlp": 0.010259, + "balance_loss_clip": 1.04778504, + "balance_loss_mlp": 1.0183717, + "epoch": 0.6319966331990621, + "flos": 24681152810880.0, + "grad_norm": 1.8070813094592593, + "language_loss": 0.72717857, + "learning_rate": 1.2600730490027583e-06, + "loss": 0.74933243, + "num_input_tokens_seen": 113173135, + "step": 5256, + "time_per_iteration": 2.7469472885131836 + }, + { + "auxiliary_loss_clip": 0.01197168, + "auxiliary_loss_mlp": 0.01030182, + "balance_loss_clip": 1.04644322, + "balance_loss_mlp": 1.02264833, + "epoch": 0.6321168760897012, + "flos": 17491764913920.0, + "grad_norm": 1.6996683750363715, + "language_loss": 0.8065241, + "learning_rate": 1.2593494044177984e-06, + "loss": 0.82879758, + "num_input_tokens_seen": 113191440, + "step": 5257, + "time_per_iteration": 2.7246220111846924 + }, + { + "auxiliary_loss_clip": 0.01177728, + "auxiliary_loss_mlp": 0.01025679, + "balance_loss_clip": 1.04938459, + "balance_loss_mlp": 1.01762009, + "epoch": 0.6322371189803403, + "flos": 18295373940480.0, + "grad_norm": 2.7791596335526085, + "language_loss": 0.80648124, + "learning_rate": 1.2586258721889448e-06, + "loss": 0.82851529, + "num_input_tokens_seen": 113208790, + "step": 5258, + "time_per_iteration": 2.6089677810668945 + }, + { + "auxiliary_loss_clip": 0.01200391, + "auxiliary_loss_mlp": 0.01029957, + "balance_loss_clip": 1.04395008, + "balance_loss_mlp": 1.02179098, + "epoch": 0.6323573618709794, + "flos": 20157270399360.0, + "grad_norm": 4.7765935083941855, + "language_loss": 0.81635988, + "learning_rate": 1.2579024524259573e-06, + "loss": 0.83866334, + "num_input_tokens_seen": 113225050, + "step": 5259, + "time_per_iteration": 2.742880344390869 + }, + { + "auxiliary_loss_clip": 0.01183726, + "auxiliary_loss_mlp": 0.01027327, + "balance_loss_clip": 1.04320073, + "balance_loss_mlp": 1.01941776, + "epoch": 0.6324776047616185, + "flos": 20042391726720.0, + "grad_norm": 1.9678340105510368, + "language_loss": 0.91648245, + "learning_rate": 1.2571791452385768e-06, + "loss": 0.93859303, + "num_input_tokens_seen": 113242315, + "step": 5260, + "time_per_iteration": 3.588127613067627 + }, + { + "auxiliary_loss_clip": 0.01193587, + "auxiliary_loss_mlp": 0.01022039, + "balance_loss_clip": 1.04873824, + "balance_loss_mlp": 1.01512444, + "epoch": 0.6325978476522576, + "flos": 30848235724800.0, + "grad_norm": 1.5750775662120349, + "language_loss": 0.77138144, + "learning_rate": 1.2564559507365301e-06, + "loss": 0.79353774, + "num_input_tokens_seen": 113264720, + "step": 5261, + "time_per_iteration": 3.6525862216949463 + }, + { + "auxiliary_loss_clip": 0.01194025, + "auxiliary_loss_mlp": 0.01025183, + "balance_loss_clip": 1.04898834, + "balance_loss_mlp": 1.01732075, + "epoch": 0.6327180905428966, + "flos": 24535104111360.0, + "grad_norm": 2.163877498082616, + "language_loss": 0.78876448, + "learning_rate": 1.2557328690295244e-06, + "loss": 0.8109566, + "num_input_tokens_seen": 113282910, + "step": 5262, + "time_per_iteration": 2.670412063598633 + }, + { + "auxiliary_loss_clip": 0.01205386, + "auxiliary_loss_mlp": 0.01023567, + "balance_loss_clip": 1.04698873, + "balance_loss_mlp": 1.01639962, + "epoch": 0.6328383334335358, + "flos": 21575274583680.0, + "grad_norm": 1.7186080467249263, + "language_loss": 0.764741, + "learning_rate": 1.255009900227251e-06, + "loss": 0.78703058, + "num_input_tokens_seen": 113301935, + "step": 5263, + "time_per_iteration": 3.6137402057647705 + }, + { + "auxiliary_loss_clip": 0.01173406, + "auxiliary_loss_mlp": 0.01023182, + "balance_loss_clip": 1.05158925, + "balance_loss_mlp": 1.01545179, + "epoch": 0.6329585763241748, + "flos": 22929861306240.0, + "grad_norm": 1.9597862880127688, + "language_loss": 0.79477894, + "learning_rate": 1.254287044439383e-06, + "loss": 0.81674486, + "num_input_tokens_seen": 113321540, + "step": 5264, + "time_per_iteration": 2.5936357975006104 + }, + { + "auxiliary_loss_clip": 0.01067587, + "auxiliary_loss_mlp": 0.01003762, + "balance_loss_clip": 1.01347232, + "balance_loss_mlp": 1.00254059, + "epoch": 0.6330788192148139, + "flos": 70936897847040.0, + "grad_norm": 0.7791702498377467, + "language_loss": 0.54518688, + "learning_rate": 1.2535643017755776e-06, + "loss": 0.56590033, + "num_input_tokens_seen": 113383730, + "step": 5265, + "time_per_iteration": 3.277998208999634 + }, + { + "auxiliary_loss_clip": 0.01199229, + "auxiliary_loss_mlp": 0.0102762, + "balance_loss_clip": 1.04451168, + "balance_loss_mlp": 1.02022862, + "epoch": 0.6331990621054531, + "flos": 21244501215360.0, + "grad_norm": 9.019002788504187, + "language_loss": 0.71793878, + "learning_rate": 1.2528416723454737e-06, + "loss": 0.74020725, + "num_input_tokens_seen": 113400400, + "step": 5266, + "time_per_iteration": 3.6407382488250732 + }, + { + "auxiliary_loss_clip": 0.01174039, + "auxiliary_loss_mlp": 0.01022766, + "balance_loss_clip": 1.05163133, + "balance_loss_mlp": 1.01547289, + "epoch": 0.6333193049960921, + "flos": 34459412526720.0, + "grad_norm": 1.5052942845521424, + "language_loss": 0.71318066, + "learning_rate": 1.2521191562586945e-06, + "loss": 0.73514867, + "num_input_tokens_seen": 113424050, + "step": 5267, + "time_per_iteration": 2.765407085418701 + }, + { + "auxiliary_loss_clip": 0.01172808, + "auxiliary_loss_mlp": 0.01379092, + "balance_loss_clip": 1.04959488, + "balance_loss_mlp": 1.00027418, + "epoch": 0.6334395478867312, + "flos": 18329883932160.0, + "grad_norm": 2.210920622494314, + "language_loss": 0.76636273, + "learning_rate": 1.2513967536248445e-06, + "loss": 0.79188174, + "num_input_tokens_seen": 113440370, + "step": 5268, + "time_per_iteration": 2.730957269668579 + }, + { + "auxiliary_loss_clip": 0.01178789, + "auxiliary_loss_mlp": 0.01024553, + "balance_loss_clip": 1.04939616, + "balance_loss_mlp": 1.01681948, + "epoch": 0.6335597907773702, + "flos": 23623152687360.0, + "grad_norm": 1.683531439775821, + "language_loss": 0.81014198, + "learning_rate": 1.2506744645535117e-06, + "loss": 0.83217537, + "num_input_tokens_seen": 113460800, + "step": 5269, + "time_per_iteration": 2.6948628425598145 + }, + { + "auxiliary_loss_clip": 0.01181155, + "auxiliary_loss_mlp": 0.0102791, + "balance_loss_clip": 1.0417949, + "balance_loss_mlp": 1.02063835, + "epoch": 0.6336800336680094, + "flos": 22710913954560.0, + "grad_norm": 2.133861377302028, + "language_loss": 0.60302025, + "learning_rate": 1.249952289154267e-06, + "loss": 0.62511092, + "num_input_tokens_seen": 113480840, + "step": 5270, + "time_per_iteration": 2.702754497528076 + }, + { + "auxiliary_loss_clip": 0.01202718, + "auxiliary_loss_mlp": 0.0102196, + "balance_loss_clip": 1.04097152, + "balance_loss_mlp": 1.01494193, + "epoch": 0.6338002765586485, + "flos": 23622757637760.0, + "grad_norm": 1.9330067188788154, + "language_loss": 0.76580369, + "learning_rate": 1.2492302275366635e-06, + "loss": 0.78805047, + "num_input_tokens_seen": 113500515, + "step": 5271, + "time_per_iteration": 2.786928176879883 + }, + { + "auxiliary_loss_clip": 0.01173549, + "auxiliary_loss_mlp": 0.0102498, + "balance_loss_clip": 1.04648364, + "balance_loss_mlp": 1.01640844, + "epoch": 0.6339205194492875, + "flos": 26505450708480.0, + "grad_norm": 5.327338705332072, + "language_loss": 0.65472931, + "learning_rate": 1.2485082798102377e-06, + "loss": 0.67671454, + "num_input_tokens_seen": 113520930, + "step": 5272, + "time_per_iteration": 2.7182321548461914 + }, + { + "auxiliary_loss_clip": 0.01205198, + "auxiliary_loss_mlp": 0.0102591, + "balance_loss_clip": 1.04622817, + "balance_loss_mlp": 1.01811421, + "epoch": 0.6340407623399267, + "flos": 18544306170240.0, + "grad_norm": 2.5449322194665838, + "language_loss": 0.6844402, + "learning_rate": 1.2477864460845084e-06, + "loss": 0.70675129, + "num_input_tokens_seen": 113537330, + "step": 5273, + "time_per_iteration": 2.7378101348876953 + }, + { + "auxiliary_loss_clip": 0.01192099, + "auxiliary_loss_mlp": 0.01027966, + "balance_loss_clip": 1.04770088, + "balance_loss_mlp": 1.01976991, + "epoch": 0.6341610052305657, + "flos": 17712579772800.0, + "grad_norm": 2.387730921757119, + "language_loss": 0.73475933, + "learning_rate": 1.2470647264689776e-06, + "loss": 0.75695992, + "num_input_tokens_seen": 113555810, + "step": 5274, + "time_per_iteration": 2.642979145050049 + }, + { + "auxiliary_loss_clip": 0.01220664, + "auxiliary_loss_mlp": 0.01024378, + "balance_loss_clip": 1.03952312, + "balance_loss_mlp": 1.01677561, + "epoch": 0.6342812481212048, + "flos": 23587026583680.0, + "grad_norm": 1.9986070677480094, + "language_loss": 0.71120882, + "learning_rate": 1.2463431210731282e-06, + "loss": 0.73365921, + "num_input_tokens_seen": 113575395, + "step": 5275, + "time_per_iteration": 2.7756834030151367 + }, + { + "auxiliary_loss_clip": 0.01217066, + "auxiliary_loss_mlp": 0.01029298, + "balance_loss_clip": 1.04295254, + "balance_loss_mlp": 1.02153754, + "epoch": 0.634401491011844, + "flos": 17821927751040.0, + "grad_norm": 2.4061428186142413, + "language_loss": 0.75904155, + "learning_rate": 1.2456216300064289e-06, + "loss": 0.78150523, + "num_input_tokens_seen": 113592945, + "step": 5276, + "time_per_iteration": 2.6909091472625732 + }, + { + "auxiliary_loss_clip": 0.01180963, + "auxiliary_loss_mlp": 0.01029077, + "balance_loss_clip": 1.044191, + "balance_loss_mlp": 1.02193964, + "epoch": 0.634521733902483, + "flos": 21358158825600.0, + "grad_norm": 2.2110802715808613, + "language_loss": 0.78229147, + "learning_rate": 1.244900253378328e-06, + "loss": 0.80439186, + "num_input_tokens_seen": 113613000, + "step": 5277, + "time_per_iteration": 2.716265916824341 + }, + { + "auxiliary_loss_clip": 0.01240382, + "auxiliary_loss_mlp": 0.0102616, + "balance_loss_clip": 1.04273844, + "balance_loss_mlp": 1.01895928, + "epoch": 0.6346419767931221, + "flos": 16545052103040.0, + "grad_norm": 5.407665675156493, + "language_loss": 0.69419897, + "learning_rate": 1.2441789912982583e-06, + "loss": 0.71686441, + "num_input_tokens_seen": 113630085, + "step": 5278, + "time_per_iteration": 3.0332131385803223 + }, + { + "auxiliary_loss_clip": 0.01187311, + "auxiliary_loss_mlp": 0.01029949, + "balance_loss_clip": 1.05040455, + "balance_loss_mlp": 1.02164006, + "epoch": 0.6347622196837612, + "flos": 24350989973760.0, + "grad_norm": 1.8989400394387241, + "language_loss": 0.65220857, + "learning_rate": 1.2434578438756346e-06, + "loss": 0.67438114, + "num_input_tokens_seen": 113650515, + "step": 5279, + "time_per_iteration": 2.9554216861724854 + }, + { + "auxiliary_loss_clip": 0.01185027, + "auxiliary_loss_mlp": 0.01022863, + "balance_loss_clip": 1.04785383, + "balance_loss_mlp": 1.01578224, + "epoch": 0.6348824625744003, + "flos": 64523178195840.0, + "grad_norm": 1.862380743477723, + "language_loss": 0.78120762, + "learning_rate": 1.242736811219855e-06, + "loss": 0.80328655, + "num_input_tokens_seen": 113676475, + "step": 5280, + "time_per_iteration": 3.0194127559661865 + }, + { + "auxiliary_loss_clip": 0.0117575, + "auxiliary_loss_mlp": 0.01028011, + "balance_loss_clip": 1.04779196, + "balance_loss_mlp": 1.02010119, + "epoch": 0.6350027054650393, + "flos": 28622133313920.0, + "grad_norm": 2.1518039196765373, + "language_loss": 0.81623095, + "learning_rate": 1.2420158934402988e-06, + "loss": 0.83826858, + "num_input_tokens_seen": 113697090, + "step": 5281, + "time_per_iteration": 2.7493693828582764 + }, + { + "auxiliary_loss_clip": 0.01181874, + "auxiliary_loss_mlp": 0.01027397, + "balance_loss_clip": 1.04150271, + "balance_loss_mlp": 1.01917791, + "epoch": 0.6351229483556785, + "flos": 23002544476800.0, + "grad_norm": 2.1619106405056647, + "language_loss": 0.8467865, + "learning_rate": 1.2412950906463286e-06, + "loss": 0.8688792, + "num_input_tokens_seen": 113714395, + "step": 5282, + "time_per_iteration": 2.6764707565307617 + }, + { + "auxiliary_loss_clip": 0.01208027, + "auxiliary_loss_mlp": 0.01022977, + "balance_loss_clip": 1.04534411, + "balance_loss_mlp": 1.01589322, + "epoch": 0.6352431912463176, + "flos": 21939300967680.0, + "grad_norm": 1.8917392767267216, + "language_loss": 0.90051627, + "learning_rate": 1.2405744029472902e-06, + "loss": 0.92282629, + "num_input_tokens_seen": 113733880, + "step": 5283, + "time_per_iteration": 2.7372095584869385 + }, + { + "auxiliary_loss_clip": 0.01187286, + "auxiliary_loss_mlp": 0.01025565, + "balance_loss_clip": 1.04582, + "balance_loss_mlp": 1.01811385, + "epoch": 0.6353634341369566, + "flos": 13735257684480.0, + "grad_norm": 1.9466309562900124, + "language_loss": 0.76432031, + "learning_rate": 1.2398538304525108e-06, + "loss": 0.78644878, + "num_input_tokens_seen": 113752505, + "step": 5284, + "time_per_iteration": 2.693690538406372 + }, + { + "auxiliary_loss_clip": 0.01195419, + "auxiliary_loss_mlp": 0.01029779, + "balance_loss_clip": 1.04622865, + "balance_loss_mlp": 1.02185142, + "epoch": 0.6354836770275958, + "flos": 19316170552320.0, + "grad_norm": 4.7562670583759425, + "language_loss": 0.75513452, + "learning_rate": 1.2391333732713016e-06, + "loss": 0.77738643, + "num_input_tokens_seen": 113770310, + "step": 5285, + "time_per_iteration": 2.6976752281188965 + }, + { + "auxiliary_loss_clip": 0.0119588, + "auxiliary_loss_mlp": 0.01028791, + "balance_loss_clip": 1.04310441, + "balance_loss_mlp": 1.02089357, + "epoch": 0.6356039199182348, + "flos": 21613375935360.0, + "grad_norm": 2.7071846160504465, + "language_loss": 0.78447342, + "learning_rate": 1.2384130315129543e-06, + "loss": 0.80672014, + "num_input_tokens_seen": 113788635, + "step": 5286, + "time_per_iteration": 3.829841375350952 + }, + { + "auxiliary_loss_clip": 0.01250316, + "auxiliary_loss_mlp": 0.01025683, + "balance_loss_clip": 1.03716516, + "balance_loss_mlp": 1.01802421, + "epoch": 0.6357241628088739, + "flos": 18111978074880.0, + "grad_norm": 3.0743002689566308, + "language_loss": 0.73759902, + "learning_rate": 1.2376928052867447e-06, + "loss": 0.76035905, + "num_input_tokens_seen": 113807755, + "step": 5287, + "time_per_iteration": 3.2647528648376465 + }, + { + "auxiliary_loss_clip": 0.01192722, + "auxiliary_loss_mlp": 0.01025932, + "balance_loss_clip": 1.04990292, + "balance_loss_mlp": 1.0184989, + "epoch": 0.6358444056995131, + "flos": 24935256599040.0, + "grad_norm": 1.822362586545449, + "language_loss": 0.77233934, + "learning_rate": 1.2369726947019299e-06, + "loss": 0.79452586, + "num_input_tokens_seen": 113828230, + "step": 5288, + "time_per_iteration": 4.316404581069946 + }, + { + "auxiliary_loss_clip": 0.01182782, + "auxiliary_loss_mlp": 0.01027268, + "balance_loss_clip": 1.04837894, + "balance_loss_mlp": 1.02002037, + "epoch": 0.6359646485901521, + "flos": 23293348986240.0, + "grad_norm": 2.2256675316141634, + "language_loss": 0.6754663, + "learning_rate": 1.2362526998677511e-06, + "loss": 0.69756681, + "num_input_tokens_seen": 113844595, + "step": 5289, + "time_per_iteration": 3.5212745666503906 + }, + { + "auxiliary_loss_clip": 0.01194451, + "auxiliary_loss_mlp": 0.01023462, + "balance_loss_clip": 1.04669356, + "balance_loss_mlp": 1.01636934, + "epoch": 0.6360848914807912, + "flos": 20887442069760.0, + "grad_norm": 1.7604840940071287, + "language_loss": 0.84304678, + "learning_rate": 1.2355328208934301e-06, + "loss": 0.86522591, + "num_input_tokens_seen": 113863470, + "step": 5290, + "time_per_iteration": 2.7610628604888916 + }, + { + "auxiliary_loss_clip": 0.0118199, + "auxiliary_loss_mlp": 0.01378784, + "balance_loss_clip": 1.04671049, + "balance_loss_mlp": 1.00027323, + "epoch": 0.6362051343714303, + "flos": 18479775386880.0, + "grad_norm": 2.833532704744228, + "language_loss": 0.72279185, + "learning_rate": 1.2348130578881728e-06, + "loss": 0.74839956, + "num_input_tokens_seen": 113881690, + "step": 5291, + "time_per_iteration": 2.6070947647094727 + }, + { + "auxiliary_loss_clip": 0.01177689, + "auxiliary_loss_mlp": 0.01025369, + "balance_loss_clip": 1.05157697, + "balance_loss_mlp": 1.01727486, + "epoch": 0.6363253772620694, + "flos": 24389594115840.0, + "grad_norm": 4.3983563854846475, + "language_loss": 0.76348996, + "learning_rate": 1.2340934109611664e-06, + "loss": 0.78552043, + "num_input_tokens_seen": 113902450, + "step": 5292, + "time_per_iteration": 3.5424880981445312 + }, + { + "auxiliary_loss_clip": 0.01196382, + "auxiliary_loss_mlp": 0.01027617, + "balance_loss_clip": 1.0483216, + "balance_loss_mlp": 1.01914144, + "epoch": 0.6364456201527084, + "flos": 25958243940480.0, + "grad_norm": 2.5771734190291635, + "language_loss": 0.68653524, + "learning_rate": 1.2333738802215798e-06, + "loss": 0.70877516, + "num_input_tokens_seen": 113922670, + "step": 5293, + "time_per_iteration": 2.733104705810547 + }, + { + "auxiliary_loss_clip": 0.01199155, + "auxiliary_loss_mlp": 0.01025012, + "balance_loss_clip": 1.04120207, + "balance_loss_mlp": 1.01749289, + "epoch": 0.6365658630433476, + "flos": 20740711011840.0, + "grad_norm": 1.8853573074570666, + "language_loss": 0.81277531, + "learning_rate": 1.2326544657785668e-06, + "loss": 0.83501697, + "num_input_tokens_seen": 113942360, + "step": 5294, + "time_per_iteration": 2.7577762603759766 + }, + { + "auxiliary_loss_clip": 0.01188998, + "auxiliary_loss_mlp": 0.01028931, + "balance_loss_clip": 1.04493654, + "balance_loss_mlp": 1.02056289, + "epoch": 0.6366861059339867, + "flos": 21434146047360.0, + "grad_norm": 2.411912599672922, + "language_loss": 0.74551666, + "learning_rate": 1.2319351677412608e-06, + "loss": 0.7676959, + "num_input_tokens_seen": 113959405, + "step": 5295, + "time_per_iteration": 2.674002170562744 + }, + { + "auxiliary_loss_clip": 0.01208937, + "auxiliary_loss_mlp": 0.01027351, + "balance_loss_clip": 1.04677725, + "balance_loss_mlp": 1.01946509, + "epoch": 0.6368063488246257, + "flos": 22267093507200.0, + "grad_norm": 1.8115194676976345, + "language_loss": 0.74264807, + "learning_rate": 1.2312159862187796e-06, + "loss": 0.76501095, + "num_input_tokens_seen": 113977815, + "step": 5296, + "time_per_iteration": 2.7461979389190674 + }, + { + "auxiliary_loss_clip": 0.01180511, + "auxiliary_loss_mlp": 0.01022651, + "balance_loss_clip": 1.05445397, + "balance_loss_mlp": 1.01476586, + "epoch": 0.6369265917152649, + "flos": 22420719976320.0, + "grad_norm": 1.5999601838208424, + "language_loss": 0.75721276, + "learning_rate": 1.2304969213202217e-06, + "loss": 0.77924442, + "num_input_tokens_seen": 113999075, + "step": 5297, + "time_per_iteration": 2.6086137294769287 + }, + { + "auxiliary_loss_clip": 0.01184211, + "auxiliary_loss_mlp": 0.0102447, + "balance_loss_clip": 1.04431009, + "balance_loss_mlp": 1.01704311, + "epoch": 0.6370468346059039, + "flos": 24718176754560.0, + "grad_norm": 3.187234603572375, + "language_loss": 0.79452491, + "learning_rate": 1.2297779731546692e-06, + "loss": 0.81661171, + "num_input_tokens_seen": 114018170, + "step": 5298, + "time_per_iteration": 2.675943613052368 + }, + { + "auxiliary_loss_clip": 0.01190024, + "auxiliary_loss_mlp": 0.01028157, + "balance_loss_clip": 1.04900146, + "balance_loss_mlp": 1.02072418, + "epoch": 0.637167077496543, + "flos": 25296589463040.0, + "grad_norm": 2.108797043272583, + "language_loss": 0.77853853, + "learning_rate": 1.2290591418311853e-06, + "loss": 0.80072033, + "num_input_tokens_seen": 114035565, + "step": 5299, + "time_per_iteration": 2.688466787338257 + }, + { + "auxiliary_loss_clip": 0.01183908, + "auxiliary_loss_mlp": 0.01033029, + "balance_loss_clip": 1.05135763, + "balance_loss_mlp": 1.02537608, + "epoch": 0.637287320387182, + "flos": 27671110871040.0, + "grad_norm": 1.5668518954260162, + "language_loss": 0.7216121, + "learning_rate": 1.2283404274588172e-06, + "loss": 0.74378145, + "num_input_tokens_seen": 114054510, + "step": 5300, + "time_per_iteration": 2.6834282875061035 + }, + { + "auxiliary_loss_clip": 0.01132664, + "auxiliary_loss_mlp": 0.01002085, + "balance_loss_clip": 1.01656449, + "balance_loss_mlp": 1.00109565, + "epoch": 0.6374075632778212, + "flos": 63173406873600.0, + "grad_norm": 0.7387893525840555, + "language_loss": 0.52706158, + "learning_rate": 1.227621830146592e-06, + "loss": 0.54840904, + "num_input_tokens_seen": 114109875, + "step": 5301, + "time_per_iteration": 3.376912832260132 + }, + { + "auxiliary_loss_clip": 0.01202617, + "auxiliary_loss_mlp": 0.01024858, + "balance_loss_clip": 1.04692209, + "balance_loss_mlp": 1.01687717, + "epoch": 0.6375278061684603, + "flos": 25558127366400.0, + "grad_norm": 1.7898388186142675, + "language_loss": 0.79084444, + "learning_rate": 1.2269033500035217e-06, + "loss": 0.81311917, + "num_input_tokens_seen": 114130010, + "step": 5302, + "time_per_iteration": 3.3973031044006348 + }, + { + "auxiliary_loss_clip": 0.0119948, + "auxiliary_loss_mlp": 0.01032745, + "balance_loss_clip": 1.04567373, + "balance_loss_mlp": 1.0250566, + "epoch": 0.6376480490590993, + "flos": 25666362023040.0, + "grad_norm": 1.9475795923013468, + "language_loss": 0.73446941, + "learning_rate": 1.2261849871385988e-06, + "loss": 0.75679165, + "num_input_tokens_seen": 114151115, + "step": 5303, + "time_per_iteration": 2.749156951904297 + }, + { + "auxiliary_loss_clip": 0.01177322, + "auxiliary_loss_mlp": 0.01028252, + "balance_loss_clip": 1.05033088, + "balance_loss_mlp": 1.020432, + "epoch": 0.6377682919497385, + "flos": 31537684350720.0, + "grad_norm": 2.091273706361628, + "language_loss": 0.62565553, + "learning_rate": 1.2254667416607972e-06, + "loss": 0.64771128, + "num_input_tokens_seen": 114172715, + "step": 5304, + "time_per_iteration": 2.672736883163452 + }, + { + "auxiliary_loss_clip": 0.01181269, + "auxiliary_loss_mlp": 0.01025162, + "balance_loss_clip": 1.04881454, + "balance_loss_mlp": 1.01732385, + "epoch": 0.6378885348403776, + "flos": 23039209284480.0, + "grad_norm": 1.8713841227004695, + "language_loss": 0.83069366, + "learning_rate": 1.2247486136790756e-06, + "loss": 0.85275793, + "num_input_tokens_seen": 114192195, + "step": 5305, + "time_per_iteration": 2.7589657306671143 + }, + { + "auxiliary_loss_clip": 0.01186175, + "auxiliary_loss_mlp": 0.01027736, + "balance_loss_clip": 1.04971492, + "balance_loss_mlp": 1.02034461, + "epoch": 0.6380087777310166, + "flos": 18697070712960.0, + "grad_norm": 2.2378087226870584, + "language_loss": 0.81020838, + "learning_rate": 1.2240306033023726e-06, + "loss": 0.83234751, + "num_input_tokens_seen": 114210020, + "step": 5306, + "time_per_iteration": 2.7308366298675537 + }, + { + "auxiliary_loss_clip": 0.01198827, + "auxiliary_loss_mlp": 0.01026122, + "balance_loss_clip": 1.04226053, + "balance_loss_mlp": 1.01885283, + "epoch": 0.6381290206216558, + "flos": 23331558078720.0, + "grad_norm": 2.20199244944394, + "language_loss": 0.72088039, + "learning_rate": 1.223312710639611e-06, + "loss": 0.74312985, + "num_input_tokens_seen": 114228740, + "step": 5307, + "time_per_iteration": 2.7348899841308594 + }, + { + "auxiliary_loss_clip": 0.0118989, + "auxiliary_loss_mlp": 0.0102635, + "balance_loss_clip": 1.04831696, + "balance_loss_mlp": 1.0184648, + "epoch": 0.6382492635122948, + "flos": 18880466578560.0, + "grad_norm": 2.0810818627553918, + "language_loss": 0.86773109, + "learning_rate": 1.2225949357996928e-06, + "loss": 0.88989353, + "num_input_tokens_seen": 114246865, + "step": 5308, + "time_per_iteration": 2.6815268993377686 + }, + { + "auxiliary_loss_clip": 0.01182114, + "auxiliary_loss_mlp": 0.01028997, + "balance_loss_clip": 1.05066419, + "balance_loss_mlp": 1.02062237, + "epoch": 0.6383695064029339, + "flos": 27819134818560.0, + "grad_norm": 1.5586616396752928, + "language_loss": 0.80370688, + "learning_rate": 1.221877278891505e-06, + "loss": 0.82581806, + "num_input_tokens_seen": 114266120, + "step": 5309, + "time_per_iteration": 2.7091662883758545 + }, + { + "auxiliary_loss_clip": 0.01190749, + "auxiliary_loss_mlp": 0.01025518, + "balance_loss_clip": 1.05077481, + "balance_loss_mlp": 1.01748288, + "epoch": 0.638489749293573, + "flos": 26395635853440.0, + "grad_norm": 3.343262368146808, + "language_loss": 0.71505255, + "learning_rate": 1.221159740023915e-06, + "loss": 0.73721522, + "num_input_tokens_seen": 114285950, + "step": 5310, + "time_per_iteration": 2.718517780303955 + }, + { + "auxiliary_loss_clip": 0.0120997, + "auxiliary_loss_mlp": 0.01379117, + "balance_loss_clip": 1.04674625, + "balance_loss_mlp": 1.00029194, + "epoch": 0.6386099921842121, + "flos": 23988328306560.0, + "grad_norm": 1.8954719492483905, + "language_loss": 0.72751081, + "learning_rate": 1.2204423193057735e-06, + "loss": 0.75340164, + "num_input_tokens_seen": 114304780, + "step": 5311, + "time_per_iteration": 2.723850727081299 + }, + { + "auxiliary_loss_clip": 0.01095525, + "auxiliary_loss_mlp": 0.01006193, + "balance_loss_clip": 1.01514995, + "balance_loss_mlp": 1.00491202, + "epoch": 0.6387302350748512, + "flos": 71731169337600.0, + "grad_norm": 0.8484274659590749, + "language_loss": 0.63353091, + "learning_rate": 1.2197250168459122e-06, + "loss": 0.65454811, + "num_input_tokens_seen": 114361180, + "step": 5312, + "time_per_iteration": 4.086288690567017 + }, + { + "auxiliary_loss_clip": 0.0118678, + "auxiliary_loss_mlp": 0.01024551, + "balance_loss_clip": 1.05048585, + "balance_loss_mlp": 1.0166775, + "epoch": 0.6388504779654903, + "flos": 14535778141440.0, + "grad_norm": 1.883911228491466, + "language_loss": 0.74626064, + "learning_rate": 1.2190078327531454e-06, + "loss": 0.76837391, + "num_input_tokens_seen": 114377425, + "step": 5313, + "time_per_iteration": 3.621591806411743 + }, + { + "auxiliary_loss_clip": 0.01182646, + "auxiliary_loss_mlp": 0.01025431, + "balance_loss_clip": 1.04772735, + "balance_loss_mlp": 1.01782894, + "epoch": 0.6389707208561294, + "flos": 22346133384960.0, + "grad_norm": 1.4565705877509292, + "language_loss": 0.72935581, + "learning_rate": 1.2182907671362697e-06, + "loss": 0.75143659, + "num_input_tokens_seen": 114398120, + "step": 5314, + "time_per_iteration": 3.6017024517059326 + }, + { + "auxiliary_loss_clip": 0.01182223, + "auxiliary_loss_mlp": 0.01030428, + "balance_loss_clip": 1.04970658, + "balance_loss_mlp": 1.02253628, + "epoch": 0.6390909637467684, + "flos": 19426883247360.0, + "grad_norm": 2.0051314280154817, + "language_loss": 0.78588116, + "learning_rate": 1.2175738201040626e-06, + "loss": 0.80800772, + "num_input_tokens_seen": 114415160, + "step": 5315, + "time_per_iteration": 2.6185858249664307 + }, + { + "auxiliary_loss_clip": 0.01183818, + "auxiliary_loss_mlp": 0.01027321, + "balance_loss_clip": 1.0486939, + "balance_loss_mlp": 1.01900077, + "epoch": 0.6392112066374076, + "flos": 24090852700800.0, + "grad_norm": 1.8544076912309808, + "language_loss": 0.78557122, + "learning_rate": 1.2168569917652855e-06, + "loss": 0.80768263, + "num_input_tokens_seen": 114435015, + "step": 5316, + "time_per_iteration": 2.7246432304382324 + }, + { + "auxiliary_loss_clip": 0.01185868, + "auxiliary_loss_mlp": 0.01023127, + "balance_loss_clip": 1.05247951, + "balance_loss_mlp": 1.01526499, + "epoch": 0.6393314495280467, + "flos": 26795141896320.0, + "grad_norm": 1.7429363776189954, + "language_loss": 0.64179099, + "learning_rate": 1.2161402822286797e-06, + "loss": 0.66388094, + "num_input_tokens_seen": 114455700, + "step": 5317, + "time_per_iteration": 2.6723439693450928 + }, + { + "auxiliary_loss_clip": 0.01199578, + "auxiliary_loss_mlp": 0.01022435, + "balance_loss_clip": 1.04504859, + "balance_loss_mlp": 1.01502347, + "epoch": 0.6394516924186857, + "flos": 20260692633600.0, + "grad_norm": 2.2202337967586043, + "language_loss": 0.78915352, + "learning_rate": 1.2154236916029703e-06, + "loss": 0.81137365, + "num_input_tokens_seen": 114473675, + "step": 5318, + "time_per_iteration": 2.668628454208374 + }, + { + "auxiliary_loss_clip": 0.01207585, + "auxiliary_loss_mlp": 0.01029805, + "balance_loss_clip": 1.04107177, + "balance_loss_mlp": 1.02251554, + "epoch": 0.6395719353093249, + "flos": 18368847210240.0, + "grad_norm": 8.09865896689444, + "language_loss": 0.73863113, + "learning_rate": 1.2147072199968627e-06, + "loss": 0.76100504, + "num_input_tokens_seen": 114492310, + "step": 5319, + "time_per_iteration": 3.6776607036590576 + }, + { + "auxiliary_loss_clip": 0.011829, + "auxiliary_loss_mlp": 0.01028319, + "balance_loss_clip": 1.04966855, + "balance_loss_mlp": 1.02073193, + "epoch": 0.6396921781999639, + "flos": 17566315591680.0, + "grad_norm": 33.67213877538318, + "language_loss": 0.72001672, + "learning_rate": 1.2139908675190454e-06, + "loss": 0.74212891, + "num_input_tokens_seen": 114511520, + "step": 5320, + "time_per_iteration": 2.6262474060058594 + }, + { + "auxiliary_loss_clip": 0.01203047, + "auxiliary_loss_mlp": 0.01023517, + "balance_loss_clip": 1.03820372, + "balance_loss_mlp": 1.01570916, + "epoch": 0.639812421090603, + "flos": 21251252972160.0, + "grad_norm": 2.154039758107581, + "language_loss": 0.75156248, + "learning_rate": 1.2132746342781883e-06, + "loss": 0.77382809, + "num_input_tokens_seen": 114532680, + "step": 5321, + "time_per_iteration": 2.8586244583129883 + }, + { + "auxiliary_loss_clip": 0.01177993, + "auxiliary_loss_mlp": 0.01026565, + "balance_loss_clip": 1.05211198, + "balance_loss_mlp": 1.01861978, + "epoch": 0.6399326639812422, + "flos": 11180967684480.0, + "grad_norm": 2.8123930847423653, + "language_loss": 0.80293703, + "learning_rate": 1.2125585203829442e-06, + "loss": 0.82498264, + "num_input_tokens_seen": 114548320, + "step": 5322, + "time_per_iteration": 2.59918475151062 + }, + { + "auxiliary_loss_clip": 0.0118608, + "auxiliary_loss_mlp": 0.01025265, + "balance_loss_clip": 1.0451405, + "balance_loss_mlp": 1.01715851, + "epoch": 0.6400529068718812, + "flos": 23911048195200.0, + "grad_norm": 2.1932879883184553, + "language_loss": 0.74015057, + "learning_rate": 1.211842525941946e-06, + "loss": 0.76226395, + "num_input_tokens_seen": 114568115, + "step": 5323, + "time_per_iteration": 2.743218183517456 + }, + { + "auxiliary_loss_clip": 0.01207338, + "auxiliary_loss_mlp": 0.0102799, + "balance_loss_clip": 1.0451299, + "balance_loss_mlp": 1.0207566, + "epoch": 0.6401731497625203, + "flos": 44018724890880.0, + "grad_norm": 1.9511217621894243, + "language_loss": 0.78835857, + "learning_rate": 1.2111266510638105e-06, + "loss": 0.81071186, + "num_input_tokens_seen": 114591040, + "step": 5324, + "time_per_iteration": 2.980823516845703 + }, + { + "auxiliary_loss_clip": 0.01208906, + "auxiliary_loss_mlp": 0.01026649, + "balance_loss_clip": 1.04159474, + "balance_loss_mlp": 1.01907277, + "epoch": 0.6402933926531594, + "flos": 20662209838080.0, + "grad_norm": 1.6637537051508895, + "language_loss": 0.79886037, + "learning_rate": 1.2104108958571346e-06, + "loss": 0.82121587, + "num_input_tokens_seen": 114609310, + "step": 5325, + "time_per_iteration": 2.802079677581787 + }, + { + "auxiliary_loss_clip": 0.01181045, + "auxiliary_loss_mlp": 0.01023256, + "balance_loss_clip": 1.05021286, + "balance_loss_mlp": 1.01613021, + "epoch": 0.6404136355437985, + "flos": 24863327614080.0, + "grad_norm": 1.5928456526495098, + "language_loss": 0.75779605, + "learning_rate": 1.2096952604304975e-06, + "loss": 0.77983904, + "num_input_tokens_seen": 114629740, + "step": 5326, + "time_per_iteration": 2.604534387588501 + }, + { + "auxiliary_loss_clip": 0.0118732, + "auxiliary_loss_mlp": 0.01029467, + "balance_loss_clip": 1.05078578, + "balance_loss_mlp": 1.02121198, + "epoch": 0.6405338784344375, + "flos": 40479548901120.0, + "grad_norm": 1.9442083258071174, + "language_loss": 0.70195639, + "learning_rate": 1.2089797448924616e-06, + "loss": 0.72412431, + "num_input_tokens_seen": 114653615, + "step": 5327, + "time_per_iteration": 2.76841402053833 + }, + { + "auxiliary_loss_clip": 0.01210477, + "auxiliary_loss_mlp": 0.01024948, + "balance_loss_clip": 1.04039788, + "balance_loss_mlp": 1.01718175, + "epoch": 0.6406541213250767, + "flos": 20886041439360.0, + "grad_norm": 1.9751310791989418, + "language_loss": 0.65549695, + "learning_rate": 1.2082643493515692e-06, + "loss": 0.6778512, + "num_input_tokens_seen": 114671935, + "step": 5328, + "time_per_iteration": 2.6944363117218018 + }, + { + "auxiliary_loss_clip": 0.01182975, + "auxiliary_loss_mlp": 0.01032919, + "balance_loss_clip": 1.04927576, + "balance_loss_mlp": 1.02531362, + "epoch": 0.6407743642157158, + "flos": 23295970679040.0, + "grad_norm": 1.7789195349267057, + "language_loss": 0.81832874, + "learning_rate": 1.207549073916346e-06, + "loss": 0.84048772, + "num_input_tokens_seen": 114692870, + "step": 5329, + "time_per_iteration": 2.6471006870269775 + }, + { + "auxiliary_loss_clip": 0.01183787, + "auxiliary_loss_mlp": 0.01030467, + "balance_loss_clip": 1.04891372, + "balance_loss_mlp": 1.02265882, + "epoch": 0.6408946071063548, + "flos": 15012636122880.0, + "grad_norm": 2.8400294423906978, + "language_loss": 0.7754041, + "learning_rate": 1.2068339186952976e-06, + "loss": 0.79754663, + "num_input_tokens_seen": 114710410, + "step": 5330, + "time_per_iteration": 2.5959901809692383 + }, + { + "auxiliary_loss_clip": 0.01188071, + "auxiliary_loss_mlp": 0.01030419, + "balance_loss_clip": 1.0509057, + "balance_loss_mlp": 1.02239645, + "epoch": 0.6410148499969939, + "flos": 22528595496960.0, + "grad_norm": 1.7828501102082634, + "language_loss": 0.7285794, + "learning_rate": 1.2061188837969136e-06, + "loss": 0.75076425, + "num_input_tokens_seen": 114730020, + "step": 5331, + "time_per_iteration": 2.738797903060913 + }, + { + "auxiliary_loss_clip": 0.01192178, + "auxiliary_loss_mlp": 0.01024093, + "balance_loss_clip": 1.04244351, + "balance_loss_mlp": 1.01539087, + "epoch": 0.641135092887633, + "flos": 12422004537600.0, + "grad_norm": 2.3531321325854977, + "language_loss": 0.84263057, + "learning_rate": 1.2054039693296631e-06, + "loss": 0.8647933, + "num_input_tokens_seen": 114748015, + "step": 5332, + "time_per_iteration": 2.739168882369995 + }, + { + "auxiliary_loss_clip": 0.01192448, + "auxiliary_loss_mlp": 0.01025949, + "balance_loss_clip": 1.04460001, + "balance_loss_mlp": 1.0185523, + "epoch": 0.6412553357782721, + "flos": 22127329687680.0, + "grad_norm": 1.6806448739188513, + "language_loss": 0.8140049, + "learning_rate": 1.2046891754019992e-06, + "loss": 0.83618891, + "num_input_tokens_seen": 114768625, + "step": 5333, + "time_per_iteration": 2.687939405441284 + }, + { + "auxiliary_loss_clip": 0.01186862, + "auxiliary_loss_mlp": 0.01024908, + "balance_loss_clip": 1.05096149, + "balance_loss_mlp": 1.01703441, + "epoch": 0.6413755786689112, + "flos": 15888605097600.0, + "grad_norm": 1.989286982671974, + "language_loss": 0.82686973, + "learning_rate": 1.2039745021223548e-06, + "loss": 0.8489874, + "num_input_tokens_seen": 114786045, + "step": 5334, + "time_per_iteration": 2.5680856704711914 + }, + { + "auxiliary_loss_clip": 0.01116153, + "auxiliary_loss_mlp": 0.01000955, + "balance_loss_clip": 1.0157795, + "balance_loss_mlp": 0.99967939, + "epoch": 0.6414958215595503, + "flos": 68039159955840.0, + "grad_norm": 0.7975217037206958, + "language_loss": 0.57078469, + "learning_rate": 1.2032599495991456e-06, + "loss": 0.59195566, + "num_input_tokens_seen": 114850785, + "step": 5335, + "time_per_iteration": 3.374798059463501 + }, + { + "auxiliary_loss_clip": 0.01182216, + "auxiliary_loss_mlp": 0.01021781, + "balance_loss_clip": 1.04921627, + "balance_loss_mlp": 1.01430047, + "epoch": 0.6416160644501894, + "flos": 44091300320640.0, + "grad_norm": 2.2318157129227987, + "language_loss": 0.69679034, + "learning_rate": 1.2025455179407685e-06, + "loss": 0.71883029, + "num_input_tokens_seen": 114871945, + "step": 5336, + "time_per_iteration": 2.800156593322754 + }, + { + "auxiliary_loss_clip": 0.011838, + "auxiliary_loss_mlp": 0.01379238, + "balance_loss_clip": 1.05113566, + "balance_loss_mlp": 1.0002346, + "epoch": 0.6417363073408284, + "flos": 20959837931520.0, + "grad_norm": 2.029960066680311, + "language_loss": 0.73614794, + "learning_rate": 1.2018312072556022e-06, + "loss": 0.76177835, + "num_input_tokens_seen": 114890445, + "step": 5337, + "time_per_iteration": 3.5316884517669678 + }, + { + "auxiliary_loss_clip": 0.01171685, + "auxiliary_loss_mlp": 0.0137903, + "balance_loss_clip": 1.04924703, + "balance_loss_mlp": 1.0002439, + "epoch": 0.6418565502314676, + "flos": 22455122227200.0, + "grad_norm": 1.8622936062213762, + "language_loss": 0.74392962, + "learning_rate": 1.2011170176520077e-06, + "loss": 0.76943678, + "num_input_tokens_seen": 114911360, + "step": 5338, + "time_per_iteration": 2.582526922225952 + }, + { + "auxiliary_loss_clip": 0.01197022, + "auxiliary_loss_mlp": 0.01024482, + "balance_loss_clip": 1.04007936, + "balance_loss_mlp": 1.01656997, + "epoch": 0.6419767931221066, + "flos": 25045502417280.0, + "grad_norm": 1.5073817521834019, + "language_loss": 0.81098616, + "learning_rate": 1.2004029492383256e-06, + "loss": 0.83320123, + "num_input_tokens_seen": 114932700, + "step": 5339, + "time_per_iteration": 3.705080032348633 + }, + { + "auxiliary_loss_clip": 0.01185002, + "auxiliary_loss_mlp": 0.01035678, + "balance_loss_clip": 1.05227923, + "balance_loss_mlp": 1.02760744, + "epoch": 0.6420970360127457, + "flos": 19463691709440.0, + "grad_norm": 2.246079159345773, + "language_loss": 0.73318994, + "learning_rate": 1.1996890021228814e-06, + "loss": 0.75539672, + "num_input_tokens_seen": 114949475, + "step": 5340, + "time_per_iteration": 2.5857019424438477 + }, + { + "auxiliary_loss_clip": 0.01187682, + "auxiliary_loss_mlp": 0.01022213, + "balance_loss_clip": 1.04535556, + "balance_loss_mlp": 1.01457119, + "epoch": 0.6422172789033849, + "flos": 40406147458560.0, + "grad_norm": 1.4502617852365927, + "language_loss": 0.70118469, + "learning_rate": 1.1989751764139785e-06, + "loss": 0.72328365, + "num_input_tokens_seen": 114973125, + "step": 5341, + "time_per_iteration": 3.7818703651428223 + }, + { + "auxiliary_loss_clip": 0.01202464, + "auxiliary_loss_mlp": 0.01023204, + "balance_loss_clip": 1.03887606, + "balance_loss_mlp": 1.01551509, + "epoch": 0.6423375217940239, + "flos": 27672870637440.0, + "grad_norm": 1.5332772770068088, + "language_loss": 0.83138645, + "learning_rate": 1.1982614722199044e-06, + "loss": 0.85364318, + "num_input_tokens_seen": 114994300, + "step": 5342, + "time_per_iteration": 2.849902868270874 + }, + { + "auxiliary_loss_clip": 0.01193353, + "auxiliary_loss_mlp": 0.01026297, + "balance_loss_clip": 1.04442036, + "balance_loss_mlp": 1.01848328, + "epoch": 0.642457764684663, + "flos": 18369242259840.0, + "grad_norm": 2.2513755475607975, + "language_loss": 0.78121227, + "learning_rate": 1.1975478896489276e-06, + "loss": 0.8034088, + "num_input_tokens_seen": 115012135, + "step": 5343, + "time_per_iteration": 2.6611039638519287 + }, + { + "auxiliary_loss_clip": 0.01169806, + "auxiliary_loss_mlp": 0.01023493, + "balance_loss_clip": 1.04743874, + "balance_loss_mlp": 1.01589954, + "epoch": 0.6425780075753021, + "flos": 19750509809280.0, + "grad_norm": 1.9253192279630007, + "language_loss": 0.76476365, + "learning_rate": 1.1968344288092981e-06, + "loss": 0.78669661, + "num_input_tokens_seen": 115028715, + "step": 5344, + "time_per_iteration": 2.656059503555298 + }, + { + "auxiliary_loss_clip": 0.01183539, + "auxiliary_loss_mlp": 0.01378935, + "balance_loss_clip": 1.04929876, + "balance_loss_mlp": 1.00020885, + "epoch": 0.6426982504659412, + "flos": 20558536208640.0, + "grad_norm": 1.6862650602471991, + "language_loss": 0.64549869, + "learning_rate": 1.1961210898092468e-06, + "loss": 0.67112345, + "num_input_tokens_seen": 115047665, + "step": 5345, + "time_per_iteration": 3.55251145362854 + }, + { + "auxiliary_loss_clip": 0.01197646, + "auxiliary_loss_mlp": 0.01030892, + "balance_loss_clip": 1.04795039, + "balance_loss_mlp": 1.02263045, + "epoch": 0.6428184933565803, + "flos": 17851984456320.0, + "grad_norm": 2.1315348136779884, + "language_loss": 0.79331446, + "learning_rate": 1.1954078727569874e-06, + "loss": 0.8155998, + "num_input_tokens_seen": 115064965, + "step": 5346, + "time_per_iteration": 2.6762149333953857 + }, + { + "auxiliary_loss_clip": 0.01206956, + "auxiliary_loss_mlp": 0.01379293, + "balance_loss_clip": 1.04709435, + "balance_loss_mlp": 1.00023639, + "epoch": 0.6429387362472194, + "flos": 22456953820800.0, + "grad_norm": 1.6510837083640886, + "language_loss": 0.7801435, + "learning_rate": 1.1946947777607141e-06, + "loss": 0.80600601, + "num_input_tokens_seen": 115086100, + "step": 5347, + "time_per_iteration": 2.6968560218811035 + }, + { + "auxiliary_loss_clip": 0.01199355, + "auxiliary_loss_mlp": 0.01028809, + "balance_loss_clip": 1.04392385, + "balance_loss_mlp": 1.02100706, + "epoch": 0.6430589791378585, + "flos": 24752579005440.0, + "grad_norm": 1.8940570163634587, + "language_loss": 0.80083907, + "learning_rate": 1.1939818049286024e-06, + "loss": 0.82312065, + "num_input_tokens_seen": 115104260, + "step": 5348, + "time_per_iteration": 2.73938250541687 + }, + { + "auxiliary_loss_clip": 0.0120013, + "auxiliary_loss_mlp": 0.0102906, + "balance_loss_clip": 1.04021621, + "balance_loss_mlp": 1.02147841, + "epoch": 0.6431792220284975, + "flos": 24901249397760.0, + "grad_norm": 1.7522352318267247, + "language_loss": 0.76009762, + "learning_rate": 1.1932689543688101e-06, + "loss": 0.78238952, + "num_input_tokens_seen": 115125365, + "step": 5349, + "time_per_iteration": 2.883500337600708 + }, + { + "auxiliary_loss_clip": 0.01192913, + "auxiliary_loss_mlp": 0.01031178, + "balance_loss_clip": 1.04991722, + "balance_loss_mlp": 1.02265429, + "epoch": 0.6432994649191367, + "flos": 21032305620480.0, + "grad_norm": 2.3032532969338724, + "language_loss": 0.72510731, + "learning_rate": 1.1925562261894756e-06, + "loss": 0.74734819, + "num_input_tokens_seen": 115144445, + "step": 5350, + "time_per_iteration": 2.7375645637512207 + }, + { + "auxiliary_loss_clip": 0.01188253, + "auxiliary_loss_mlp": 0.01024597, + "balance_loss_clip": 1.04718804, + "balance_loss_mlp": 1.01746547, + "epoch": 0.6434197078097758, + "flos": 30884433655680.0, + "grad_norm": 1.6268199961224952, + "language_loss": 0.77506828, + "learning_rate": 1.1918436204987207e-06, + "loss": 0.79719687, + "num_input_tokens_seen": 115166305, + "step": 5351, + "time_per_iteration": 2.7946066856384277 + }, + { + "auxiliary_loss_clip": 0.01177983, + "auxiliary_loss_mlp": 0.01029277, + "balance_loss_clip": 1.04806447, + "balance_loss_mlp": 1.02185655, + "epoch": 0.6435399507004148, + "flos": 15012492468480.0, + "grad_norm": 2.0723624907925764, + "language_loss": 0.81428921, + "learning_rate": 1.191131137404645e-06, + "loss": 0.83636183, + "num_input_tokens_seen": 115183045, + "step": 5352, + "time_per_iteration": 2.5423028469085693 + }, + { + "auxiliary_loss_clip": 0.01184594, + "auxiliary_loss_mlp": 0.01032027, + "balance_loss_clip": 1.04314613, + "balance_loss_mlp": 1.02394462, + "epoch": 0.643660193591054, + "flos": 19901981462400.0, + "grad_norm": 1.824049373855697, + "language_loss": 0.77494901, + "learning_rate": 1.190418777015333e-06, + "loss": 0.79711521, + "num_input_tokens_seen": 115201955, + "step": 5353, + "time_per_iteration": 2.7405130863189697 + }, + { + "auxiliary_loss_clip": 0.0119218, + "auxiliary_loss_mlp": 0.0102265, + "balance_loss_clip": 1.04897606, + "balance_loss_mlp": 1.01534891, + "epoch": 0.643780436481693, + "flos": 24133622820480.0, + "grad_norm": 1.4338343538215497, + "language_loss": 0.73407823, + "learning_rate": 1.1897065394388487e-06, + "loss": 0.75622654, + "num_input_tokens_seen": 115222395, + "step": 5354, + "time_per_iteration": 2.6182034015655518 + }, + { + "auxiliary_loss_clip": 0.01194609, + "auxiliary_loss_mlp": 0.01035769, + "balance_loss_clip": 1.05232608, + "balance_loss_mlp": 1.02812195, + "epoch": 0.6439006793723321, + "flos": 23148808657920.0, + "grad_norm": 1.6711960179770495, + "language_loss": 0.76704973, + "learning_rate": 1.1889944247832385e-06, + "loss": 0.78935349, + "num_input_tokens_seen": 115242635, + "step": 5355, + "time_per_iteration": 2.693814277648926 + }, + { + "auxiliary_loss_clip": 0.01184171, + "auxiliary_loss_mlp": 0.01027524, + "balance_loss_clip": 1.04770756, + "balance_loss_mlp": 1.01997852, + "epoch": 0.6440209222629713, + "flos": 23617909301760.0, + "grad_norm": 1.9054362636618705, + "language_loss": 0.70930493, + "learning_rate": 1.1882824331565283e-06, + "loss": 0.73142195, + "num_input_tokens_seen": 115262095, + "step": 5356, + "time_per_iteration": 2.6282436847686768 + }, + { + "auxiliary_loss_clip": 0.01192257, + "auxiliary_loss_mlp": 0.01024052, + "balance_loss_clip": 1.04268527, + "balance_loss_mlp": 1.01685762, + "epoch": 0.6441411651536103, + "flos": 16544872535040.0, + "grad_norm": 3.75031056306803, + "language_loss": 0.8920083, + "learning_rate": 1.1875705646667287e-06, + "loss": 0.91417146, + "num_input_tokens_seen": 115279985, + "step": 5357, + "time_per_iteration": 2.719318151473999 + }, + { + "auxiliary_loss_clip": 0.01176763, + "auxiliary_loss_mlp": 0.01028775, + "balance_loss_clip": 1.04576373, + "balance_loss_mlp": 1.02099657, + "epoch": 0.6442614080442494, + "flos": 25410965345280.0, + "grad_norm": 2.6517252362404276, + "language_loss": 0.75510103, + "learning_rate": 1.1868588194218282e-06, + "loss": 0.77715647, + "num_input_tokens_seen": 115300365, + "step": 5358, + "time_per_iteration": 2.6392414569854736 + }, + { + "auxiliary_loss_clip": 0.01196123, + "auxiliary_loss_mlp": 0.01028609, + "balance_loss_clip": 1.04632509, + "balance_loss_mlp": 1.02058053, + "epoch": 0.6443816509348885, + "flos": 28294017552000.0, + "grad_norm": 1.733164862899651, + "language_loss": 0.7403394, + "learning_rate": 1.1861471975297979e-06, + "loss": 0.76258671, + "num_input_tokens_seen": 115322060, + "step": 5359, + "time_per_iteration": 2.7512435913085938 + }, + { + "auxiliary_loss_clip": 0.01195915, + "auxiliary_loss_mlp": 0.01029914, + "balance_loss_clip": 1.04623437, + "balance_loss_mlp": 1.02200413, + "epoch": 0.6445018938255276, + "flos": 36690075964800.0, + "grad_norm": 2.1227910332703503, + "language_loss": 0.70992243, + "learning_rate": 1.185435699098591e-06, + "loss": 0.73218071, + "num_input_tokens_seen": 115348255, + "step": 5360, + "time_per_iteration": 2.844221353530884 + }, + { + "auxiliary_loss_clip": 0.01195365, + "auxiliary_loss_mlp": 0.01030099, + "balance_loss_clip": 1.04823732, + "balance_loss_mlp": 1.02222228, + "epoch": 0.6446221367161666, + "flos": 14501411804160.0, + "grad_norm": 2.354377603308917, + "language_loss": 0.78088623, + "learning_rate": 1.1847243242361403e-06, + "loss": 0.80314088, + "num_input_tokens_seen": 115366845, + "step": 5361, + "time_per_iteration": 2.6490111351013184 + }, + { + "auxiliary_loss_clip": 0.01192321, + "auxiliary_loss_mlp": 0.0102611, + "balance_loss_clip": 1.046826, + "balance_loss_mlp": 1.01848686, + "epoch": 0.6447423796068057, + "flos": 24609367480320.0, + "grad_norm": 2.1616617325028544, + "language_loss": 0.78128409, + "learning_rate": 1.1840130730503624e-06, + "loss": 0.80346847, + "num_input_tokens_seen": 115388125, + "step": 5362, + "time_per_iteration": 2.7627203464508057 + }, + { + "auxiliary_loss_clip": 0.01173974, + "auxiliary_loss_mlp": 0.01024005, + "balance_loss_clip": 1.04975533, + "balance_loss_mlp": 1.01591063, + "epoch": 0.6448626224974449, + "flos": 25047298097280.0, + "grad_norm": 1.996935039899444, + "language_loss": 0.7456702, + "learning_rate": 1.1833019456491518e-06, + "loss": 0.76765001, + "num_input_tokens_seen": 115409655, + "step": 5363, + "time_per_iteration": 3.5349667072296143 + }, + { + "auxiliary_loss_clip": 0.0118574, + "auxiliary_loss_mlp": 0.01024868, + "balance_loss_clip": 1.05019212, + "balance_loss_mlp": 1.01739311, + "epoch": 0.6449828653880839, + "flos": 22530355263360.0, + "grad_norm": 1.8712880331888193, + "language_loss": 0.78727365, + "learning_rate": 1.1825909421403871e-06, + "loss": 0.80937976, + "num_input_tokens_seen": 115428750, + "step": 5364, + "time_per_iteration": 2.6814587116241455 + }, + { + "auxiliary_loss_clip": 0.01184928, + "auxiliary_loss_mlp": 0.01024279, + "balance_loss_clip": 1.04891539, + "balance_loss_mlp": 1.01676857, + "epoch": 0.645103108278723, + "flos": 25695736369920.0, + "grad_norm": 1.859583082781644, + "language_loss": 0.76351678, + "learning_rate": 1.181880062631926e-06, + "loss": 0.78560889, + "num_input_tokens_seen": 115448085, + "step": 5365, + "time_per_iteration": 3.557772397994995 + }, + { + "auxiliary_loss_clip": 0.01185445, + "auxiliary_loss_mlp": 0.01029155, + "balance_loss_clip": 1.04703331, + "balance_loss_mlp": 1.02101326, + "epoch": 0.6452233511693621, + "flos": 27450331925760.0, + "grad_norm": 1.9625456843134053, + "language_loss": 0.84411156, + "learning_rate": 1.1811693072316093e-06, + "loss": 0.86625755, + "num_input_tokens_seen": 115465765, + "step": 5366, + "time_per_iteration": 3.5982508659362793 + }, + { + "auxiliary_loss_clip": 0.01173962, + "auxiliary_loss_mlp": 0.01378924, + "balance_loss_clip": 1.04890037, + "balance_loss_mlp": 1.00017571, + "epoch": 0.6453435940600012, + "flos": 19208618254080.0, + "grad_norm": 3.919188179171582, + "language_loss": 0.83542645, + "learning_rate": 1.1804586760472574e-06, + "loss": 0.86095524, + "num_input_tokens_seen": 115482230, + "step": 5367, + "time_per_iteration": 2.66928768157959 + }, + { + "auxiliary_loss_clip": 0.01198724, + "auxiliary_loss_mlp": 0.0102581, + "balance_loss_clip": 1.04432917, + "balance_loss_mlp": 1.01804972, + "epoch": 0.6454638369506402, + "flos": 25737680476800.0, + "grad_norm": 2.078639503492047, + "language_loss": 0.80288881, + "learning_rate": 1.1797481691866736e-06, + "loss": 0.82513422, + "num_input_tokens_seen": 115499455, + "step": 5368, + "time_per_iteration": 2.72037672996521 + }, + { + "auxiliary_loss_clip": 0.01180178, + "auxiliary_loss_mlp": 0.01029747, + "balance_loss_clip": 1.04482818, + "balance_loss_mlp": 1.02244532, + "epoch": 0.6455840798412794, + "flos": 20989176364800.0, + "grad_norm": 2.2379748378128044, + "language_loss": 0.83023602, + "learning_rate": 1.1790377867576393e-06, + "loss": 0.85233527, + "num_input_tokens_seen": 115517205, + "step": 5369, + "time_per_iteration": 2.6844935417175293 + }, + { + "auxiliary_loss_clip": 0.01193686, + "auxiliary_loss_mlp": 0.0102879, + "balance_loss_clip": 1.04582202, + "balance_loss_mlp": 1.02158976, + "epoch": 0.6457043227319185, + "flos": 26067556005120.0, + "grad_norm": 1.7298031489629502, + "language_loss": 0.76900846, + "learning_rate": 1.1783275288679203e-06, + "loss": 0.79123318, + "num_input_tokens_seen": 115534370, + "step": 5370, + "time_per_iteration": 2.7405805587768555 + }, + { + "auxiliary_loss_clip": 0.01081283, + "auxiliary_loss_mlp": 0.01003965, + "balance_loss_clip": 1.01309156, + "balance_loss_mlp": 1.0026778, + "epoch": 0.6458245656225575, + "flos": 60370831088640.0, + "grad_norm": 0.8415691516497723, + "language_loss": 0.57132554, + "learning_rate": 1.177617395625262e-06, + "loss": 0.59217799, + "num_input_tokens_seen": 115592345, + "step": 5371, + "time_per_iteration": 4.076406955718994 + }, + { + "auxiliary_loss_clip": 0.01185586, + "auxiliary_loss_mlp": 0.01031065, + "balance_loss_clip": 1.0517478, + "balance_loss_mlp": 1.02367115, + "epoch": 0.6459448085131967, + "flos": 23076771932160.0, + "grad_norm": 2.047884881029733, + "language_loss": 0.75253183, + "learning_rate": 1.1769073871373908e-06, + "loss": 0.77469832, + "num_input_tokens_seen": 115612550, + "step": 5372, + "time_per_iteration": 2.6703603267669678 + }, + { + "auxiliary_loss_clip": 0.01195458, + "auxiliary_loss_mlp": 0.01026385, + "balance_loss_clip": 1.04296207, + "balance_loss_mlp": 1.01816583, + "epoch": 0.6460650514038357, + "flos": 22598190097920.0, + "grad_norm": 1.5516702644135376, + "language_loss": 0.83858752, + "learning_rate": 1.176197503512015e-06, + "loss": 0.86080599, + "num_input_tokens_seen": 115632265, + "step": 5373, + "time_per_iteration": 2.7072906494140625 + }, + { + "auxiliary_loss_clip": 0.01184623, + "auxiliary_loss_mlp": 0.01029281, + "balance_loss_clip": 1.04613984, + "balance_loss_mlp": 1.02198195, + "epoch": 0.6461852942944748, + "flos": 20266726118400.0, + "grad_norm": 2.1661698021090636, + "language_loss": 0.82229936, + "learning_rate": 1.1754877448568223e-06, + "loss": 0.84443843, + "num_input_tokens_seen": 115651720, + "step": 5374, + "time_per_iteration": 2.6560397148132324 + }, + { + "auxiliary_loss_clip": 0.01190906, + "auxiliary_loss_mlp": 0.01022983, + "balance_loss_clip": 1.04734087, + "balance_loss_mlp": 1.01558638, + "epoch": 0.646305537185114, + "flos": 23367109564800.0, + "grad_norm": 2.224126308577742, + "language_loss": 0.89741987, + "learning_rate": 1.1747781112794837e-06, + "loss": 0.91955876, + "num_input_tokens_seen": 115668215, + "step": 5375, + "time_per_iteration": 2.7088968753814697 + }, + { + "auxiliary_loss_clip": 0.01194378, + "auxiliary_loss_mlp": 0.010288, + "balance_loss_clip": 1.04565859, + "balance_loss_mlp": 1.02151, + "epoch": 0.646425780075753, + "flos": 24277480790400.0, + "grad_norm": 1.605006437924, + "language_loss": 0.83261341, + "learning_rate": 1.1740686028876487e-06, + "loss": 0.85484517, + "num_input_tokens_seen": 115687080, + "step": 5376, + "time_per_iteration": 2.7672383785247803 + }, + { + "auxiliary_loss_clip": 0.01177395, + "auxiliary_loss_mlp": 0.01028824, + "balance_loss_clip": 1.04850316, + "balance_loss_mlp": 1.02109909, + "epoch": 0.6465460229663921, + "flos": 20813968800000.0, + "grad_norm": 2.4123293740199765, + "language_loss": 0.75314546, + "learning_rate": 1.1733592197889507e-06, + "loss": 0.7752077, + "num_input_tokens_seen": 115703990, + "step": 5377, + "time_per_iteration": 2.655881881713867 + }, + { + "auxiliary_loss_clip": 0.01177181, + "auxiliary_loss_mlp": 0.01024613, + "balance_loss_clip": 1.04931855, + "balance_loss_mlp": 1.0173831, + "epoch": 0.6466662658570312, + "flos": 22853299466880.0, + "grad_norm": 3.04119184331384, + "language_loss": 0.72662181, + "learning_rate": 1.1726499620910014e-06, + "loss": 0.74863976, + "num_input_tokens_seen": 115724270, + "step": 5378, + "time_per_iteration": 2.7216594219207764 + }, + { + "auxiliary_loss_clip": 0.01182084, + "auxiliary_loss_mlp": 0.01027606, + "balance_loss_clip": 1.04949784, + "balance_loss_mlp": 1.01964259, + "epoch": 0.6467865087476703, + "flos": 15304553953920.0, + "grad_norm": 1.9941392707845311, + "language_loss": 0.78191572, + "learning_rate": 1.1719408299013955e-06, + "loss": 0.80401266, + "num_input_tokens_seen": 115742995, + "step": 5379, + "time_per_iteration": 2.8024027347564697 + }, + { + "auxiliary_loss_clip": 0.0117248, + "auxiliary_loss_mlp": 0.01024908, + "balance_loss_clip": 1.05089521, + "balance_loss_mlp": 1.01777935, + "epoch": 0.6469067516383094, + "flos": 19573650218880.0, + "grad_norm": 2.467780495652918, + "language_loss": 0.75518256, + "learning_rate": 1.1712318233277067e-06, + "loss": 0.77715647, + "num_input_tokens_seen": 115762015, + "step": 5380, + "time_per_iteration": 2.842061758041382 + }, + { + "auxiliary_loss_clip": 0.01081186, + "auxiliary_loss_mlp": 0.01001589, + "balance_loss_clip": 1.01370168, + "balance_loss_mlp": 1.00025415, + "epoch": 0.6470269945289485, + "flos": 65098002522240.0, + "grad_norm": 0.7513433001531942, + "language_loss": 0.57832134, + "learning_rate": 1.1705229424774916e-06, + "loss": 0.59914911, + "num_input_tokens_seen": 115816285, + "step": 5381, + "time_per_iteration": 3.2180826663970947 + }, + { + "auxiliary_loss_clip": 0.01185492, + "auxiliary_loss_mlp": 0.01025373, + "balance_loss_clip": 1.04427183, + "balance_loss_mlp": 1.01751757, + "epoch": 0.6471472374195876, + "flos": 30696943639680.0, + "grad_norm": 1.63903769949151, + "language_loss": 0.63982445, + "learning_rate": 1.1698141874582867e-06, + "loss": 0.66193306, + "num_input_tokens_seen": 115837330, + "step": 5382, + "time_per_iteration": 2.72243595123291 + }, + { + "auxiliary_loss_clip": 0.01172309, + "auxiliary_loss_mlp": 0.01028583, + "balance_loss_clip": 1.05219507, + "balance_loss_mlp": 1.02162969, + "epoch": 0.6472674803102266, + "flos": 20521835487360.0, + "grad_norm": 1.8090754279799022, + "language_loss": 0.7184397, + "learning_rate": 1.169105558377609e-06, + "loss": 0.74044859, + "num_input_tokens_seen": 115857420, + "step": 5383, + "time_per_iteration": 2.60947585105896 + }, + { + "auxiliary_loss_clip": 0.01207642, + "auxiliary_loss_mlp": 0.01379033, + "balance_loss_clip": 1.04764962, + "balance_loss_mlp": 1.00021577, + "epoch": 0.6473877232008658, + "flos": 24715447320960.0, + "grad_norm": 1.8994640887858685, + "language_loss": 0.78175926, + "learning_rate": 1.1683970553429587e-06, + "loss": 0.80762601, + "num_input_tokens_seen": 115878875, + "step": 5384, + "time_per_iteration": 2.8854024410247803 + }, + { + "auxiliary_loss_clip": 0.01202516, + "auxiliary_loss_mlp": 0.01025386, + "balance_loss_clip": 1.04680419, + "balance_loss_mlp": 1.01768196, + "epoch": 0.6475079660915048, + "flos": 15885552441600.0, + "grad_norm": 1.7980729788131886, + "language_loss": 0.82021618, + "learning_rate": 1.1676886784618128e-06, + "loss": 0.8424952, + "num_input_tokens_seen": 115895540, + "step": 5385, + "time_per_iteration": 2.7353415489196777 + }, + { + "auxiliary_loss_clip": 0.01182835, + "auxiliary_loss_mlp": 0.01024132, + "balance_loss_clip": 1.05006337, + "balance_loss_mlp": 1.01613939, + "epoch": 0.6476282089821439, + "flos": 17381590922880.0, + "grad_norm": 2.8415089574406696, + "language_loss": 0.83447564, + "learning_rate": 1.1669804278416332e-06, + "loss": 0.85654533, + "num_input_tokens_seen": 115910265, + "step": 5386, + "time_per_iteration": 2.6857125759124756 + }, + { + "auxiliary_loss_clip": 0.01196418, + "auxiliary_loss_mlp": 0.01023353, + "balance_loss_clip": 1.04798639, + "balance_loss_mlp": 1.01564646, + "epoch": 0.6477484518727831, + "flos": 20194078861440.0, + "grad_norm": 2.4002569073397213, + "language_loss": 0.71682751, + "learning_rate": 1.1662723035898602e-06, + "loss": 0.73902524, + "num_input_tokens_seen": 115930025, + "step": 5387, + "time_per_iteration": 2.7391953468322754 + }, + { + "auxiliary_loss_clip": 0.01180523, + "auxiliary_loss_mlp": 0.01024877, + "balance_loss_clip": 1.0485189, + "balance_loss_mlp": 1.01693738, + "epoch": 0.6478686947634221, + "flos": 25410426641280.0, + "grad_norm": 2.166442588788713, + "language_loss": 0.82018346, + "learning_rate": 1.165564305813915e-06, + "loss": 0.84223747, + "num_input_tokens_seen": 115949025, + "step": 5388, + "time_per_iteration": 2.7289533615112305 + }, + { + "auxiliary_loss_clip": 0.01181822, + "auxiliary_loss_mlp": 0.01028699, + "balance_loss_clip": 1.04869449, + "balance_loss_mlp": 1.02139783, + "epoch": 0.6479889376540612, + "flos": 20083581648000.0, + "grad_norm": 1.7038386008099409, + "language_loss": 0.81029791, + "learning_rate": 1.1648564346212019e-06, + "loss": 0.83240312, + "num_input_tokens_seen": 115968145, + "step": 5389, + "time_per_iteration": 3.58335542678833 + }, + { + "auxiliary_loss_clip": 0.01178889, + "auxiliary_loss_mlp": 0.01027866, + "balance_loss_clip": 1.05003643, + "balance_loss_mlp": 1.02096367, + "epoch": 0.6481091805447003, + "flos": 26758082039040.0, + "grad_norm": 2.2017402786962608, + "language_loss": 0.76165903, + "learning_rate": 1.164148690119104e-06, + "loss": 0.78372657, + "num_input_tokens_seen": 115989425, + "step": 5390, + "time_per_iteration": 2.697908878326416 + }, + { + "auxiliary_loss_clip": 0.01168595, + "auxiliary_loss_mlp": 0.01025293, + "balance_loss_clip": 1.04828072, + "balance_loss_mlp": 1.01812816, + "epoch": 0.6482294234353394, + "flos": 23952094462080.0, + "grad_norm": 2.134545590757931, + "language_loss": 0.74165601, + "learning_rate": 1.163441072414985e-06, + "loss": 0.76359487, + "num_input_tokens_seen": 116009630, + "step": 5391, + "time_per_iteration": 3.6698496341705322 + }, + { + "auxiliary_loss_clip": 0.01179737, + "auxiliary_loss_mlp": 0.01031493, + "balance_loss_clip": 1.04952204, + "balance_loss_mlp": 1.02406645, + "epoch": 0.6483496663259785, + "flos": 26209833776640.0, + "grad_norm": 1.8602467452204223, + "language_loss": 0.69893128, + "learning_rate": 1.16273358161619e-06, + "loss": 0.72104359, + "num_input_tokens_seen": 116029965, + "step": 5392, + "time_per_iteration": 3.7255969047546387 + }, + { + "auxiliary_loss_clip": 0.0119892, + "auxiliary_loss_mlp": 0.01024944, + "balance_loss_clip": 1.04847968, + "balance_loss_mlp": 1.01760936, + "epoch": 0.6484699092166175, + "flos": 20922239370240.0, + "grad_norm": 5.395989410321068, + "language_loss": 0.83872503, + "learning_rate": 1.1620262178300446e-06, + "loss": 0.86096364, + "num_input_tokens_seen": 116048580, + "step": 5393, + "time_per_iteration": 2.7280709743499756 + }, + { + "auxiliary_loss_clip": 0.0119866, + "auxiliary_loss_mlp": 0.01022579, + "balance_loss_clip": 1.04430676, + "balance_loss_mlp": 1.01503611, + "epoch": 0.6485901521072567, + "flos": 33072865678080.0, + "grad_norm": 1.717566353235211, + "language_loss": 0.75771344, + "learning_rate": 1.1613189811638563e-06, + "loss": 0.77992576, + "num_input_tokens_seen": 116070305, + "step": 5394, + "time_per_iteration": 2.8294577598571777 + }, + { + "auxiliary_loss_clip": 0.01183029, + "auxiliary_loss_mlp": 0.01024149, + "balance_loss_clip": 1.04967368, + "balance_loss_mlp": 1.01702023, + "epoch": 0.6487103949978957, + "flos": 22274060745600.0, + "grad_norm": 1.8173614014638535, + "language_loss": 0.7795704, + "learning_rate": 1.1606118717249117e-06, + "loss": 0.80164218, + "num_input_tokens_seen": 116090405, + "step": 5395, + "time_per_iteration": 2.727219581604004 + }, + { + "auxiliary_loss_clip": 0.01173853, + "auxiliary_loss_mlp": 0.01027363, + "balance_loss_clip": 1.04760146, + "balance_loss_mlp": 1.01979899, + "epoch": 0.6488306378885348, + "flos": 22930400010240.0, + "grad_norm": 2.0282269997268196, + "language_loss": 0.67838836, + "learning_rate": 1.1599048896204787e-06, + "loss": 0.70040059, + "num_input_tokens_seen": 116110285, + "step": 5396, + "time_per_iteration": 2.665961980819702 + }, + { + "auxiliary_loss_clip": 0.0120054, + "auxiliary_loss_mlp": 0.01026534, + "balance_loss_clip": 1.04428852, + "balance_loss_mlp": 1.01875603, + "epoch": 0.648950880779174, + "flos": 20376110010240.0, + "grad_norm": 1.7108037424617328, + "language_loss": 0.805884, + "learning_rate": 1.1591980349578061e-06, + "loss": 0.82815468, + "num_input_tokens_seen": 116128955, + "step": 5397, + "time_per_iteration": 3.6467080116271973 + }, + { + "auxiliary_loss_clip": 0.01101513, + "auxiliary_loss_mlp": 0.0100143, + "balance_loss_clip": 1.01016688, + "balance_loss_mlp": 1.0000596, + "epoch": 0.649071123669813, + "flos": 59930889310080.0, + "grad_norm": 0.7387573372324544, + "language_loss": 0.54230285, + "learning_rate": 1.158491307844123e-06, + "loss": 0.5633322, + "num_input_tokens_seen": 116188875, + "step": 5398, + "time_per_iteration": 3.270744800567627 + }, + { + "auxiliary_loss_clip": 0.01190816, + "auxiliary_loss_mlp": 0.01028518, + "balance_loss_clip": 1.04936135, + "balance_loss_mlp": 1.02109122, + "epoch": 0.6491913665604521, + "flos": 20446566537600.0, + "grad_norm": 1.6503886623685728, + "language_loss": 0.83503163, + "learning_rate": 1.1577847083866387e-06, + "loss": 0.85722494, + "num_input_tokens_seen": 116207910, + "step": 5399, + "time_per_iteration": 2.7999844551086426 + }, + { + "auxiliary_loss_clip": 0.0117669, + "auxiliary_loss_mlp": 0.0102681, + "balance_loss_clip": 1.0445559, + "balance_loss_mlp": 1.01906753, + "epoch": 0.6493116094510912, + "flos": 16946820702720.0, + "grad_norm": 1.9587305685215959, + "language_loss": 0.72230136, + "learning_rate": 1.1570782366925453e-06, + "loss": 0.74433637, + "num_input_tokens_seen": 116226425, + "step": 5400, + "time_per_iteration": 2.6425514221191406 + }, + { + "auxiliary_loss_clip": 0.01190786, + "auxiliary_loss_mlp": 0.01029107, + "balance_loss_clip": 1.04467177, + "balance_loss_mlp": 1.0214119, + "epoch": 0.6494318523417303, + "flos": 18802935072000.0, + "grad_norm": 1.6537624208969344, + "language_loss": 0.7538197, + "learning_rate": 1.1563718928690132e-06, + "loss": 0.77601862, + "num_input_tokens_seen": 116243860, + "step": 5401, + "time_per_iteration": 2.642829418182373 + }, + { + "auxiliary_loss_clip": 0.0119889, + "auxiliary_loss_mlp": 0.0103056, + "balance_loss_clip": 1.04722881, + "balance_loss_mlp": 1.023139, + "epoch": 0.6495520952323693, + "flos": 18982847318400.0, + "grad_norm": 2.8853240190959544, + "language_loss": 0.7100395, + "learning_rate": 1.1556656770231942e-06, + "loss": 0.73233396, + "num_input_tokens_seen": 116260055, + "step": 5402, + "time_per_iteration": 2.653926372528076 + }, + { + "auxiliary_loss_clip": 0.01179173, + "auxiliary_loss_mlp": 0.01026797, + "balance_loss_clip": 1.04575288, + "balance_loss_mlp": 1.01976943, + "epoch": 0.6496723381230085, + "flos": 22745388032640.0, + "grad_norm": 1.5425201019754873, + "language_loss": 0.75970042, + "learning_rate": 1.1549595892622207e-06, + "loss": 0.7817601, + "num_input_tokens_seen": 116278825, + "step": 5403, + "time_per_iteration": 2.6561717987060547 + }, + { + "auxiliary_loss_clip": 0.01113622, + "auxiliary_loss_mlp": 0.01000861, + "balance_loss_clip": 1.01474857, + "balance_loss_mlp": 0.99956763, + "epoch": 0.6497925810136476, + "flos": 62145283887360.0, + "grad_norm": 0.8214946898225934, + "language_loss": 0.5904336, + "learning_rate": 1.1542536296932047e-06, + "loss": 0.61157846, + "num_input_tokens_seen": 116342360, + "step": 5404, + "time_per_iteration": 3.2172324657440186 + }, + { + "auxiliary_loss_clip": 0.01204866, + "auxiliary_loss_mlp": 0.01026628, + "balance_loss_clip": 1.04466653, + "balance_loss_mlp": 1.01794958, + "epoch": 0.6499128239042866, + "flos": 20156731695360.0, + "grad_norm": 1.7268927948806507, + "language_loss": 0.70197845, + "learning_rate": 1.1535477984232414e-06, + "loss": 0.72429335, + "num_input_tokens_seen": 116362235, + "step": 5405, + "time_per_iteration": 2.7211709022521973 + }, + { + "auxiliary_loss_clip": 0.01208124, + "auxiliary_loss_mlp": 0.01029553, + "balance_loss_clip": 1.04086661, + "balance_loss_mlp": 1.02219141, + "epoch": 0.6500330667949258, + "flos": 24462420940800.0, + "grad_norm": 1.971768965948585, + "language_loss": 0.76447004, + "learning_rate": 1.152842095559404e-06, + "loss": 0.78684682, + "num_input_tokens_seen": 116382895, + "step": 5406, + "time_per_iteration": 2.7210352420806885 + }, + { + "auxiliary_loss_clip": 0.01194412, + "auxiliary_loss_mlp": 0.01020198, + "balance_loss_clip": 1.04611206, + "balance_loss_mlp": 1.01327491, + "epoch": 0.6501533096855648, + "flos": 25477399549440.0, + "grad_norm": 1.6527032920024687, + "language_loss": 0.76815099, + "learning_rate": 1.1521365212087474e-06, + "loss": 0.79029709, + "num_input_tokens_seen": 116402880, + "step": 5407, + "time_per_iteration": 2.730943202972412 + }, + { + "auxiliary_loss_clip": 0.01182627, + "auxiliary_loss_mlp": 0.01025249, + "balance_loss_clip": 1.04743695, + "balance_loss_mlp": 1.01711345, + "epoch": 0.6502735525762039, + "flos": 44819245347840.0, + "grad_norm": 1.6608805327096103, + "language_loss": 0.70600367, + "learning_rate": 1.1514310754783062e-06, + "loss": 0.72808242, + "num_input_tokens_seen": 116425830, + "step": 5408, + "time_per_iteration": 2.895686388015747 + }, + { + "auxiliary_loss_clip": 0.01193855, + "auxiliary_loss_mlp": 0.01028343, + "balance_loss_clip": 1.04800737, + "balance_loss_mlp": 1.02096117, + "epoch": 0.6503937954668431, + "flos": 28658546726400.0, + "grad_norm": 2.249244046018649, + "language_loss": 0.73161781, + "learning_rate": 1.1507257584750964e-06, + "loss": 0.75383973, + "num_input_tokens_seen": 116446010, + "step": 5409, + "time_per_iteration": 2.721442937850952 + }, + { + "auxiliary_loss_clip": 0.01174873, + "auxiliary_loss_mlp": 0.01027295, + "balance_loss_clip": 1.0521996, + "balance_loss_mlp": 1.01958179, + "epoch": 0.6505140383574821, + "flos": 20922562592640.0, + "grad_norm": 1.8314584761578063, + "language_loss": 0.77295184, + "learning_rate": 1.150020570306113e-06, + "loss": 0.79497349, + "num_input_tokens_seen": 116465150, + "step": 5410, + "time_per_iteration": 2.576352834701538 + }, + { + "auxiliary_loss_clip": 0.01179875, + "auxiliary_loss_mlp": 0.01031319, + "balance_loss_clip": 1.04209447, + "balance_loss_mlp": 1.02327847, + "epoch": 0.6506342812481212, + "flos": 20595236929920.0, + "grad_norm": 1.8165308130747249, + "language_loss": 0.74878114, + "learning_rate": 1.1493155110783338e-06, + "loss": 0.7708931, + "num_input_tokens_seen": 116483675, + "step": 5411, + "time_per_iteration": 2.6208078861236572 + }, + { + "auxiliary_loss_clip": 0.01182637, + "auxiliary_loss_mlp": 0.01027201, + "balance_loss_clip": 1.04964876, + "balance_loss_mlp": 1.01961291, + "epoch": 0.6507545241387603, + "flos": 30226478279040.0, + "grad_norm": 1.7994654287467176, + "language_loss": 0.70507681, + "learning_rate": 1.1486105808987155e-06, + "loss": 0.72717512, + "num_input_tokens_seen": 116505165, + "step": 5412, + "time_per_iteration": 2.710911750793457 + }, + { + "auxiliary_loss_clip": 0.01185535, + "auxiliary_loss_mlp": 0.01023654, + "balance_loss_clip": 1.05076313, + "balance_loss_mlp": 1.01547015, + "epoch": 0.6508747670293994, + "flos": 17128241320320.0, + "grad_norm": 2.0252154480151403, + "language_loss": 0.81195951, + "learning_rate": 1.1479057798741947e-06, + "loss": 0.83405143, + "num_input_tokens_seen": 116523220, + "step": 5413, + "time_per_iteration": 2.596022367477417 + }, + { + "auxiliary_loss_clip": 0.01089891, + "auxiliary_loss_mlp": 0.01002956, + "balance_loss_clip": 1.01600957, + "balance_loss_mlp": 1.00190735, + "epoch": 0.6509950099200384, + "flos": 68559826573440.0, + "grad_norm": 0.7901042256692048, + "language_loss": 0.5331881, + "learning_rate": 1.14720110811169e-06, + "loss": 0.55411661, + "num_input_tokens_seen": 116580450, + "step": 5414, + "time_per_iteration": 3.253159999847412 + }, + { + "auxiliary_loss_clip": 0.01187067, + "auxiliary_loss_mlp": 0.01025498, + "balance_loss_clip": 1.05158293, + "balance_loss_mlp": 1.0174036, + "epoch": 0.6511152528106776, + "flos": 22347462188160.0, + "grad_norm": 2.366500921497808, + "language_loss": 0.76484263, + "learning_rate": 1.146496565718098e-06, + "loss": 0.78696823, + "num_input_tokens_seen": 116601020, + "step": 5415, + "time_per_iteration": 3.5411922931671143 + }, + { + "auxiliary_loss_clip": 0.01189665, + "auxiliary_loss_mlp": 0.01018896, + "balance_loss_clip": 1.04787433, + "balance_loss_mlp": 1.01165438, + "epoch": 0.6512354957013167, + "flos": 20522158709760.0, + "grad_norm": 3.0409287680159958, + "language_loss": 0.75958955, + "learning_rate": 1.1457921528002996e-06, + "loss": 0.78167516, + "num_input_tokens_seen": 116619455, + "step": 5416, + "time_per_iteration": 2.643118381500244 + }, + { + "auxiliary_loss_clip": 0.01172102, + "auxiliary_loss_mlp": 0.01378725, + "balance_loss_clip": 1.04935384, + "balance_loss_mlp": 1.00016868, + "epoch": 0.6513557385919557, + "flos": 32337342881280.0, + "grad_norm": 2.507627764503028, + "language_loss": 0.72113788, + "learning_rate": 1.1450878694651522e-06, + "loss": 0.74664611, + "num_input_tokens_seen": 116640020, + "step": 5417, + "time_per_iteration": 3.5611321926116943 + }, + { + "auxiliary_loss_clip": 0.01208028, + "auxiliary_loss_mlp": 0.0102582, + "balance_loss_clip": 1.04224288, + "balance_loss_mlp": 1.01827383, + "epoch": 0.6514759814825949, + "flos": 12093206417280.0, + "grad_norm": 2.3965509206479942, + "language_loss": 0.63758707, + "learning_rate": 1.1443837158194954e-06, + "loss": 0.65992558, + "num_input_tokens_seen": 116655165, + "step": 5418, + "time_per_iteration": 2.6105048656463623 + }, + { + "auxiliary_loss_clip": 0.01200362, + "auxiliary_loss_mlp": 0.01022223, + "balance_loss_clip": 1.04859376, + "balance_loss_mlp": 1.01465333, + "epoch": 0.651596224373234, + "flos": 22526907557760.0, + "grad_norm": 1.6081219981507633, + "language_loss": 0.74242854, + "learning_rate": 1.1436796919701484e-06, + "loss": 0.7646544, + "num_input_tokens_seen": 116673880, + "step": 5419, + "time_per_iteration": 3.6956496238708496 + }, + { + "auxiliary_loss_clip": 0.0119021, + "auxiliary_loss_mlp": 0.01026959, + "balance_loss_clip": 1.04746473, + "balance_loss_mlp": 1.01932931, + "epoch": 0.651716467263873, + "flos": 27818955250560.0, + "grad_norm": 2.076957208918215, + "language_loss": 0.61947632, + "learning_rate": 1.1429757980239115e-06, + "loss": 0.64164799, + "num_input_tokens_seen": 116694305, + "step": 5420, + "time_per_iteration": 2.7299818992614746 + }, + { + "auxiliary_loss_clip": 0.01174049, + "auxiliary_loss_mlp": 0.01031183, + "balance_loss_clip": 1.04944897, + "balance_loss_mlp": 1.02268314, + "epoch": 0.6518367101545122, + "flos": 24316300414080.0, + "grad_norm": 2.3359042561971193, + "language_loss": 0.81254995, + "learning_rate": 1.1422720340875636e-06, + "loss": 0.83460224, + "num_input_tokens_seen": 116713055, + "step": 5421, + "time_per_iteration": 2.5948848724365234 + }, + { + "auxiliary_loss_clip": 0.01186024, + "auxiliary_loss_mlp": 0.01027234, + "balance_loss_clip": 1.04856539, + "balance_loss_mlp": 1.01970017, + "epoch": 0.6519569530451512, + "flos": 20011939971840.0, + "grad_norm": 3.4211058863963792, + "language_loss": 0.78994656, + "learning_rate": 1.1415684002678671e-06, + "loss": 0.81207919, + "num_input_tokens_seen": 116731815, + "step": 5422, + "time_per_iteration": 2.6277806758880615 + }, + { + "auxiliary_loss_clip": 0.01192445, + "auxiliary_loss_mlp": 0.01024483, + "balance_loss_clip": 1.04637861, + "balance_loss_mlp": 1.01632285, + "epoch": 0.6520771959357903, + "flos": 21576064682880.0, + "grad_norm": 2.3761010479037403, + "language_loss": 0.77662086, + "learning_rate": 1.1408648966715617e-06, + "loss": 0.7987901, + "num_input_tokens_seen": 116749335, + "step": 5423, + "time_per_iteration": 3.566775321960449 + }, + { + "auxiliary_loss_clip": 0.01188972, + "auxiliary_loss_mlp": 0.01029984, + "balance_loss_clip": 1.04381573, + "balance_loss_mlp": 1.02180636, + "epoch": 0.6521974388264293, + "flos": 22711021695360.0, + "grad_norm": 1.8089787966110364, + "language_loss": 0.72496557, + "learning_rate": 1.1401615234053683e-06, + "loss": 0.74715513, + "num_input_tokens_seen": 116768155, + "step": 5424, + "time_per_iteration": 2.698035478591919 + }, + { + "auxiliary_loss_clip": 0.0119054, + "auxiliary_loss_mlp": 0.01029816, + "balance_loss_clip": 1.04544044, + "balance_loss_mlp": 1.02191281, + "epoch": 0.6523176817170685, + "flos": 23002939526400.0, + "grad_norm": 2.02555910557439, + "language_loss": 0.75496566, + "learning_rate": 1.1394582805759885e-06, + "loss": 0.77716923, + "num_input_tokens_seen": 116787435, + "step": 5425, + "time_per_iteration": 2.633366584777832 + }, + { + "auxiliary_loss_clip": 0.01179099, + "auxiliary_loss_mlp": 0.01023786, + "balance_loss_clip": 1.0480783, + "balance_loss_mlp": 1.0153873, + "epoch": 0.6524379246077076, + "flos": 21688249835520.0, + "grad_norm": 1.9460779457336144, + "language_loss": 0.75835812, + "learning_rate": 1.1387551682901022e-06, + "loss": 0.78038698, + "num_input_tokens_seen": 116808040, + "step": 5426, + "time_per_iteration": 2.636065721511841 + }, + { + "auxiliary_loss_clip": 0.01194832, + "auxiliary_loss_mlp": 0.01030508, + "balance_loss_clip": 1.04468632, + "balance_loss_mlp": 1.02247906, + "epoch": 0.6525581674983466, + "flos": 19390936711680.0, + "grad_norm": 2.004017003958974, + "language_loss": 0.70593548, + "learning_rate": 1.138052186654373e-06, + "loss": 0.72818887, + "num_input_tokens_seen": 116825510, + "step": 5427, + "time_per_iteration": 2.6560378074645996 + }, + { + "auxiliary_loss_clip": 0.01190725, + "auxiliary_loss_mlp": 0.01033343, + "balance_loss_clip": 1.04627967, + "balance_loss_mlp": 1.02551103, + "epoch": 0.6526784103889858, + "flos": 17165444832000.0, + "grad_norm": 2.5049162624030843, + "language_loss": 0.88022763, + "learning_rate": 1.1373493357754417e-06, + "loss": 0.90246832, + "num_input_tokens_seen": 116844415, + "step": 5428, + "time_per_iteration": 2.6660566329956055 + }, + { + "auxiliary_loss_clip": 0.01170368, + "auxiliary_loss_mlp": 0.0102535, + "balance_loss_clip": 1.04666352, + "balance_loss_mlp": 1.01792586, + "epoch": 0.6527986532796248, + "flos": 18989168112000.0, + "grad_norm": 2.0871784779588483, + "language_loss": 0.77509093, + "learning_rate": 1.1366466157599303e-06, + "loss": 0.79704821, + "num_input_tokens_seen": 116863690, + "step": 5429, + "time_per_iteration": 2.5707809925079346 + }, + { + "auxiliary_loss_clip": 0.01203162, + "auxiliary_loss_mlp": 0.01378875, + "balance_loss_clip": 1.0421766, + "balance_loss_mlp": 1.00024629, + "epoch": 0.6529188961702639, + "flos": 14238581011200.0, + "grad_norm": 2.5154091109959738, + "language_loss": 0.75939143, + "learning_rate": 1.1359440267144412e-06, + "loss": 0.7852118, + "num_input_tokens_seen": 116881145, + "step": 5430, + "time_per_iteration": 2.8525893688201904 + }, + { + "auxiliary_loss_clip": 0.01184516, + "auxiliary_loss_mlp": 0.0102959, + "balance_loss_clip": 1.04949141, + "balance_loss_mlp": 1.02191842, + "epoch": 0.653039139060903, + "flos": 36682929158400.0, + "grad_norm": 2.626020123659012, + "language_loss": 0.74177217, + "learning_rate": 1.1352415687455556e-06, + "loss": 0.76391327, + "num_input_tokens_seen": 116902405, + "step": 5431, + "time_per_iteration": 2.8218746185302734 + }, + { + "auxiliary_loss_clip": 0.01183187, + "auxiliary_loss_mlp": 0.01026734, + "balance_loss_clip": 1.04952693, + "balance_loss_mlp": 1.01862788, + "epoch": 0.6531593819515421, + "flos": 25376275785600.0, + "grad_norm": 2.4762394340446807, + "language_loss": 0.63969755, + "learning_rate": 1.1345392419598362e-06, + "loss": 0.66179681, + "num_input_tokens_seen": 116921285, + "step": 5432, + "time_per_iteration": 2.676400661468506 + }, + { + "auxiliary_loss_clip": 0.01174112, + "auxiliary_loss_mlp": 0.0103146, + "balance_loss_clip": 1.04602814, + "balance_loss_mlp": 1.02373552, + "epoch": 0.6532796248421812, + "flos": 21178533888000.0, + "grad_norm": 1.842889480426148, + "language_loss": 0.72366321, + "learning_rate": 1.1338370464638263e-06, + "loss": 0.74571896, + "num_input_tokens_seen": 116940685, + "step": 5433, + "time_per_iteration": 2.6400399208068848 + }, + { + "auxiliary_loss_clip": 0.01173506, + "auxiliary_loss_mlp": 0.01023494, + "balance_loss_clip": 1.048944, + "balance_loss_mlp": 1.01606178, + "epoch": 0.6533998677328203, + "flos": 17675950878720.0, + "grad_norm": 2.649405537559041, + "language_loss": 0.64052808, + "learning_rate": 1.1331349823640474e-06, + "loss": 0.66249806, + "num_input_tokens_seen": 116958115, + "step": 5434, + "time_per_iteration": 2.6273081302642822 + }, + { + "auxiliary_loss_clip": 0.01180154, + "auxiliary_loss_mlp": 0.01378383, + "balance_loss_clip": 1.04729056, + "balance_loss_mlp": 1.0002228, + "epoch": 0.6535201106234594, + "flos": 28400384701440.0, + "grad_norm": 2.7012870356305116, + "language_loss": 0.77971828, + "learning_rate": 1.132433049767003e-06, + "loss": 0.80530369, + "num_input_tokens_seen": 116976030, + "step": 5435, + "time_per_iteration": 2.726398468017578 + }, + { + "auxiliary_loss_clip": 0.01189538, + "auxiliary_loss_mlp": 0.01019054, + "balance_loss_clip": 1.04789758, + "balance_loss_mlp": 1.01196671, + "epoch": 0.6536403535140984, + "flos": 23586667447680.0, + "grad_norm": 1.5269881874507516, + "language_loss": 0.80997062, + "learning_rate": 1.1317312487791748e-06, + "loss": 0.83205652, + "num_input_tokens_seen": 116997680, + "step": 5436, + "time_per_iteration": 2.7141425609588623 + }, + { + "auxiliary_loss_clip": 0.01174041, + "auxiliary_loss_mlp": 0.01025006, + "balance_loss_clip": 1.04602051, + "balance_loss_mlp": 1.01729298, + "epoch": 0.6537605964047376, + "flos": 21579476474880.0, + "grad_norm": 9.326610582061123, + "language_loss": 0.73362225, + "learning_rate": 1.1310295795070253e-06, + "loss": 0.75561267, + "num_input_tokens_seen": 117017620, + "step": 5437, + "time_per_iteration": 2.7387421131134033 + }, + { + "auxiliary_loss_clip": 0.01211123, + "auxiliary_loss_mlp": 0.01022164, + "balance_loss_clip": 1.04365253, + "balance_loss_mlp": 1.01448727, + "epoch": 0.6538808392953767, + "flos": 26833997433600.0, + "grad_norm": 1.842631738040884, + "language_loss": 0.81084371, + "learning_rate": 1.1303280420569982e-06, + "loss": 0.83317661, + "num_input_tokens_seen": 117039505, + "step": 5438, + "time_per_iteration": 2.782607316970825 + }, + { + "auxiliary_loss_clip": 0.01174339, + "auxiliary_loss_mlp": 0.01028309, + "balance_loss_clip": 1.0463326, + "balance_loss_mlp": 1.02078652, + "epoch": 0.6540010821860157, + "flos": 30738241301760.0, + "grad_norm": 1.6811922804630992, + "language_loss": 0.77331412, + "learning_rate": 1.1296266365355158e-06, + "loss": 0.7953406, + "num_input_tokens_seen": 117062890, + "step": 5439, + "time_per_iteration": 2.702214002609253 + }, + { + "auxiliary_loss_clip": 0.01204619, + "auxiliary_loss_mlp": 0.0102748, + "balance_loss_clip": 1.04721451, + "balance_loss_mlp": 1.01958883, + "epoch": 0.6541213250766549, + "flos": 26907147480960.0, + "grad_norm": 1.96699760559037, + "language_loss": 0.73761636, + "learning_rate": 1.1289253630489806e-06, + "loss": 0.75993735, + "num_input_tokens_seen": 117083940, + "step": 5440, + "time_per_iteration": 2.681300401687622 + }, + { + "auxiliary_loss_clip": 0.01186841, + "auxiliary_loss_mlp": 0.01029288, + "balance_loss_clip": 1.04841924, + "balance_loss_mlp": 1.02105045, + "epoch": 0.6542415679672939, + "flos": 19172384409600.0, + "grad_norm": 2.0374017291045985, + "language_loss": 0.7207725, + "learning_rate": 1.1282242217037753e-06, + "loss": 0.74293375, + "num_input_tokens_seen": 117101440, + "step": 5441, + "time_per_iteration": 3.552530288696289 + }, + { + "auxiliary_loss_clip": 0.01197638, + "auxiliary_loss_mlp": 0.0103101, + "balance_loss_clip": 1.03814244, + "balance_loss_mlp": 1.02314258, + "epoch": 0.654361810857933, + "flos": 48173517100800.0, + "grad_norm": 1.93215412017445, + "language_loss": 0.62318116, + "learning_rate": 1.127523212606262e-06, + "loss": 0.64546764, + "num_input_tokens_seen": 117124265, + "step": 5442, + "time_per_iteration": 2.9870705604553223 + }, + { + "auxiliary_loss_clip": 0.0117817, + "auxiliary_loss_mlp": 0.01027364, + "balance_loss_clip": 1.04633737, + "balance_loss_mlp": 1.01948476, + "epoch": 0.6544820537485722, + "flos": 26943165843840.0, + "grad_norm": 1.5784819513982715, + "language_loss": 0.72948968, + "learning_rate": 1.1268223358627835e-06, + "loss": 0.75154495, + "num_input_tokens_seen": 117146755, + "step": 5443, + "time_per_iteration": 3.494569778442383 + }, + { + "auxiliary_loss_clip": 0.01173204, + "auxiliary_loss_mlp": 0.01026564, + "balance_loss_clip": 1.04939187, + "balance_loss_mlp": 1.0188539, + "epoch": 0.6546022966392112, + "flos": 20886328748160.0, + "grad_norm": 1.8507597334800001, + "language_loss": 0.72148281, + "learning_rate": 1.126121591579663e-06, + "loss": 0.74348044, + "num_input_tokens_seen": 117165960, + "step": 5444, + "time_per_iteration": 3.5095598697662354 + }, + { + "auxiliary_loss_clip": 0.01178697, + "auxiliary_loss_mlp": 0.01025331, + "balance_loss_clip": 1.04878497, + "balance_loss_mlp": 1.01789236, + "epoch": 0.6547225395298503, + "flos": 24936693143040.0, + "grad_norm": 1.6564998311811796, + "language_loss": 0.69004714, + "learning_rate": 1.1254209798632018e-06, + "loss": 0.71208739, + "num_input_tokens_seen": 117186980, + "step": 5445, + "time_per_iteration": 2.6961288452148438 + }, + { + "auxiliary_loss_clip": 0.01201744, + "auxiliary_loss_mlp": 0.01023194, + "balance_loss_clip": 1.03938961, + "balance_loss_mlp": 1.01553512, + "epoch": 0.6548427824204894, + "flos": 22565942663040.0, + "grad_norm": 1.7076456875356212, + "language_loss": 0.84690517, + "learning_rate": 1.124720500819683e-06, + "loss": 0.86915457, + "num_input_tokens_seen": 117205135, + "step": 5446, + "time_per_iteration": 2.717576026916504 + }, + { + "auxiliary_loss_clip": 0.01173222, + "auxiliary_loss_mlp": 0.01026352, + "balance_loss_clip": 1.04935455, + "balance_loss_mlp": 1.01853168, + "epoch": 0.6549630253111285, + "flos": 18442500048000.0, + "grad_norm": 2.2110817502115427, + "language_loss": 0.82298803, + "learning_rate": 1.1240201545553682e-06, + "loss": 0.84498382, + "num_input_tokens_seen": 117222935, + "step": 5447, + "time_per_iteration": 2.5877528190612793 + }, + { + "auxiliary_loss_clip": 0.01198386, + "auxiliary_loss_mlp": 0.01026834, + "balance_loss_clip": 1.04627335, + "balance_loss_mlp": 1.01880193, + "epoch": 0.6550832682017675, + "flos": 25187313312000.0, + "grad_norm": 1.9260976311656293, + "language_loss": 0.73342371, + "learning_rate": 1.1233199411764987e-06, + "loss": 0.75567585, + "num_input_tokens_seen": 117242370, + "step": 5448, + "time_per_iteration": 2.734208583831787 + }, + { + "auxiliary_loss_clip": 0.01184155, + "auxiliary_loss_mlp": 0.01028584, + "balance_loss_clip": 1.04193103, + "balance_loss_mlp": 1.02072179, + "epoch": 0.6552035110924067, + "flos": 22748153379840.0, + "grad_norm": 1.7176198334647896, + "language_loss": 0.68905151, + "learning_rate": 1.1226198607892978e-06, + "loss": 0.7111789, + "num_input_tokens_seen": 117262930, + "step": 5449, + "time_per_iteration": 3.598304033279419 + }, + { + "auxiliary_loss_clip": 0.01211709, + "auxiliary_loss_mlp": 0.01031116, + "balance_loss_clip": 1.04703844, + "balance_loss_mlp": 1.02344513, + "epoch": 0.6553237539830458, + "flos": 21799178012160.0, + "grad_norm": 3.2455517099084292, + "language_loss": 0.80019492, + "learning_rate": 1.1219199134999664e-06, + "loss": 0.82262313, + "num_input_tokens_seen": 117281430, + "step": 5450, + "time_per_iteration": 2.8103978633880615 + }, + { + "auxiliary_loss_clip": 0.01192153, + "auxiliary_loss_mlp": 0.01027373, + "balance_loss_clip": 1.04713202, + "balance_loss_mlp": 1.01933217, + "epoch": 0.6554439968736848, + "flos": 20887226588160.0, + "grad_norm": 2.089938460230114, + "language_loss": 0.7849021, + "learning_rate": 1.1212200994146863e-06, + "loss": 0.80709732, + "num_input_tokens_seen": 117299185, + "step": 5451, + "time_per_iteration": 2.656121253967285 + }, + { + "auxiliary_loss_clip": 0.01192771, + "auxiliary_loss_mlp": 0.01031131, + "balance_loss_clip": 1.03899562, + "balance_loss_mlp": 1.02332234, + "epoch": 0.655564239764324, + "flos": 16139045698560.0, + "grad_norm": 1.9389421928802228, + "language_loss": 0.75747788, + "learning_rate": 1.120520418639618e-06, + "loss": 0.77971691, + "num_input_tokens_seen": 117317720, + "step": 5452, + "time_per_iteration": 2.7524914741516113 + }, + { + "auxiliary_loss_clip": 0.01183699, + "auxiliary_loss_mlp": 0.01027776, + "balance_loss_clip": 1.05047321, + "balance_loss_mlp": 1.02022409, + "epoch": 0.655684482654963, + "flos": 29570354496000.0, + "grad_norm": 6.214552861638131, + "language_loss": 0.83272016, + "learning_rate": 1.119820871280903e-06, + "loss": 0.85483491, + "num_input_tokens_seen": 117338795, + "step": 5453, + "time_per_iteration": 2.6760146617889404 + }, + { + "auxiliary_loss_clip": 0.01180355, + "auxiliary_loss_mlp": 0.01028438, + "balance_loss_clip": 1.04832768, + "balance_loss_mlp": 1.02025437, + "epoch": 0.6558047255456021, + "flos": 29789409588480.0, + "grad_norm": 2.6481168548383685, + "language_loss": 0.73901677, + "learning_rate": 1.1191214574446614e-06, + "loss": 0.7611047, + "num_input_tokens_seen": 117359040, + "step": 5454, + "time_per_iteration": 2.686100721359253 + }, + { + "auxiliary_loss_clip": 0.01183202, + "auxiliary_loss_mlp": 0.01026634, + "balance_loss_clip": 1.04527545, + "balance_loss_mlp": 1.0187782, + "epoch": 0.6559249684362413, + "flos": 29059166090880.0, + "grad_norm": 1.5455310842617695, + "language_loss": 0.8009994, + "learning_rate": 1.118422177236995e-06, + "loss": 0.82309777, + "num_input_tokens_seen": 117380865, + "step": 5455, + "time_per_iteration": 2.741837739944458 + }, + { + "auxiliary_loss_clip": 0.0119116, + "auxiliary_loss_mlp": 0.01031436, + "balance_loss_clip": 1.04665029, + "balance_loss_mlp": 1.02366316, + "epoch": 0.6560452113268803, + "flos": 20225464369920.0, + "grad_norm": 1.952781535387301, + "language_loss": 0.85610479, + "learning_rate": 1.1177230307639835e-06, + "loss": 0.87833071, + "num_input_tokens_seen": 117398405, + "step": 5456, + "time_per_iteration": 2.685612678527832 + }, + { + "auxiliary_loss_clip": 0.01196129, + "auxiliary_loss_mlp": 0.01022151, + "balance_loss_clip": 1.04490924, + "balance_loss_mlp": 1.01471543, + "epoch": 0.6561654542175194, + "flos": 25045538330880.0, + "grad_norm": 1.6783494037134383, + "language_loss": 0.78837311, + "learning_rate": 1.1170240181316865e-06, + "loss": 0.81055593, + "num_input_tokens_seen": 117419850, + "step": 5457, + "time_per_iteration": 2.6994621753692627 + }, + { + "auxiliary_loss_clip": 0.01190767, + "auxiliary_loss_mlp": 0.01025204, + "balance_loss_clip": 1.04143119, + "balance_loss_mlp": 1.01765788, + "epoch": 0.6562856971081584, + "flos": 22856711258880.0, + "grad_norm": 2.130496461231111, + "language_loss": 0.79261708, + "learning_rate": 1.1163251394461442e-06, + "loss": 0.81477684, + "num_input_tokens_seen": 117438330, + "step": 5458, + "time_per_iteration": 2.791262149810791 + }, + { + "auxiliary_loss_clip": 0.01177004, + "auxiliary_loss_mlp": 0.0102428, + "balance_loss_clip": 1.04644394, + "balance_loss_mlp": 1.0167942, + "epoch": 0.6564059399987976, + "flos": 18872565586560.0, + "grad_norm": 2.1052055773073977, + "language_loss": 0.82161522, + "learning_rate": 1.1156263948133746e-06, + "loss": 0.84362805, + "num_input_tokens_seen": 117454985, + "step": 5459, + "time_per_iteration": 2.6147706508636475 + }, + { + "auxiliary_loss_clip": 0.01196216, + "auxiliary_loss_mlp": 0.01378995, + "balance_loss_clip": 1.04268622, + "balance_loss_mlp": 1.00023341, + "epoch": 0.6565261828894366, + "flos": 25484187219840.0, + "grad_norm": 1.7275216015008015, + "language_loss": 0.7758832, + "learning_rate": 1.1149277843393787e-06, + "loss": 0.80163532, + "num_input_tokens_seen": 117476145, + "step": 5460, + "time_per_iteration": 2.7401113510131836 + }, + { + "auxiliary_loss_clip": 0.01207593, + "auxiliary_loss_mlp": 0.01378774, + "balance_loss_clip": 1.03780222, + "balance_loss_mlp": 1.00020146, + "epoch": 0.6566464257800757, + "flos": 19683500987520.0, + "grad_norm": 2.552209470122296, + "language_loss": 0.63417125, + "learning_rate": 1.1142293081301342e-06, + "loss": 0.66003489, + "num_input_tokens_seen": 117494025, + "step": 5461, + "time_per_iteration": 2.8092641830444336 + }, + { + "auxiliary_loss_clip": 0.01187376, + "auxiliary_loss_mlp": 0.010211, + "balance_loss_clip": 1.04766417, + "balance_loss_mlp": 1.0138998, + "epoch": 0.6567666686707149, + "flos": 23514127931520.0, + "grad_norm": 1.6012606276308419, + "language_loss": 0.67834008, + "learning_rate": 1.1135309662915995e-06, + "loss": 0.70042479, + "num_input_tokens_seen": 117514190, + "step": 5462, + "time_per_iteration": 2.761206865310669 + }, + { + "auxiliary_loss_clip": 0.01211734, + "auxiliary_loss_mlp": 0.01021311, + "balance_loss_clip": 1.04113817, + "balance_loss_mlp": 1.01340747, + "epoch": 0.6568869115613539, + "flos": 32781342896640.0, + "grad_norm": 2.2635145912768198, + "language_loss": 0.60597241, + "learning_rate": 1.112832758929712e-06, + "loss": 0.62830287, + "num_input_tokens_seen": 117536800, + "step": 5463, + "time_per_iteration": 2.8574702739715576 + }, + { + "auxiliary_loss_clip": 0.01179878, + "auxiliary_loss_mlp": 0.010293, + "balance_loss_clip": 1.04911637, + "balance_loss_mlp": 1.02127147, + "epoch": 0.657007154451993, + "flos": 18442428220800.0, + "grad_norm": 2.615514964998461, + "language_loss": 0.74846554, + "learning_rate": 1.11213468615039e-06, + "loss": 0.77055728, + "num_input_tokens_seen": 117556230, + "step": 5464, + "time_per_iteration": 2.649663209915161 + }, + { + "auxiliary_loss_clip": 0.01210751, + "auxiliary_loss_mlp": 0.01021146, + "balance_loss_clip": 1.03923893, + "balance_loss_mlp": 1.01360941, + "epoch": 0.6571273973426321, + "flos": 25156717902720.0, + "grad_norm": 1.590347505351862, + "language_loss": 0.75183594, + "learning_rate": 1.1114367480595292e-06, + "loss": 0.7741549, + "num_input_tokens_seen": 117577310, + "step": 5465, + "time_per_iteration": 2.8114919662475586 + }, + { + "auxiliary_loss_clip": 0.0121479, + "auxiliary_loss_mlp": 0.01026011, + "balance_loss_clip": 1.04221129, + "balance_loss_mlp": 1.0169332, + "epoch": 0.6572476402332712, + "flos": 17529830352000.0, + "grad_norm": 1.8825456053584273, + "language_loss": 0.81380725, + "learning_rate": 1.1107389447630086e-06, + "loss": 0.83621526, + "num_input_tokens_seen": 117596010, + "step": 5466, + "time_per_iteration": 2.7108876705169678 + }, + { + "auxiliary_loss_clip": 0.01182698, + "auxiliary_loss_mlp": 0.01378555, + "balance_loss_clip": 1.04321921, + "balance_loss_mlp": 1.00017047, + "epoch": 0.6573678831239103, + "flos": 17014260487680.0, + "grad_norm": 2.2551953764470514, + "language_loss": 0.78166765, + "learning_rate": 1.1100412763666818e-06, + "loss": 0.80728018, + "num_input_tokens_seen": 117611270, + "step": 5467, + "time_per_iteration": 3.683776378631592 + }, + { + "auxiliary_loss_clip": 0.0119304, + "auxiliary_loss_mlp": 0.01024009, + "balance_loss_clip": 1.04822516, + "balance_loss_mlp": 1.01668358, + "epoch": 0.6574881260145494, + "flos": 23910078528000.0, + "grad_norm": 2.401831570390094, + "language_loss": 0.80115378, + "learning_rate": 1.1093437429763865e-06, + "loss": 0.82332432, + "num_input_tokens_seen": 117631535, + "step": 5468, + "time_per_iteration": 3.530486822128296 + }, + { + "auxiliary_loss_clip": 0.01180594, + "auxiliary_loss_mlp": 0.0102663, + "balance_loss_clip": 1.04727757, + "balance_loss_mlp": 1.01942086, + "epoch": 0.6576083689051885, + "flos": 11218458504960.0, + "grad_norm": 1.9952258422789133, + "language_loss": 0.7359606, + "learning_rate": 1.1086463446979361e-06, + "loss": 0.7580328, + "num_input_tokens_seen": 117649885, + "step": 5469, + "time_per_iteration": 2.5653040409088135 + }, + { + "auxiliary_loss_clip": 0.01184533, + "auxiliary_loss_mlp": 0.01027554, + "balance_loss_clip": 1.05024886, + "balance_loss_mlp": 1.01921487, + "epoch": 0.6577286117958275, + "flos": 22455553190400.0, + "grad_norm": 1.9601870192669129, + "language_loss": 0.77582896, + "learning_rate": 1.1079490816371277e-06, + "loss": 0.79794985, + "num_input_tokens_seen": 117669650, + "step": 5470, + "time_per_iteration": 2.640389919281006 + }, + { + "auxiliary_loss_clip": 0.01181466, + "auxiliary_loss_mlp": 0.01379024, + "balance_loss_clip": 1.04759228, + "balance_loss_mlp": 1.0001657, + "epoch": 0.6578488546864667, + "flos": 21872184405120.0, + "grad_norm": 2.1486738187691374, + "language_loss": 0.74587703, + "learning_rate": 1.1072519538997352e-06, + "loss": 0.77148187, + "num_input_tokens_seen": 117688790, + "step": 5471, + "time_per_iteration": 3.56889009475708 + }, + { + "auxiliary_loss_clip": 0.01191577, + "auxiliary_loss_mlp": 0.01026862, + "balance_loss_clip": 1.04404497, + "balance_loss_mlp": 1.0190717, + "epoch": 0.6579690975771058, + "flos": 23543753673600.0, + "grad_norm": 1.7587335565438564, + "language_loss": 0.82282346, + "learning_rate": 1.1065549615915095e-06, + "loss": 0.84500784, + "num_input_tokens_seen": 117708620, + "step": 5472, + "time_per_iteration": 2.6895265579223633 + }, + { + "auxiliary_loss_clip": 0.01184183, + "auxiliary_loss_mlp": 0.01031134, + "balance_loss_clip": 1.05285609, + "balance_loss_mlp": 1.02306914, + "epoch": 0.6580893404677448, + "flos": 32743995730560.0, + "grad_norm": 3.0037491904279072, + "language_loss": 0.78420812, + "learning_rate": 1.105858104818187e-06, + "loss": 0.8063612, + "num_input_tokens_seen": 117729775, + "step": 5473, + "time_per_iteration": 2.7206971645355225 + }, + { + "auxiliary_loss_clip": 0.01185777, + "auxiliary_loss_mlp": 0.01026646, + "balance_loss_clip": 1.04851985, + "balance_loss_mlp": 1.01856971, + "epoch": 0.658209583358384, + "flos": 15888138220800.0, + "grad_norm": 2.43776280206725, + "language_loss": 0.7515322, + "learning_rate": 1.105161383685478e-06, + "loss": 0.77365649, + "num_input_tokens_seen": 117746160, + "step": 5474, + "time_per_iteration": 3.5353760719299316 + }, + { + "auxiliary_loss_clip": 0.01104643, + "auxiliary_loss_mlp": 0.0100203, + "balance_loss_clip": 1.01331365, + "balance_loss_mlp": 1.00074291, + "epoch": 0.658329826249023, + "flos": 62695902447360.0, + "grad_norm": 0.7271054230891241, + "language_loss": 0.5634498, + "learning_rate": 1.1044647982990771e-06, + "loss": 0.58451653, + "num_input_tokens_seen": 117808045, + "step": 5475, + "time_per_iteration": 3.2078685760498047 + }, + { + "auxiliary_loss_clip": 0.01192184, + "auxiliary_loss_mlp": 0.01027084, + "balance_loss_clip": 1.04739881, + "balance_loss_mlp": 1.01845312, + "epoch": 0.6584500691396621, + "flos": 31722624501120.0, + "grad_norm": 2.2937727402492767, + "language_loss": 0.64572388, + "learning_rate": 1.1037683487646536e-06, + "loss": 0.66791654, + "num_input_tokens_seen": 117828330, + "step": 5476, + "time_per_iteration": 2.7580296993255615 + }, + { + "auxiliary_loss_clip": 0.01187751, + "auxiliary_loss_mlp": 0.01378847, + "balance_loss_clip": 1.04782701, + "balance_loss_mlp": 1.00014412, + "epoch": 0.6585703120303013, + "flos": 18406086635520.0, + "grad_norm": 2.4701921502218913, + "language_loss": 0.76813495, + "learning_rate": 1.1030720351878583e-06, + "loss": 0.79380095, + "num_input_tokens_seen": 117846450, + "step": 5477, + "time_per_iteration": 2.6711387634277344 + }, + { + "auxiliary_loss_clip": 0.01092585, + "auxiliary_loss_mlp": 0.01005494, + "balance_loss_clip": 1.01134181, + "balance_loss_mlp": 1.00420082, + "epoch": 0.6586905549209403, + "flos": 58309880434560.0, + "grad_norm": 0.8056159742080125, + "language_loss": 0.57636547, + "learning_rate": 1.102375857674323e-06, + "loss": 0.59734619, + "num_input_tokens_seen": 117908365, + "step": 5478, + "time_per_iteration": 3.243305206298828 + }, + { + "auxiliary_loss_clip": 0.0118581, + "auxiliary_loss_mlp": 0.01029122, + "balance_loss_clip": 1.04329097, + "balance_loss_mlp": 1.02211869, + "epoch": 0.6588107978115794, + "flos": 22782627457920.0, + "grad_norm": 1.700295160186772, + "language_loss": 0.90396345, + "learning_rate": 1.1016798163296561e-06, + "loss": 0.92611271, + "num_input_tokens_seen": 117927565, + "step": 5479, + "time_per_iteration": 2.648953676223755 + }, + { + "auxiliary_loss_clip": 0.01184388, + "auxiliary_loss_mlp": 0.01024073, + "balance_loss_clip": 1.04810059, + "balance_loss_mlp": 1.01618135, + "epoch": 0.6589310407022185, + "flos": 20667525050880.0, + "grad_norm": 2.1205164775162437, + "language_loss": 0.66061902, + "learning_rate": 1.1009839112594471e-06, + "loss": 0.68270373, + "num_input_tokens_seen": 117945590, + "step": 5480, + "time_per_iteration": 2.6581740379333496 + }, + { + "auxiliary_loss_clip": 0.01182644, + "auxiliary_loss_mlp": 0.01028275, + "balance_loss_clip": 1.04722738, + "balance_loss_mlp": 1.02025533, + "epoch": 0.6590512835928576, + "flos": 25630595055360.0, + "grad_norm": 2.0886662215040848, + "language_loss": 0.71852285, + "learning_rate": 1.1002881425692638e-06, + "loss": 0.74063206, + "num_input_tokens_seen": 117966020, + "step": 5481, + "time_per_iteration": 2.709667921066284 + }, + { + "auxiliary_loss_clip": 0.01175321, + "auxiliary_loss_mlp": 0.0102572, + "balance_loss_clip": 1.046556, + "balance_loss_mlp": 1.01804316, + "epoch": 0.6591715264834966, + "flos": 23726108044800.0, + "grad_norm": 1.8724404460943624, + "language_loss": 0.75586152, + "learning_rate": 1.0995925103646532e-06, + "loss": 0.77787197, + "num_input_tokens_seen": 117984620, + "step": 5482, + "time_per_iteration": 2.6206629276275635 + }, + { + "auxiliary_loss_clip": 0.01192598, + "auxiliary_loss_mlp": 0.0102577, + "balance_loss_clip": 1.04505467, + "balance_loss_mlp": 1.01888573, + "epoch": 0.6592917693741358, + "flos": 35773850822400.0, + "grad_norm": 1.442011814989223, + "language_loss": 0.66408372, + "learning_rate": 1.0988970147511437e-06, + "loss": 0.68626738, + "num_input_tokens_seen": 118006500, + "step": 5483, + "time_per_iteration": 2.8196098804473877 + }, + { + "auxiliary_loss_clip": 0.01190761, + "auxiliary_loss_mlp": 0.01024521, + "balance_loss_clip": 1.04802513, + "balance_loss_mlp": 1.01629543, + "epoch": 0.6594120122647749, + "flos": 21396834794880.0, + "grad_norm": 3.3703502823146794, + "language_loss": 0.8051635, + "learning_rate": 1.0982016558342405e-06, + "loss": 0.82731628, + "num_input_tokens_seen": 118025470, + "step": 5484, + "time_per_iteration": 2.7161011695861816 + }, + { + "auxiliary_loss_clip": 0.01172505, + "auxiliary_loss_mlp": 0.01032277, + "balance_loss_clip": 1.04879904, + "balance_loss_mlp": 1.02464795, + "epoch": 0.6595322551554139, + "flos": 19351829779200.0, + "grad_norm": 1.8695344821352566, + "language_loss": 0.71016455, + "learning_rate": 1.0975064337194291e-06, + "loss": 0.73221231, + "num_input_tokens_seen": 118043515, + "step": 5485, + "time_per_iteration": 2.6657278537750244 + }, + { + "auxiliary_loss_clip": 0.01190653, + "auxiliary_loss_mlp": 0.01025685, + "balance_loss_clip": 1.04315507, + "balance_loss_mlp": 1.01800501, + "epoch": 0.6596524980460531, + "flos": 16837113588480.0, + "grad_norm": 1.4217932510798572, + "language_loss": 0.70712054, + "learning_rate": 1.0968113485121743e-06, + "loss": 0.72928393, + "num_input_tokens_seen": 118063105, + "step": 5486, + "time_per_iteration": 2.734748601913452 + }, + { + "auxiliary_loss_clip": 0.01180991, + "auxiliary_loss_mlp": 0.01378976, + "balance_loss_clip": 1.04582667, + "balance_loss_mlp": 1.00012863, + "epoch": 0.6597727409366921, + "flos": 21798567480960.0, + "grad_norm": 1.9066775982915378, + "language_loss": 0.80443132, + "learning_rate": 1.0961164003179185e-06, + "loss": 0.83003092, + "num_input_tokens_seen": 118081615, + "step": 5487, + "time_per_iteration": 2.670727014541626 + }, + { + "auxiliary_loss_clip": 0.01194173, + "auxiliary_loss_mlp": 0.01024714, + "balance_loss_clip": 1.04311645, + "balance_loss_mlp": 1.01726365, + "epoch": 0.6598929838273312, + "flos": 23730704985600.0, + "grad_norm": 2.045562061774251, + "language_loss": 0.84085035, + "learning_rate": 1.0954215892420884e-06, + "loss": 0.86303926, + "num_input_tokens_seen": 118102315, + "step": 5488, + "time_per_iteration": 2.7016162872314453 + }, + { + "auxiliary_loss_clip": 0.01202453, + "auxiliary_loss_mlp": 0.01031479, + "balance_loss_clip": 1.04730284, + "balance_loss_mlp": 1.02341461, + "epoch": 0.6600132267179702, + "flos": 19974520978560.0, + "grad_norm": 1.9846904853606142, + "language_loss": 0.70643085, + "learning_rate": 1.094726915390082e-06, + "loss": 0.7287702, + "num_input_tokens_seen": 118120650, + "step": 5489, + "time_per_iteration": 2.7126033306121826 + }, + { + "auxiliary_loss_clip": 0.01182459, + "auxiliary_loss_mlp": 0.0102825, + "balance_loss_clip": 1.04858303, + "balance_loss_mlp": 1.02025723, + "epoch": 0.6601334696086094, + "flos": 22342649765760.0, + "grad_norm": 1.920783632597329, + "language_loss": 0.69420683, + "learning_rate": 1.0940323788672836e-06, + "loss": 0.71631396, + "num_input_tokens_seen": 118139825, + "step": 5490, + "time_per_iteration": 2.5979909896850586 + }, + { + "auxiliary_loss_clip": 0.01180092, + "auxiliary_loss_mlp": 0.01025827, + "balance_loss_clip": 1.04919434, + "balance_loss_mlp": 1.01858532, + "epoch": 0.6602537124992485, + "flos": 25703098657920.0, + "grad_norm": 1.7269715592089296, + "language_loss": 0.73603153, + "learning_rate": 1.0933379797790522e-06, + "loss": 0.75809073, + "num_input_tokens_seen": 118159240, + "step": 5491, + "time_per_iteration": 2.708857774734497 + }, + { + "auxiliary_loss_clip": 0.01172917, + "auxiliary_loss_mlp": 0.01028961, + "balance_loss_clip": 1.04944968, + "balance_loss_mlp": 1.02116787, + "epoch": 0.6603739553898875, + "flos": 25848572739840.0, + "grad_norm": 2.4603300180926526, + "language_loss": 0.71697378, + "learning_rate": 1.0926437182307293e-06, + "loss": 0.73899257, + "num_input_tokens_seen": 118178050, + "step": 5492, + "time_per_iteration": 2.632115125656128 + }, + { + "auxiliary_loss_clip": 0.0119428, + "auxiliary_loss_mlp": 0.01023058, + "balance_loss_clip": 1.04500985, + "balance_loss_mlp": 1.01563692, + "epoch": 0.6604941982805267, + "flos": 24570296461440.0, + "grad_norm": 1.8155557963237636, + "language_loss": 0.77874815, + "learning_rate": 1.0919495943276338e-06, + "loss": 0.80092156, + "num_input_tokens_seen": 118199070, + "step": 5493, + "time_per_iteration": 3.686401128768921 + }, + { + "auxiliary_loss_clip": 0.01196792, + "auxiliary_loss_mlp": 0.01033129, + "balance_loss_clip": 1.04094827, + "balance_loss_mlp": 1.0252552, + "epoch": 0.6606144411711657, + "flos": 13261775581440.0, + "grad_norm": 4.42978491386882, + "language_loss": 0.76267219, + "learning_rate": 1.0912556081750611e-06, + "loss": 0.78497136, + "num_input_tokens_seen": 118217000, + "step": 5494, + "time_per_iteration": 3.6307425498962402 + }, + { + "auxiliary_loss_clip": 0.01185992, + "auxiliary_loss_mlp": 0.01029571, + "balance_loss_clip": 1.04605913, + "balance_loss_mlp": 1.02178049, + "epoch": 0.6607346840618048, + "flos": 25155281358720.0, + "grad_norm": 2.0383894790585115, + "language_loss": 0.76409835, + "learning_rate": 1.0905617598782909e-06, + "loss": 0.78625393, + "num_input_tokens_seen": 118237205, + "step": 5495, + "time_per_iteration": 2.6771562099456787 + }, + { + "auxiliary_loss_clip": 0.01198245, + "auxiliary_loss_mlp": 0.01027407, + "balance_loss_clip": 1.04342628, + "balance_loss_mlp": 1.01992631, + "epoch": 0.660854926952444, + "flos": 17638029095040.0, + "grad_norm": 2.508484579529935, + "language_loss": 0.81688905, + "learning_rate": 1.0898680495425775e-06, + "loss": 0.8391456, + "num_input_tokens_seen": 118255495, + "step": 5496, + "time_per_iteration": 2.8771986961364746 + }, + { + "auxiliary_loss_clip": 0.01192437, + "auxiliary_loss_mlp": 0.01024005, + "balance_loss_clip": 1.04711938, + "balance_loss_mlp": 1.01641178, + "epoch": 0.660975169843083, + "flos": 16836000266880.0, + "grad_norm": 1.7399581429238427, + "language_loss": 0.80351788, + "learning_rate": 1.0891744772731594e-06, + "loss": 0.82568228, + "num_input_tokens_seen": 118273310, + "step": 5497, + "time_per_iteration": 3.495831251144409 + }, + { + "auxiliary_loss_clip": 0.01181068, + "auxiliary_loss_mlp": 0.01024535, + "balance_loss_clip": 1.04609764, + "balance_loss_mlp": 1.01750124, + "epoch": 0.6610954127337221, + "flos": 26870410846080.0, + "grad_norm": 1.6241064194233474, + "language_loss": 0.65761685, + "learning_rate": 1.088481043175248e-06, + "loss": 0.67967284, + "num_input_tokens_seen": 118293880, + "step": 5498, + "time_per_iteration": 2.660104513168335 + }, + { + "auxiliary_loss_clip": 0.01178212, + "auxiliary_loss_mlp": 0.0102536, + "balance_loss_clip": 1.0445118, + "balance_loss_mlp": 1.01815355, + "epoch": 0.6612156556243612, + "flos": 26465697331200.0, + "grad_norm": 2.1783534616864086, + "language_loss": 0.75496113, + "learning_rate": 1.0877877473540368e-06, + "loss": 0.77699685, + "num_input_tokens_seen": 118314465, + "step": 5499, + "time_per_iteration": 2.74314546585083 + }, + { + "auxiliary_loss_clip": 0.01173889, + "auxiliary_loss_mlp": 0.01026653, + "balance_loss_clip": 1.04964352, + "balance_loss_mlp": 1.01941109, + "epoch": 0.6613358985150003, + "flos": 19791915212160.0, + "grad_norm": 1.8087442917014325, + "language_loss": 0.72499287, + "learning_rate": 1.0870945899147002e-06, + "loss": 0.74699831, + "num_input_tokens_seen": 118331110, + "step": 5500, + "time_per_iteration": 3.4458203315734863 + }, + { + "auxiliary_loss_clip": 0.01181293, + "auxiliary_loss_mlp": 0.01027968, + "balance_loss_clip": 1.04934311, + "balance_loss_mlp": 1.02088404, + "epoch": 0.6614561414056394, + "flos": 26831627136000.0, + "grad_norm": 1.8086673361479098, + "language_loss": 0.76335663, + "learning_rate": 1.0864015709623879e-06, + "loss": 0.78544921, + "num_input_tokens_seen": 118351980, + "step": 5501, + "time_per_iteration": 2.7052218914031982 + }, + { + "auxiliary_loss_clip": 0.01184851, + "auxiliary_loss_mlp": 0.01021228, + "balance_loss_clip": 1.04782939, + "balance_loss_mlp": 1.01387548, + "epoch": 0.6615763842962785, + "flos": 22894597128960.0, + "grad_norm": 2.657721596580447, + "language_loss": 0.79920608, + "learning_rate": 1.0857086906022313e-06, + "loss": 0.82126689, + "num_input_tokens_seen": 118370315, + "step": 5502, + "time_per_iteration": 2.631204605102539 + }, + { + "auxiliary_loss_clip": 0.01207042, + "auxiliary_loss_mlp": 0.01026683, + "balance_loss_clip": 1.04390228, + "balance_loss_mlp": 1.01907158, + "epoch": 0.6616966271869176, + "flos": 24790321221120.0, + "grad_norm": 1.8922869443985, + "language_loss": 0.73366094, + "learning_rate": 1.0850159489393388e-06, + "loss": 0.75599825, + "num_input_tokens_seen": 118389575, + "step": 5503, + "time_per_iteration": 2.8517208099365234 + }, + { + "auxiliary_loss_clip": 0.01187346, + "auxiliary_loss_mlp": 0.01025632, + "balance_loss_clip": 1.03906369, + "balance_loss_mlp": 1.01810706, + "epoch": 0.6618168700775566, + "flos": 17202109639680.0, + "grad_norm": 1.789844901137012, + "language_loss": 0.8230471, + "learning_rate": 1.0843233460787992e-06, + "loss": 0.84517682, + "num_input_tokens_seen": 118406790, + "step": 5504, + "time_per_iteration": 2.651590585708618 + }, + { + "auxiliary_loss_clip": 0.01188048, + "auxiliary_loss_mlp": 0.01029965, + "balance_loss_clip": 1.04613829, + "balance_loss_mlp": 1.02185845, + "epoch": 0.6619371129681958, + "flos": 25447091448960.0, + "grad_norm": 1.7469000851552505, + "language_loss": 0.77941239, + "learning_rate": 1.0836308821256805e-06, + "loss": 0.80159253, + "num_input_tokens_seen": 118427590, + "step": 5505, + "time_per_iteration": 2.7676830291748047 + }, + { + "auxiliary_loss_clip": 0.01178571, + "auxiliary_loss_mlp": 0.01021972, + "balance_loss_clip": 1.04858613, + "balance_loss_mlp": 1.01469374, + "epoch": 0.6620573558588349, + "flos": 18040444139520.0, + "grad_norm": 2.079434348557534, + "language_loss": 0.77799034, + "learning_rate": 1.0829385571850282e-06, + "loss": 0.79999578, + "num_input_tokens_seen": 118444570, + "step": 5506, + "time_per_iteration": 2.649899959564209 + }, + { + "auxiliary_loss_clip": 0.01175937, + "auxiliary_loss_mlp": 0.01026079, + "balance_loss_clip": 1.04958797, + "balance_loss_mlp": 1.01698923, + "epoch": 0.6621775987494739, + "flos": 17785586165760.0, + "grad_norm": 2.5610184256469686, + "language_loss": 0.83758801, + "learning_rate": 1.0822463713618679e-06, + "loss": 0.85960817, + "num_input_tokens_seen": 118461425, + "step": 5507, + "time_per_iteration": 2.6528854370117188 + }, + { + "auxiliary_loss_clip": 0.01200281, + "auxiliary_loss_mlp": 0.01028385, + "balance_loss_clip": 1.04545724, + "balance_loss_mlp": 1.02122331, + "epoch": 0.6622978416401131, + "flos": 17492590926720.0, + "grad_norm": 2.034884225440026, + "language_loss": 0.85116619, + "learning_rate": 1.0815543247612034e-06, + "loss": 0.87345284, + "num_input_tokens_seen": 118478495, + "step": 5508, + "time_per_iteration": 2.695185899734497 + }, + { + "auxiliary_loss_clip": 0.01186995, + "auxiliary_loss_mlp": 0.01027476, + "balance_loss_clip": 1.04325211, + "balance_loss_mlp": 1.01982832, + "epoch": 0.6624180845307521, + "flos": 21648352803840.0, + "grad_norm": 1.8368209651615952, + "language_loss": 0.83221394, + "learning_rate": 1.0808624174880168e-06, + "loss": 0.85435867, + "num_input_tokens_seen": 118499145, + "step": 5509, + "time_per_iteration": 2.6373324394226074 + }, + { + "auxiliary_loss_clip": 0.01169282, + "auxiliary_loss_mlp": 0.01022534, + "balance_loss_clip": 1.04846287, + "balance_loss_mlp": 1.01545644, + "epoch": 0.6625383274213912, + "flos": 23805902108160.0, + "grad_norm": 1.855408337150547, + "language_loss": 0.79954195, + "learning_rate": 1.080170649647272e-06, + "loss": 0.82146013, + "num_input_tokens_seen": 118518950, + "step": 5510, + "time_per_iteration": 2.6179730892181396 + }, + { + "auxiliary_loss_clip": 0.01169956, + "auxiliary_loss_mlp": 0.01027069, + "balance_loss_clip": 1.04865456, + "balance_loss_mlp": 1.02000928, + "epoch": 0.6626585703120303, + "flos": 33262941473280.0, + "grad_norm": 1.869485235429019, + "language_loss": 0.67488766, + "learning_rate": 1.0794790213439068e-06, + "loss": 0.69685793, + "num_input_tokens_seen": 118545850, + "step": 5511, + "time_per_iteration": 2.7011749744415283 + }, + { + "auxiliary_loss_clip": 0.01207321, + "auxiliary_loss_mlp": 0.01025227, + "balance_loss_clip": 1.04390538, + "balance_loss_mlp": 1.01671517, + "epoch": 0.6627788132026694, + "flos": 22085780630400.0, + "grad_norm": 2.1688734075276597, + "language_loss": 0.78181833, + "learning_rate": 1.078787532682843e-06, + "loss": 0.80414385, + "num_input_tokens_seen": 118563325, + "step": 5512, + "time_per_iteration": 2.789236545562744 + }, + { + "auxiliary_loss_clip": 0.01178304, + "auxiliary_loss_mlp": 0.01026759, + "balance_loss_clip": 1.0474925, + "balance_loss_mlp": 1.01943934, + "epoch": 0.6628990560933085, + "flos": 36173608260480.0, + "grad_norm": 2.4750995563709495, + "language_loss": 0.75869191, + "learning_rate": 1.0780961837689773e-06, + "loss": 0.78074253, + "num_input_tokens_seen": 118582835, + "step": 5513, + "time_per_iteration": 2.7400693893432617 + }, + { + "auxiliary_loss_clip": 0.01182267, + "auxiliary_loss_mlp": 0.0102755, + "balance_loss_clip": 1.0459305, + "balance_loss_mlp": 1.02024817, + "epoch": 0.6630192989839476, + "flos": 18513567106560.0, + "grad_norm": 1.6058062061388556, + "language_loss": 0.69808507, + "learning_rate": 1.0774049747071883e-06, + "loss": 0.72018325, + "num_input_tokens_seen": 118600715, + "step": 5514, + "time_per_iteration": 2.7001285552978516 + }, + { + "auxiliary_loss_clip": 0.01205891, + "auxiliary_loss_mlp": 0.01025155, + "balance_loss_clip": 1.04647386, + "balance_loss_mlp": 1.01771092, + "epoch": 0.6631395418745867, + "flos": 35809510049280.0, + "grad_norm": 1.6416830177864972, + "language_loss": 0.68118334, + "learning_rate": 1.076713905602332e-06, + "loss": 0.70349383, + "num_input_tokens_seen": 118621290, + "step": 5515, + "time_per_iteration": 2.8772714138031006 + }, + { + "auxiliary_loss_clip": 0.01183015, + "auxiliary_loss_mlp": 0.0102744, + "balance_loss_clip": 1.04772305, + "balance_loss_mlp": 1.02025795, + "epoch": 0.6632597847652257, + "flos": 20047742853120.0, + "grad_norm": 1.8164612152148245, + "language_loss": 0.81264418, + "learning_rate": 1.07602297655924e-06, + "loss": 0.83474869, + "num_input_tokens_seen": 118639610, + "step": 5516, + "time_per_iteration": 2.634650707244873 + }, + { + "auxiliary_loss_clip": 0.0117655, + "auxiliary_loss_mlp": 0.01027038, + "balance_loss_clip": 1.05244505, + "balance_loss_mlp": 1.01962376, + "epoch": 0.6633800276558649, + "flos": 21214480423680.0, + "grad_norm": 1.7928393465225156, + "language_loss": 0.8142879, + "learning_rate": 1.0753321876827292e-06, + "loss": 0.8363238, + "num_input_tokens_seen": 118658895, + "step": 5517, + "time_per_iteration": 2.6386969089508057 + }, + { + "auxiliary_loss_clip": 0.01171798, + "auxiliary_loss_mlp": 0.01024915, + "balance_loss_clip": 1.04842961, + "balance_loss_mlp": 1.01770329, + "epoch": 0.663500270546504, + "flos": 23987753688960.0, + "grad_norm": 2.00513347743413, + "language_loss": 0.73963088, + "learning_rate": 1.0746415390775893e-06, + "loss": 0.76159799, + "num_input_tokens_seen": 118677025, + "step": 5518, + "time_per_iteration": 2.673635959625244 + }, + { + "auxiliary_loss_clip": 0.01175314, + "auxiliary_loss_mlp": 0.01024365, + "balance_loss_clip": 1.0524286, + "balance_loss_mlp": 1.0170157, + "epoch": 0.663620513437143, + "flos": 17932389050880.0, + "grad_norm": 3.823100842962606, + "language_loss": 0.76476026, + "learning_rate": 1.0739510308485939e-06, + "loss": 0.78675699, + "num_input_tokens_seen": 118694240, + "step": 5519, + "time_per_iteration": 3.5111217498779297 + }, + { + "auxiliary_loss_clip": 0.01107639, + "auxiliary_loss_mlp": 0.01000474, + "balance_loss_clip": 1.01188815, + "balance_loss_mlp": 0.99926978, + "epoch": 0.6637407563277821, + "flos": 57840241086720.0, + "grad_norm": 0.8129811174506838, + "language_loss": 0.62526953, + "learning_rate": 1.07326066310049e-06, + "loss": 0.64635074, + "num_input_tokens_seen": 118758365, + "step": 5520, + "time_per_iteration": 4.3136210441589355 + }, + { + "auxiliary_loss_clip": 0.01190306, + "auxiliary_loss_mlp": 0.01025851, + "balance_loss_clip": 1.04304564, + "balance_loss_mlp": 1.01797688, + "epoch": 0.6638609992184212, + "flos": 27306007079040.0, + "grad_norm": 4.56677052989168, + "language_loss": 0.79665029, + "learning_rate": 1.0725704359380059e-06, + "loss": 0.81881189, + "num_input_tokens_seen": 118778220, + "step": 5521, + "time_per_iteration": 2.738121509552002 + }, + { + "auxiliary_loss_clip": 0.01174176, + "auxiliary_loss_mlp": 0.01024586, + "balance_loss_clip": 1.04961765, + "balance_loss_mlp": 1.01722455, + "epoch": 0.6639812421090603, + "flos": 18624854419200.0, + "grad_norm": 1.9906474265288279, + "language_loss": 0.71789145, + "learning_rate": 1.0718803494658497e-06, + "loss": 0.73987901, + "num_input_tokens_seen": 118797110, + "step": 5522, + "time_per_iteration": 2.5792088508605957 + }, + { + "auxiliary_loss_clip": 0.01225115, + "auxiliary_loss_mlp": 0.01027187, + "balance_loss_clip": 1.03904772, + "balance_loss_mlp": 1.01943254, + "epoch": 0.6641014849996993, + "flos": 15924479806080.0, + "grad_norm": 2.062099604705986, + "language_loss": 0.83454704, + "learning_rate": 1.071190403788707e-06, + "loss": 0.85706997, + "num_input_tokens_seen": 118812415, + "step": 5523, + "time_per_iteration": 3.81178879737854 + }, + { + "auxiliary_loss_clip": 0.01205297, + "auxiliary_loss_mlp": 0.01026032, + "balance_loss_clip": 1.04630518, + "balance_loss_mlp": 1.01730037, + "epoch": 0.6642217278903385, + "flos": 26505486622080.0, + "grad_norm": 1.8032001815049679, + "language_loss": 0.75175422, + "learning_rate": 1.0705005990112415e-06, + "loss": 0.77406752, + "num_input_tokens_seen": 118832195, + "step": 5524, + "time_per_iteration": 3.4653024673461914 + }, + { + "auxiliary_loss_clip": 0.01196117, + "auxiliary_loss_mlp": 0.01025818, + "balance_loss_clip": 1.04599237, + "balance_loss_mlp": 1.01817107, + "epoch": 0.6643419707809776, + "flos": 15377308951680.0, + "grad_norm": 9.927841930693235, + "language_loss": 0.74319923, + "learning_rate": 1.0698109352380957e-06, + "loss": 0.76541853, + "num_input_tokens_seen": 118849795, + "step": 5525, + "time_per_iteration": 2.712953805923462 + }, + { + "auxiliary_loss_clip": 0.01171037, + "auxiliary_loss_mlp": 0.01026012, + "balance_loss_clip": 1.04930139, + "balance_loss_mlp": 1.01862693, + "epoch": 0.6644622136716166, + "flos": 25117610970240.0, + "grad_norm": 2.4189938334141092, + "language_loss": 0.77712286, + "learning_rate": 1.0691214125738909e-06, + "loss": 0.79909331, + "num_input_tokens_seen": 118870000, + "step": 5526, + "time_per_iteration": 3.782078266143799 + }, + { + "auxiliary_loss_clip": 0.01066275, + "auxiliary_loss_mlp": 0.01000133, + "balance_loss_clip": 1.01210165, + "balance_loss_mlp": 0.99897689, + "epoch": 0.6645824565622558, + "flos": 66201717680640.0, + "grad_norm": 0.797143755540754, + "language_loss": 0.57468456, + "learning_rate": 1.0684320311232287e-06, + "loss": 0.5953486, + "num_input_tokens_seen": 118932905, + "step": 5527, + "time_per_iteration": 3.256420612335205 + }, + { + "auxiliary_loss_clip": 0.01184095, + "auxiliary_loss_mlp": 0.01028822, + "balance_loss_clip": 1.0450983, + "balance_loss_mlp": 1.02013135, + "epoch": 0.6647026994528948, + "flos": 25082131311360.0, + "grad_norm": 1.9271667203930665, + "language_loss": 0.81535506, + "learning_rate": 1.0677427909906865e-06, + "loss": 0.83748424, + "num_input_tokens_seen": 118953355, + "step": 5528, + "time_per_iteration": 2.7951691150665283 + }, + { + "auxiliary_loss_clip": 0.01176943, + "auxiliary_loss_mlp": 0.01026562, + "balance_loss_clip": 1.05096626, + "balance_loss_mlp": 1.0192008, + "epoch": 0.6648229423435339, + "flos": 18222187979520.0, + "grad_norm": 1.9870777002832425, + "language_loss": 0.72084367, + "learning_rate": 1.0670536922808216e-06, + "loss": 0.74287874, + "num_input_tokens_seen": 118973480, + "step": 5529, + "time_per_iteration": 2.6109206676483154 + }, + { + "auxiliary_loss_clip": 0.01191689, + "auxiliary_loss_mlp": 0.01025941, + "balance_loss_clip": 1.0489881, + "balance_loss_mlp": 1.01832652, + "epoch": 0.6649431852341731, + "flos": 18296882311680.0, + "grad_norm": 2.0828237809790484, + "language_loss": 0.7156359, + "learning_rate": 1.06636473509817e-06, + "loss": 0.73781216, + "num_input_tokens_seen": 118989860, + "step": 5530, + "time_per_iteration": 2.6718790531158447 + }, + { + "auxiliary_loss_clip": 0.01182462, + "auxiliary_loss_mlp": 0.01378682, + "balance_loss_clip": 1.04392409, + "balance_loss_mlp": 1.00017309, + "epoch": 0.6650634281248121, + "flos": 17019575700480.0, + "grad_norm": 2.536214035279983, + "language_loss": 0.80426502, + "learning_rate": 1.0656759195472447e-06, + "loss": 0.82987642, + "num_input_tokens_seen": 119007150, + "step": 5531, + "time_per_iteration": 2.690462589263916 + }, + { + "auxiliary_loss_clip": 0.01090221, + "auxiliary_loss_mlp": 0.00999329, + "balance_loss_clip": 1.01235425, + "balance_loss_mlp": 0.99810117, + "epoch": 0.6651836710154512, + "flos": 69294810666240.0, + "grad_norm": 0.779017403498313, + "language_loss": 0.5971598, + "learning_rate": 1.0649872457325414e-06, + "loss": 0.61805528, + "num_input_tokens_seen": 119068435, + "step": 5532, + "time_per_iteration": 3.195887804031372 + }, + { + "auxiliary_loss_clip": 0.01079786, + "auxiliary_loss_mlp": 0.00999771, + "balance_loss_clip": 1.01194429, + "balance_loss_mlp": 0.99851972, + "epoch": 0.6653039139060903, + "flos": 66883444882560.0, + "grad_norm": 0.8581359301233291, + "language_loss": 0.55086726, + "learning_rate": 1.0642987137585278e-06, + "loss": 0.57166284, + "num_input_tokens_seen": 119127960, + "step": 5533, + "time_per_iteration": 3.1650779247283936 + }, + { + "auxiliary_loss_clip": 0.01188993, + "auxiliary_loss_mlp": 0.0103014, + "balance_loss_clip": 1.04704893, + "balance_loss_mlp": 1.02286839, + "epoch": 0.6654241567967294, + "flos": 21470056669440.0, + "grad_norm": 1.7354365457583634, + "language_loss": 0.82501495, + "learning_rate": 1.0636103237296561e-06, + "loss": 0.84720629, + "num_input_tokens_seen": 119146885, + "step": 5534, + "time_per_iteration": 2.736976385116577 + }, + { + "auxiliary_loss_clip": 0.0118177, + "auxiliary_loss_mlp": 0.01024549, + "balance_loss_clip": 1.05217552, + "balance_loss_mlp": 1.01774216, + "epoch": 0.6655443996873684, + "flos": 25119514391040.0, + "grad_norm": 1.7761110293403353, + "language_loss": 0.84130621, + "learning_rate": 1.062922075750353e-06, + "loss": 0.86336946, + "num_input_tokens_seen": 119166900, + "step": 5535, + "time_per_iteration": 2.6242313385009766 + }, + { + "auxiliary_loss_clip": 0.01202792, + "auxiliary_loss_mlp": 0.01023238, + "balance_loss_clip": 1.04716825, + "balance_loss_mlp": 1.01592422, + "epoch": 0.6656646425780076, + "flos": 17457326749440.0, + "grad_norm": 2.0718274907570224, + "language_loss": 0.72080803, + "learning_rate": 1.0622339699250267e-06, + "loss": 0.7430684, + "num_input_tokens_seen": 119184820, + "step": 5536, + "time_per_iteration": 2.761080741882324 + }, + { + "auxiliary_loss_clip": 0.01197993, + "auxiliary_loss_mlp": 0.01021582, + "balance_loss_clip": 1.04417372, + "balance_loss_mlp": 1.01406622, + "epoch": 0.6657848854686467, + "flos": 23434190213760.0, + "grad_norm": 1.6419363294907683, + "language_loss": 0.79204416, + "learning_rate": 1.0615460063580624e-06, + "loss": 0.81423986, + "num_input_tokens_seen": 119203295, + "step": 5537, + "time_per_iteration": 2.7824923992156982 + }, + { + "auxiliary_loss_clip": 0.01192312, + "auxiliary_loss_mlp": 0.01024221, + "balance_loss_clip": 1.04703045, + "balance_loss_mlp": 1.01662779, + "epoch": 0.6659051283592857, + "flos": 11509909459200.0, + "grad_norm": 3.2091240871908977, + "language_loss": 0.72503704, + "learning_rate": 1.060858185153821e-06, + "loss": 0.7472024, + "num_input_tokens_seen": 119221395, + "step": 5538, + "time_per_iteration": 2.6517858505249023 + }, + { + "auxiliary_loss_clip": 0.01194584, + "auxiliary_loss_mlp": 0.01026084, + "balance_loss_clip": 1.0467751, + "balance_loss_mlp": 1.01799607, + "epoch": 0.6660253712499249, + "flos": 20594554571520.0, + "grad_norm": 5.582406933756351, + "language_loss": 0.76506662, + "learning_rate": 1.0601705064166474e-06, + "loss": 0.78727329, + "num_input_tokens_seen": 119239790, + "step": 5539, + "time_per_iteration": 2.764200210571289 + }, + { + "auxiliary_loss_clip": 0.01186819, + "auxiliary_loss_mlp": 0.01023498, + "balance_loss_clip": 1.04691386, + "balance_loss_mlp": 1.01619947, + "epoch": 0.666145614140564, + "flos": 21251504367360.0, + "grad_norm": 2.144383080296833, + "language_loss": 0.73019457, + "learning_rate": 1.0594829702508596e-06, + "loss": 0.7522977, + "num_input_tokens_seen": 119257505, + "step": 5540, + "time_per_iteration": 2.715040683746338 + }, + { + "auxiliary_loss_clip": 0.01199061, + "auxiliary_loss_mlp": 0.01025518, + "balance_loss_clip": 1.04407144, + "balance_loss_mlp": 1.01834142, + "epoch": 0.666265857031203, + "flos": 33726188200320.0, + "grad_norm": 2.5826696956783333, + "language_loss": 0.55279791, + "learning_rate": 1.0587955767607592e-06, + "loss": 0.57504368, + "num_input_tokens_seen": 119279365, + "step": 5541, + "time_per_iteration": 2.869196891784668 + }, + { + "auxiliary_loss_clip": 0.0117163, + "auxiliary_loss_mlp": 0.0102709, + "balance_loss_clip": 1.04859507, + "balance_loss_mlp": 1.02018762, + "epoch": 0.6663860999218422, + "flos": 17456644391040.0, + "grad_norm": 2.463111190427129, + "language_loss": 0.76808262, + "learning_rate": 1.0581083260506206e-06, + "loss": 0.79006982, + "num_input_tokens_seen": 119296150, + "step": 5542, + "time_per_iteration": 2.5649774074554443 + }, + { + "auxiliary_loss_clip": 0.01189946, + "auxiliary_loss_mlp": 0.01025926, + "balance_loss_clip": 1.04669142, + "balance_loss_mlp": 1.01894069, + "epoch": 0.6665063428124812, + "flos": 17676740977920.0, + "grad_norm": 2.2598569326922333, + "language_loss": 0.77043545, + "learning_rate": 1.0574212182246993e-06, + "loss": 0.79259419, + "num_input_tokens_seen": 119314845, + "step": 5543, + "time_per_iteration": 2.726135492324829 + }, + { + "auxiliary_loss_clip": 0.01195418, + "auxiliary_loss_mlp": 0.01023943, + "balance_loss_clip": 1.04744244, + "balance_loss_mlp": 1.01593804, + "epoch": 0.6666265857031203, + "flos": 27673265687040.0, + "grad_norm": 2.7117129015888795, + "language_loss": 0.76124132, + "learning_rate": 1.0567342533872303e-06, + "loss": 0.78343493, + "num_input_tokens_seen": 119334875, + "step": 5544, + "time_per_iteration": 2.6880998611450195 + }, + { + "auxiliary_loss_clip": 0.01192954, + "auxiliary_loss_mlp": 0.01026556, + "balance_loss_clip": 1.04843783, + "balance_loss_mlp": 1.01907539, + "epoch": 0.6667468285937594, + "flos": 25046831220480.0, + "grad_norm": 1.7127482609064164, + "language_loss": 0.80813676, + "learning_rate": 1.0560474316424255e-06, + "loss": 0.83033192, + "num_input_tokens_seen": 119354635, + "step": 5545, + "time_per_iteration": 3.7004239559173584 + }, + { + "auxiliary_loss_clip": 0.01187759, + "auxiliary_loss_mlp": 0.0102459, + "balance_loss_clip": 1.04467607, + "balance_loss_mlp": 1.01684117, + "epoch": 0.6668670714843985, + "flos": 22780472641920.0, + "grad_norm": 2.5127884303836265, + "language_loss": 0.73646986, + "learning_rate": 1.0553607530944746e-06, + "loss": 0.75859338, + "num_input_tokens_seen": 119372690, + "step": 5546, + "time_per_iteration": 3.6497256755828857 + }, + { + "auxiliary_loss_clip": 0.01199141, + "auxiliary_loss_mlp": 0.01024384, + "balance_loss_clip": 1.04428077, + "balance_loss_mlp": 1.01627195, + "epoch": 0.6669873143750376, + "flos": 22163886754560.0, + "grad_norm": 2.0030070783421907, + "language_loss": 0.89697999, + "learning_rate": 1.0546742178475463e-06, + "loss": 0.9192152, + "num_input_tokens_seen": 119391685, + "step": 5547, + "time_per_iteration": 2.712569236755371 + }, + { + "auxiliary_loss_clip": 0.01212922, + "auxiliary_loss_mlp": 0.01022965, + "balance_loss_clip": 1.04608238, + "balance_loss_mlp": 1.01596773, + "epoch": 0.6671075572656767, + "flos": 20514832335360.0, + "grad_norm": 3.0531770533015594, + "language_loss": 0.86720216, + "learning_rate": 1.0539878260057868e-06, + "loss": 0.889561, + "num_input_tokens_seen": 119410725, + "step": 5548, + "time_per_iteration": 2.7240450382232666 + }, + { + "auxiliary_loss_clip": 0.01183252, + "auxiliary_loss_mlp": 0.01024288, + "balance_loss_clip": 1.0498333, + "balance_loss_mlp": 1.01569939, + "epoch": 0.6672278001563158, + "flos": 17931203902080.0, + "grad_norm": 3.3877654590020994, + "language_loss": 0.68772829, + "learning_rate": 1.0533015776733226e-06, + "loss": 0.70980364, + "num_input_tokens_seen": 119426875, + "step": 5549, + "time_per_iteration": 3.6101768016815186 + }, + { + "auxiliary_loss_clip": 0.01191027, + "auxiliary_loss_mlp": 0.01026389, + "balance_loss_clip": 1.05010319, + "balance_loss_mlp": 1.01833093, + "epoch": 0.6673480430469548, + "flos": 22342146975360.0, + "grad_norm": 2.7323653288129464, + "language_loss": 0.7850731, + "learning_rate": 1.0526154729542566e-06, + "loss": 0.80724728, + "num_input_tokens_seen": 119446935, + "step": 5550, + "time_per_iteration": 2.661208152770996 + }, + { + "auxiliary_loss_clip": 0.0119947, + "auxiliary_loss_mlp": 0.01027281, + "balance_loss_clip": 1.04612374, + "balance_loss_mlp": 1.0195291, + "epoch": 0.6674682859375939, + "flos": 20703830722560.0, + "grad_norm": 2.4316819544627184, + "language_loss": 0.80498606, + "learning_rate": 1.0519295119526699e-06, + "loss": 0.82725364, + "num_input_tokens_seen": 119463240, + "step": 5551, + "time_per_iteration": 2.7199814319610596 + }, + { + "auxiliary_loss_clip": 0.01193288, + "auxiliary_loss_mlp": 0.01028604, + "balance_loss_clip": 1.04650199, + "balance_loss_mlp": 1.0210197, + "epoch": 0.667588528828233, + "flos": 26206673379840.0, + "grad_norm": 1.7565344813807273, + "language_loss": 0.83030176, + "learning_rate": 1.0512436947726227e-06, + "loss": 0.8525207, + "num_input_tokens_seen": 119484655, + "step": 5552, + "time_per_iteration": 3.612508773803711 + }, + { + "auxiliary_loss_clip": 0.01199326, + "auxiliary_loss_mlp": 0.01027246, + "balance_loss_clip": 1.04386675, + "balance_loss_mlp": 1.01925921, + "epoch": 0.6677087717188721, + "flos": 23071025756160.0, + "grad_norm": 2.463129836681595, + "language_loss": 0.65422463, + "learning_rate": 1.0505580215181517e-06, + "loss": 0.67649031, + "num_input_tokens_seen": 119502895, + "step": 5553, + "time_per_iteration": 2.734232187271118 + }, + { + "auxiliary_loss_clip": 0.01097481, + "auxiliary_loss_mlp": 0.00998889, + "balance_loss_clip": 1.01317525, + "balance_loss_mlp": 0.99770337, + "epoch": 0.6678290146095112, + "flos": 70941315219840.0, + "grad_norm": 0.7836142308860142, + "language_loss": 0.5662781, + "learning_rate": 1.0498724922932753e-06, + "loss": 0.58724183, + "num_input_tokens_seen": 119561010, + "step": 5554, + "time_per_iteration": 3.181422472000122 + }, + { + "auxiliary_loss_clip": 0.01177134, + "auxiliary_loss_mlp": 0.01026236, + "balance_loss_clip": 1.05076075, + "balance_loss_mlp": 1.01810014, + "epoch": 0.6679492575001503, + "flos": 18661088263680.0, + "grad_norm": 2.665619658190276, + "language_loss": 0.86174983, + "learning_rate": 1.0491871072019851e-06, + "loss": 0.88378352, + "num_input_tokens_seen": 119578900, + "step": 5555, + "time_per_iteration": 2.528498411178589 + }, + { + "auxiliary_loss_clip": 0.01200178, + "auxiliary_loss_mlp": 0.01027122, + "balance_loss_clip": 1.04203868, + "balance_loss_mlp": 1.01947486, + "epoch": 0.6680695003907894, + "flos": 29711985822720.0, + "grad_norm": 1.888657203256423, + "language_loss": 0.63742924, + "learning_rate": 1.0485018663482555e-06, + "loss": 0.6597023, + "num_input_tokens_seen": 119598920, + "step": 5556, + "time_per_iteration": 2.77152419090271 + }, + { + "auxiliary_loss_clip": 0.01174837, + "auxiliary_loss_mlp": 0.01021794, + "balance_loss_clip": 1.04646826, + "balance_loss_mlp": 1.01457596, + "epoch": 0.6681897432814284, + "flos": 28218964083840.0, + "grad_norm": 2.7317238075577017, + "language_loss": 0.70453131, + "learning_rate": 1.0478167698360354e-06, + "loss": 0.72649765, + "num_input_tokens_seen": 119618220, + "step": 5557, + "time_per_iteration": 2.622169256210327 + }, + { + "auxiliary_loss_clip": 0.01172151, + "auxiliary_loss_mlp": 0.01033606, + "balance_loss_clip": 1.04479718, + "balance_loss_mlp": 1.02617371, + "epoch": 0.6683099861720676, + "flos": 25046543911680.0, + "grad_norm": 2.119206128655778, + "language_loss": 0.70038342, + "learning_rate": 1.0471318177692556e-06, + "loss": 0.72244102, + "num_input_tokens_seen": 119638520, + "step": 5558, + "time_per_iteration": 2.624345064163208 + }, + { + "auxiliary_loss_clip": 0.01210168, + "auxiliary_loss_mlp": 0.01025582, + "balance_loss_clip": 1.04319406, + "balance_loss_mlp": 1.01811028, + "epoch": 0.6684302290627067, + "flos": 22996977868800.0, + "grad_norm": 2.7477209037551225, + "language_loss": 0.7593224, + "learning_rate": 1.046447010251821e-06, + "loss": 0.78167987, + "num_input_tokens_seen": 119655850, + "step": 5559, + "time_per_iteration": 2.7228550910949707 + }, + { + "auxiliary_loss_clip": 0.01188572, + "auxiliary_loss_mlp": 0.01025017, + "balance_loss_clip": 1.0483402, + "balance_loss_mlp": 1.01754308, + "epoch": 0.6685504719533457, + "flos": 26573824247040.0, + "grad_norm": 1.951003385069968, + "language_loss": 0.76250947, + "learning_rate": 1.0457623473876157e-06, + "loss": 0.78464532, + "num_input_tokens_seen": 119675355, + "step": 5560, + "time_per_iteration": 2.7367959022521973 + }, + { + "auxiliary_loss_clip": 0.01170359, + "auxiliary_loss_mlp": 0.01023325, + "balance_loss_clip": 1.0488193, + "balance_loss_mlp": 1.01608932, + "epoch": 0.6686707148439849, + "flos": 28986087870720.0, + "grad_norm": 5.219191265617622, + "language_loss": 0.70999479, + "learning_rate": 1.0450778292805046e-06, + "loss": 0.73193157, + "num_input_tokens_seen": 119695340, + "step": 5561, + "time_per_iteration": 2.6345856189727783 + }, + { + "auxiliary_loss_clip": 0.01184461, + "auxiliary_loss_mlp": 0.010254, + "balance_loss_clip": 1.04752374, + "balance_loss_mlp": 1.01768088, + "epoch": 0.6687909577346239, + "flos": 23623152687360.0, + "grad_norm": 1.5321869339086618, + "language_loss": 0.78418672, + "learning_rate": 1.0443934560343267e-06, + "loss": 0.80628532, + "num_input_tokens_seen": 119716750, + "step": 5562, + "time_per_iteration": 2.7633841037750244 + }, + { + "auxiliary_loss_clip": 0.01185363, + "auxiliary_loss_mlp": 0.01025643, + "balance_loss_clip": 1.04313922, + "balance_loss_mlp": 1.01776314, + "epoch": 0.668911200625263, + "flos": 23148593176320.0, + "grad_norm": 1.8955204491670072, + "language_loss": 0.78222251, + "learning_rate": 1.0437092277529034e-06, + "loss": 0.80433255, + "num_input_tokens_seen": 119736005, + "step": 5563, + "time_per_iteration": 2.6834521293640137 + }, + { + "auxiliary_loss_clip": 0.01182837, + "auxiliary_loss_mlp": 0.01029164, + "balance_loss_clip": 1.04353237, + "balance_loss_mlp": 1.02186227, + "epoch": 0.6690314435159022, + "flos": 18551919853440.0, + "grad_norm": 2.0637614087494556, + "language_loss": 0.73639089, + "learning_rate": 1.0430251445400292e-06, + "loss": 0.75851095, + "num_input_tokens_seen": 119754050, + "step": 5564, + "time_per_iteration": 2.626030683517456 + }, + { + "auxiliary_loss_clip": 0.01235539, + "auxiliary_loss_mlp": 0.01024145, + "balance_loss_clip": 1.04293358, + "balance_loss_mlp": 1.01686072, + "epoch": 0.6691516864065412, + "flos": 31759540704000.0, + "grad_norm": 2.198022173550645, + "language_loss": 0.62760866, + "learning_rate": 1.0423412064994787e-06, + "loss": 0.65020543, + "num_input_tokens_seen": 119774820, + "step": 5565, + "time_per_iteration": 3.1318984031677246 + }, + { + "auxiliary_loss_clip": 0.01196885, + "auxiliary_loss_mlp": 0.01022121, + "balance_loss_clip": 1.04309964, + "balance_loss_mlp": 1.01530194, + "epoch": 0.6692719292971803, + "flos": 34933864296960.0, + "grad_norm": 1.9955950345049358, + "language_loss": 0.74180257, + "learning_rate": 1.0416574137350064e-06, + "loss": 0.76399261, + "num_input_tokens_seen": 119795525, + "step": 5566, + "time_per_iteration": 2.9387149810791016 + }, + { + "auxiliary_loss_clip": 0.01171026, + "auxiliary_loss_mlp": 0.01022268, + "balance_loss_clip": 1.0462327, + "balance_loss_mlp": 1.01431048, + "epoch": 0.6693921721878194, + "flos": 20449188230400.0, + "grad_norm": 2.596383643685039, + "language_loss": 0.81054711, + "learning_rate": 1.0409737663503428e-06, + "loss": 0.83248007, + "num_input_tokens_seen": 119813905, + "step": 5567, + "time_per_iteration": 2.637279987335205 + }, + { + "auxiliary_loss_clip": 0.01174561, + "auxiliary_loss_mlp": 0.01027641, + "balance_loss_clip": 1.0436132, + "balance_loss_mlp": 1.0199821, + "epoch": 0.6695124150784585, + "flos": 16614538963200.0, + "grad_norm": 2.0897102457668484, + "language_loss": 0.82366216, + "learning_rate": 1.040290264449196e-06, + "loss": 0.84568417, + "num_input_tokens_seen": 119832010, + "step": 5568, + "time_per_iteration": 2.6136553287506104 + }, + { + "auxiliary_loss_clip": 0.01176388, + "auxiliary_loss_mlp": 0.01026406, + "balance_loss_clip": 1.0481267, + "balance_loss_mlp": 1.01873183, + "epoch": 0.6696326579690975, + "flos": 26652145852800.0, + "grad_norm": 4.586968086456213, + "language_loss": 0.64154518, + "learning_rate": 1.0396069081352532e-06, + "loss": 0.66357315, + "num_input_tokens_seen": 119851165, + "step": 5569, + "time_per_iteration": 2.666175127029419 + }, + { + "auxiliary_loss_clip": 0.01065899, + "auxiliary_loss_mlp": 0.01001904, + "balance_loss_clip": 1.01141119, + "balance_loss_mlp": 1.00062835, + "epoch": 0.6697529008597367, + "flos": 66964603662720.0, + "grad_norm": 0.8251658152294253, + "language_loss": 0.56059819, + "learning_rate": 1.0389236975121782e-06, + "loss": 0.58127618, + "num_input_tokens_seen": 119906015, + "step": 5570, + "time_per_iteration": 3.1064677238464355 + }, + { + "auxiliary_loss_clip": 0.01175599, + "auxiliary_loss_mlp": 0.0103095, + "balance_loss_clip": 1.0501225, + "balance_loss_mlp": 1.02333903, + "epoch": 0.6698731437503758, + "flos": 20886939279360.0, + "grad_norm": 1.8177651926391007, + "language_loss": 0.7145946, + "learning_rate": 1.0382406326836147e-06, + "loss": 0.73666006, + "num_input_tokens_seen": 119925160, + "step": 5571, + "time_per_iteration": 3.492668390274048 + }, + { + "auxiliary_loss_clip": 0.01186201, + "auxiliary_loss_mlp": 0.01025305, + "balance_loss_clip": 1.04887772, + "balance_loss_mlp": 1.01756823, + "epoch": 0.6699933866410148, + "flos": 20409470766720.0, + "grad_norm": 2.1761029560279606, + "language_loss": 0.76199281, + "learning_rate": 1.0375577137531828e-06, + "loss": 0.78410786, + "num_input_tokens_seen": 119943720, + "step": 5572, + "time_per_iteration": 3.8412294387817383 + }, + { + "auxiliary_loss_clip": 0.0119148, + "auxiliary_loss_mlp": 0.01026116, + "balance_loss_clip": 1.04496562, + "balance_loss_mlp": 1.01856995, + "epoch": 0.670113629531654, + "flos": 29023075900800.0, + "grad_norm": 1.555834679731226, + "language_loss": 0.72185576, + "learning_rate": 1.0368749408244802e-06, + "loss": 0.74403173, + "num_input_tokens_seen": 119966640, + "step": 5573, + "time_per_iteration": 2.76656436920166 + }, + { + "auxiliary_loss_clip": 0.01174413, + "auxiliary_loss_mlp": 0.01027253, + "balance_loss_clip": 1.04739094, + "balance_loss_mlp": 1.02026772, + "epoch": 0.670233872422293, + "flos": 19791699730560.0, + "grad_norm": 3.5545350674031004, + "language_loss": 0.785074, + "learning_rate": 1.0361923140010836e-06, + "loss": 0.80709076, + "num_input_tokens_seen": 119985125, + "step": 5574, + "time_per_iteration": 2.6852643489837646 + }, + { + "auxiliary_loss_clip": 0.01184473, + "auxiliary_loss_mlp": 0.0102392, + "balance_loss_clip": 1.04719758, + "balance_loss_mlp": 1.0161413, + "epoch": 0.6703541153129321, + "flos": 24243689070720.0, + "grad_norm": 1.9897281302181422, + "language_loss": 0.63437188, + "learning_rate": 1.0355098333865455e-06, + "loss": 0.65645576, + "num_input_tokens_seen": 120004355, + "step": 5575, + "time_per_iteration": 3.579127311706543 + }, + { + "auxiliary_loss_clip": 0.01178089, + "auxiliary_loss_mlp": 0.01031818, + "balance_loss_clip": 1.05032325, + "balance_loss_mlp": 1.02481484, + "epoch": 0.6704743582035713, + "flos": 26688523351680.0, + "grad_norm": 1.8021596604303993, + "language_loss": 0.6925295, + "learning_rate": 1.0348274990844006e-06, + "loss": 0.71462858, + "num_input_tokens_seen": 120027115, + "step": 5576, + "time_per_iteration": 2.687769651412964 + }, + { + "auxiliary_loss_clip": 0.01178353, + "auxiliary_loss_mlp": 0.01030618, + "balance_loss_clip": 1.04774737, + "balance_loss_mlp": 1.02337313, + "epoch": 0.6705946010942103, + "flos": 23514379326720.0, + "grad_norm": 2.0486063128710352, + "language_loss": 0.7262696, + "learning_rate": 1.034145311198155e-06, + "loss": 0.74835938, + "num_input_tokens_seen": 120047130, + "step": 5577, + "time_per_iteration": 2.658998489379883 + }, + { + "auxiliary_loss_clip": 0.01170257, + "auxiliary_loss_mlp": 0.01025998, + "balance_loss_clip": 1.04835379, + "balance_loss_mlp": 1.0188818, + "epoch": 0.6707148439848494, + "flos": 24061011477120.0, + "grad_norm": 1.64783926408295, + "language_loss": 0.6365521, + "learning_rate": 1.0334632698312989e-06, + "loss": 0.65851462, + "num_input_tokens_seen": 120067925, + "step": 5578, + "time_per_iteration": 3.4854607582092285 + }, + { + "auxiliary_loss_clip": 0.01181501, + "auxiliary_loss_mlp": 0.01025437, + "balance_loss_clip": 1.04462862, + "balance_loss_mlp": 1.01707458, + "epoch": 0.6708350868754885, + "flos": 22528667324160.0, + "grad_norm": 1.8060363543716682, + "language_loss": 0.75584817, + "learning_rate": 1.032781375087295e-06, + "loss": 0.7779175, + "num_input_tokens_seen": 120087825, + "step": 5579, + "time_per_iteration": 2.692117691040039 + }, + { + "auxiliary_loss_clip": 0.01192097, + "auxiliary_loss_mlp": 0.01027292, + "balance_loss_clip": 1.04857361, + "balance_loss_mlp": 1.02038085, + "epoch": 0.6709553297661276, + "flos": 25227749047680.0, + "grad_norm": 1.5169941795730495, + "language_loss": 0.67571032, + "learning_rate": 1.0320996270695891e-06, + "loss": 0.69790423, + "num_input_tokens_seen": 120108895, + "step": 5580, + "time_per_iteration": 2.6927592754364014 + }, + { + "auxiliary_loss_clip": 0.0119472, + "auxiliary_loss_mlp": 0.01024869, + "balance_loss_clip": 1.04217052, + "balance_loss_mlp": 1.01780009, + "epoch": 0.6710755726567667, + "flos": 20448757267200.0, + "grad_norm": 1.8569095436034129, + "language_loss": 0.732952, + "learning_rate": 1.0314180258815998e-06, + "loss": 0.75514787, + "num_input_tokens_seen": 120127535, + "step": 5581, + "time_per_iteration": 2.666337013244629 + }, + { + "auxiliary_loss_clip": 0.01185918, + "auxiliary_loss_mlp": 0.01024605, + "balance_loss_clip": 1.04188943, + "balance_loss_mlp": 1.01761937, + "epoch": 0.6711958155474057, + "flos": 25995411538560.0, + "grad_norm": 1.695431761635086, + "language_loss": 0.74382782, + "learning_rate": 1.0307365716267247e-06, + "loss": 0.76593304, + "num_input_tokens_seen": 120147980, + "step": 5582, + "time_per_iteration": 2.8066248893737793 + }, + { + "auxiliary_loss_clip": 0.01177999, + "auxiliary_loss_mlp": 0.01028816, + "balance_loss_clip": 1.04581261, + "balance_loss_mlp": 1.02117467, + "epoch": 0.6713160584380449, + "flos": 19937712516480.0, + "grad_norm": 2.277474640601085, + "language_loss": 0.77933586, + "learning_rate": 1.0300552644083423e-06, + "loss": 0.801404, + "num_input_tokens_seen": 120166905, + "step": 5583, + "time_per_iteration": 2.559889554977417 + }, + { + "auxiliary_loss_clip": 0.01205047, + "auxiliary_loss_mlp": 0.01031522, + "balance_loss_clip": 1.04769206, + "balance_loss_mlp": 1.02231264, + "epoch": 0.6714363013286839, + "flos": 18223373128320.0, + "grad_norm": 2.822177333879381, + "language_loss": 0.73072326, + "learning_rate": 1.0293741043298036e-06, + "loss": 0.75308895, + "num_input_tokens_seen": 120185255, + "step": 5584, + "time_per_iteration": 2.693446397781372 + }, + { + "auxiliary_loss_clip": 0.01203731, + "auxiliary_loss_mlp": 0.01023633, + "balance_loss_clip": 1.04979968, + "balance_loss_mlp": 1.01615262, + "epoch": 0.671556544219323, + "flos": 25812374808960.0, + "grad_norm": 3.206656554731369, + "language_loss": 0.71609461, + "learning_rate": 1.0286930914944436e-06, + "loss": 0.73836827, + "num_input_tokens_seen": 120205070, + "step": 5585, + "time_per_iteration": 2.7310967445373535 + }, + { + "auxiliary_loss_clip": 0.01170752, + "auxiliary_loss_mlp": 0.01025564, + "balance_loss_clip": 1.04684305, + "balance_loss_mlp": 1.01750517, + "epoch": 0.6716767871099621, + "flos": 15850431918720.0, + "grad_norm": 2.305730622299003, + "language_loss": 0.77195525, + "learning_rate": 1.0280122260055684e-06, + "loss": 0.79391837, + "num_input_tokens_seen": 120220780, + "step": 5586, + "time_per_iteration": 2.708860397338867 + }, + { + "auxiliary_loss_clip": 0.01172695, + "auxiliary_loss_mlp": 0.01026191, + "balance_loss_clip": 1.04902363, + "balance_loss_mlp": 1.01872289, + "epoch": 0.6717970300006012, + "flos": 19756112330880.0, + "grad_norm": 2.196955983375942, + "language_loss": 0.82103348, + "learning_rate": 1.0273315079664652e-06, + "loss": 0.84302229, + "num_input_tokens_seen": 120238735, + "step": 5587, + "time_per_iteration": 2.676482677459717 + }, + { + "auxiliary_loss_clip": 0.01183904, + "auxiliary_loss_mlp": 0.01024777, + "balance_loss_clip": 1.04856491, + "balance_loss_mlp": 1.0171566, + "epoch": 0.6719172728912403, + "flos": 25485049146240.0, + "grad_norm": 2.152106442332462, + "language_loss": 0.74278969, + "learning_rate": 1.0266509374803992e-06, + "loss": 0.76487648, + "num_input_tokens_seen": 120259895, + "step": 5588, + "time_per_iteration": 2.746321439743042 + }, + { + "auxiliary_loss_clip": 0.0117343, + "auxiliary_loss_mlp": 0.01378819, + "balance_loss_clip": 1.04964817, + "balance_loss_mlp": 1.00027394, + "epoch": 0.6720375157818794, + "flos": 15880344969600.0, + "grad_norm": 2.694539096164984, + "language_loss": 0.8436985, + "learning_rate": 1.0259705146506123e-06, + "loss": 0.86922097, + "num_input_tokens_seen": 120274790, + "step": 5589, + "time_per_iteration": 2.7270188331604004 + }, + { + "auxiliary_loss_clip": 0.01183482, + "auxiliary_loss_mlp": 0.01027776, + "balance_loss_clip": 1.04782462, + "balance_loss_mlp": 1.02073383, + "epoch": 0.6721577586725185, + "flos": 32010843231360.0, + "grad_norm": 3.3593688636181667, + "language_loss": 0.7741673, + "learning_rate": 1.025290239580324e-06, + "loss": 0.79627991, + "num_input_tokens_seen": 120295460, + "step": 5590, + "time_per_iteration": 2.783212184906006 + }, + { + "auxiliary_loss_clip": 0.01204893, + "auxiliary_loss_mlp": 0.01021698, + "balance_loss_clip": 1.04113674, + "balance_loss_mlp": 1.01402688, + "epoch": 0.6722780015631575, + "flos": 20737873837440.0, + "grad_norm": 1.7587613524092616, + "language_loss": 0.75551313, + "learning_rate": 1.0246101123727313e-06, + "loss": 0.7777791, + "num_input_tokens_seen": 120314440, + "step": 5591, + "time_per_iteration": 2.7405056953430176 + }, + { + "auxiliary_loss_clip": 0.01180596, + "auxiliary_loss_mlp": 0.01023194, + "balance_loss_clip": 1.04638743, + "balance_loss_mlp": 1.01585388, + "epoch": 0.6723982444537967, + "flos": 16909617191040.0, + "grad_norm": 1.8055249032342693, + "language_loss": 0.78591752, + "learning_rate": 1.0239301331310085e-06, + "loss": 0.80795544, + "num_input_tokens_seen": 120332060, + "step": 5592, + "time_per_iteration": 2.5893120765686035 + }, + { + "auxiliary_loss_clip": 0.01179831, + "auxiliary_loss_mlp": 0.0102412, + "balance_loss_clip": 1.04778016, + "balance_loss_mlp": 1.01720333, + "epoch": 0.6725184873444358, + "flos": 20667812359680.0, + "grad_norm": 2.1255144951640235, + "language_loss": 0.88489014, + "learning_rate": 1.0232503019583088e-06, + "loss": 0.90692961, + "num_input_tokens_seen": 120351670, + "step": 5593, + "time_per_iteration": 2.6749589443206787 + }, + { + "auxiliary_loss_clip": 0.01181901, + "auxiliary_loss_mlp": 0.01024694, + "balance_loss_clip": 1.05232191, + "balance_loss_mlp": 1.01735711, + "epoch": 0.6726387302350748, + "flos": 23727616416000.0, + "grad_norm": 1.8054142266255746, + "language_loss": 0.69997627, + "learning_rate": 1.0225706189577619e-06, + "loss": 0.7220422, + "num_input_tokens_seen": 120370195, + "step": 5594, + "time_per_iteration": 2.6319074630737305 + }, + { + "auxiliary_loss_clip": 0.0118388, + "auxiliary_loss_mlp": 0.01024049, + "balance_loss_clip": 1.04960966, + "balance_loss_mlp": 1.01630688, + "epoch": 0.672758973125714, + "flos": 15188274650880.0, + "grad_norm": 2.129833206721924, + "language_loss": 0.74818385, + "learning_rate": 1.021891084232475e-06, + "loss": 0.77026308, + "num_input_tokens_seen": 120388130, + "step": 5595, + "time_per_iteration": 2.5849056243896484 + }, + { + "auxiliary_loss_clip": 0.01178966, + "auxiliary_loss_mlp": 0.01026079, + "balance_loss_clip": 1.04557419, + "balance_loss_mlp": 1.01831889, + "epoch": 0.672879216016353, + "flos": 18077252601600.0, + "grad_norm": 3.128850320777895, + "language_loss": 0.80216169, + "learning_rate": 1.0212116978855325e-06, + "loss": 0.82421213, + "num_input_tokens_seen": 120406145, + "step": 5596, + "time_per_iteration": 2.5751876831054688 + }, + { + "auxiliary_loss_clip": 0.01197856, + "auxiliary_loss_mlp": 0.01024621, + "balance_loss_clip": 1.04514897, + "balance_loss_mlp": 1.01727152, + "epoch": 0.6729994589069921, + "flos": 23476349802240.0, + "grad_norm": 1.7021315511836745, + "language_loss": 0.79135734, + "learning_rate": 1.020532460019997e-06, + "loss": 0.81358206, + "num_input_tokens_seen": 120425395, + "step": 5597, + "time_per_iteration": 3.67409348487854 + }, + { + "auxiliary_loss_clip": 0.0122796, + "auxiliary_loss_mlp": 0.01026659, + "balance_loss_clip": 1.04097486, + "balance_loss_mlp": 1.01941156, + "epoch": 0.6731197017976313, + "flos": 26322018929280.0, + "grad_norm": 1.9420741501957315, + "language_loss": 0.71297818, + "learning_rate": 1.0198533707389096e-06, + "loss": 0.73552436, + "num_input_tokens_seen": 120446270, + "step": 5598, + "time_per_iteration": 4.059406518936157 + }, + { + "auxiliary_loss_clip": 0.01180507, + "auxiliary_loss_mlp": 0.01378772, + "balance_loss_clip": 1.04992974, + "balance_loss_mlp": 1.00024986, + "epoch": 0.6732399446882703, + "flos": 21616428591360.0, + "grad_norm": 2.7698881287957424, + "language_loss": 0.73366904, + "learning_rate": 1.0191744301452853e-06, + "loss": 0.75926179, + "num_input_tokens_seen": 120465570, + "step": 5599, + "time_per_iteration": 2.788445472717285 + }, + { + "auxiliary_loss_clip": 0.01168948, + "auxiliary_loss_mlp": 0.0102806, + "balance_loss_clip": 1.04748893, + "balance_loss_mlp": 1.02099705, + "epoch": 0.6733601875789094, + "flos": 25880173729920.0, + "grad_norm": 1.7937728459694735, + "language_loss": 0.70470798, + "learning_rate": 1.0184956383421208e-06, + "loss": 0.72667801, + "num_input_tokens_seen": 120484220, + "step": 5600, + "time_per_iteration": 2.6584348678588867 + }, + { + "auxiliary_loss_clip": 0.0118538, + "auxiliary_loss_mlp": 0.01023746, + "balance_loss_clip": 1.04926443, + "balance_loss_mlp": 1.01647401, + "epoch": 0.6734804304695485, + "flos": 22929573997440.0, + "grad_norm": 4.985881955781259, + "language_loss": 0.65783465, + "learning_rate": 1.017816995432387e-06, + "loss": 0.67992586, + "num_input_tokens_seen": 120503320, + "step": 5601, + "time_per_iteration": 3.5600199699401855 + }, + { + "auxiliary_loss_clip": 0.01189254, + "auxiliary_loss_mlp": 0.01030776, + "balance_loss_clip": 1.04745936, + "balance_loss_mlp": 1.02339435, + "epoch": 0.6736006733601876, + "flos": 18697968552960.0, + "grad_norm": 1.858177657441831, + "language_loss": 0.74294353, + "learning_rate": 1.0171385015190353e-06, + "loss": 0.76514375, + "num_input_tokens_seen": 120523180, + "step": 5602, + "time_per_iteration": 2.7216334342956543 + }, + { + "auxiliary_loss_clip": 0.01183111, + "auxiliary_loss_mlp": 0.01378466, + "balance_loss_clip": 1.04707932, + "balance_loss_mlp": 1.00020933, + "epoch": 0.6737209162508266, + "flos": 19427745173760.0, + "grad_norm": 5.2500026453324145, + "language_loss": 0.73343033, + "learning_rate": 1.0164601567049908e-06, + "loss": 0.75904614, + "num_input_tokens_seen": 120541710, + "step": 5603, + "time_per_iteration": 2.6876463890075684 + }, + { + "auxiliary_loss_clip": 0.01188464, + "auxiliary_loss_mlp": 0.0102442, + "balance_loss_clip": 1.0476265, + "balance_loss_mlp": 1.01666558, + "epoch": 0.6738411591414658, + "flos": 20158060498560.0, + "grad_norm": 1.7231320659329563, + "language_loss": 0.80576038, + "learning_rate": 1.015781961093158e-06, + "loss": 0.82788926, + "num_input_tokens_seen": 120561030, + "step": 5604, + "time_per_iteration": 4.3676371574401855 + }, + { + "auxiliary_loss_clip": 0.01191062, + "auxiliary_loss_mlp": 0.01024793, + "balance_loss_clip": 1.04376531, + "balance_loss_mlp": 1.01766741, + "epoch": 0.6739614020321049, + "flos": 21653847584640.0, + "grad_norm": 1.5114377976428666, + "language_loss": 0.77398837, + "learning_rate": 1.0151039147864197e-06, + "loss": 0.79614693, + "num_input_tokens_seen": 120581005, + "step": 5605, + "time_per_iteration": 2.7114672660827637 + }, + { + "auxiliary_loss_clip": 0.01220387, + "auxiliary_loss_mlp": 0.0102493, + "balance_loss_clip": 1.04544806, + "balance_loss_mlp": 1.0169909, + "epoch": 0.6740816449227439, + "flos": 19171702051200.0, + "grad_norm": 1.9652238327909748, + "language_loss": 0.65797502, + "learning_rate": 1.0144260178876336e-06, + "loss": 0.68042821, + "num_input_tokens_seen": 120600350, + "step": 5606, + "time_per_iteration": 3.0000882148742676 + }, + { + "auxiliary_loss_clip": 0.01192461, + "auxiliary_loss_mlp": 0.01027611, + "balance_loss_clip": 1.04496849, + "balance_loss_mlp": 1.01979733, + "epoch": 0.6742018878133831, + "flos": 21097015971840.0, + "grad_norm": 2.291158678996028, + "language_loss": 0.67298174, + "learning_rate": 1.0137482704996388e-06, + "loss": 0.69518244, + "num_input_tokens_seen": 120614700, + "step": 5607, + "time_per_iteration": 2.840780735015869 + }, + { + "auxiliary_loss_clip": 0.01202003, + "auxiliary_loss_mlp": 0.0102879, + "balance_loss_clip": 1.0460695, + "balance_loss_mlp": 1.02097559, + "epoch": 0.6743221307040221, + "flos": 23549966726400.0, + "grad_norm": 3.4533139443698238, + "language_loss": 0.78925467, + "learning_rate": 1.0130706727252461e-06, + "loss": 0.81156254, + "num_input_tokens_seen": 120631755, + "step": 5608, + "time_per_iteration": 2.867241859436035 + }, + { + "auxiliary_loss_clip": 0.01204372, + "auxiliary_loss_mlp": 0.01027066, + "balance_loss_clip": 1.04703724, + "balance_loss_mlp": 1.01935911, + "epoch": 0.6744423735946612, + "flos": 16249542912000.0, + "grad_norm": 2.376282282328894, + "language_loss": 0.67767102, + "learning_rate": 1.0123932246672468e-06, + "loss": 0.69998538, + "num_input_tokens_seen": 120645900, + "step": 5609, + "time_per_iteration": 2.6337826251983643 + }, + { + "auxiliary_loss_clip": 0.01114332, + "auxiliary_loss_mlp": 0.01373947, + "balance_loss_clip": 1.01160479, + "balance_loss_mlp": 0.99976122, + "epoch": 0.6745626164853004, + "flos": 57843257829120.0, + "grad_norm": 0.7491991824202496, + "language_loss": 0.55818194, + "learning_rate": 1.0117159264284114e-06, + "loss": 0.58306473, + "num_input_tokens_seen": 120709070, + "step": 5610, + "time_per_iteration": 3.477623701095581 + }, + { + "auxiliary_loss_clip": 0.01192046, + "auxiliary_loss_mlp": 0.01022787, + "balance_loss_clip": 1.04689956, + "balance_loss_mlp": 1.01519895, + "epoch": 0.6746828593759394, + "flos": 20485027025280.0, + "grad_norm": 1.6294176080168281, + "language_loss": 0.76885188, + "learning_rate": 1.0110387781114837e-06, + "loss": 0.79100025, + "num_input_tokens_seen": 120727685, + "step": 5611, + "time_per_iteration": 2.665667772293091 + }, + { + "auxiliary_loss_clip": 0.01169451, + "auxiliary_loss_mlp": 0.01026338, + "balance_loss_clip": 1.04843974, + "balance_loss_mlp": 1.01843476, + "epoch": 0.6748031022665785, + "flos": 19208223204480.0, + "grad_norm": 1.9887055686739155, + "language_loss": 0.77026135, + "learning_rate": 1.0103617798191872e-06, + "loss": 0.79221928, + "num_input_tokens_seen": 120747160, + "step": 5612, + "time_per_iteration": 2.6115028858184814 + }, + { + "auxiliary_loss_clip": 0.01185748, + "auxiliary_loss_mlp": 0.01022582, + "balance_loss_clip": 1.04514456, + "balance_loss_mlp": 1.01508939, + "epoch": 0.6749233451572175, + "flos": 15195026407680.0, + "grad_norm": 2.923734106937653, + "language_loss": 0.82790595, + "learning_rate": 1.0096849316542217e-06, + "loss": 0.84998924, + "num_input_tokens_seen": 120763710, + "step": 5613, + "time_per_iteration": 2.6426401138305664 + }, + { + "auxiliary_loss_clip": 0.01210767, + "auxiliary_loss_mlp": 0.01018993, + "balance_loss_clip": 1.0384109, + "balance_loss_mlp": 1.01135719, + "epoch": 0.6750435880478567, + "flos": 26499489050880.0, + "grad_norm": 2.06671238932974, + "language_loss": 0.74973106, + "learning_rate": 1.0090082337192643e-06, + "loss": 0.77202857, + "num_input_tokens_seen": 120783355, + "step": 5614, + "time_per_iteration": 2.8398380279541016 + }, + { + "auxiliary_loss_clip": 0.01206489, + "auxiliary_loss_mlp": 0.01023674, + "balance_loss_clip": 1.03734267, + "balance_loss_mlp": 1.01649785, + "epoch": 0.6751638309384957, + "flos": 23404313076480.0, + "grad_norm": 2.7032546564427187, + "language_loss": 0.78901267, + "learning_rate": 1.0083316861169705e-06, + "loss": 0.81131434, + "num_input_tokens_seen": 120802090, + "step": 5615, + "time_per_iteration": 2.850677013397217 + }, + { + "auxiliary_loss_clip": 0.01203701, + "auxiliary_loss_mlp": 0.01026978, + "balance_loss_clip": 1.04452336, + "balance_loss_mlp": 1.01868081, + "epoch": 0.6752840738291348, + "flos": 23441408847360.0, + "grad_norm": 1.9787219156260676, + "language_loss": 0.7160939, + "learning_rate": 1.0076552889499713e-06, + "loss": 0.7384007, + "num_input_tokens_seen": 120822855, + "step": 5616, + "time_per_iteration": 2.825817346572876 + }, + { + "auxiliary_loss_clip": 0.01180328, + "auxiliary_loss_mlp": 0.01025027, + "balance_loss_clip": 1.04940438, + "balance_loss_mlp": 1.01776075, + "epoch": 0.675404316719774, + "flos": 30335826257280.0, + "grad_norm": 1.8785160482728827, + "language_loss": 0.73389363, + "learning_rate": 1.006979042320876e-06, + "loss": 0.75594717, + "num_input_tokens_seen": 120843070, + "step": 5617, + "time_per_iteration": 2.7055158615112305 + }, + { + "auxiliary_loss_clip": 0.01182891, + "auxiliary_loss_mlp": 0.0102244, + "balance_loss_clip": 1.04273224, + "balance_loss_mlp": 1.01523042, + "epoch": 0.675524559610413, + "flos": 23622613983360.0, + "grad_norm": 2.020052852617401, + "language_loss": 0.63028288, + "learning_rate": 1.0063029463322702e-06, + "loss": 0.65233624, + "num_input_tokens_seen": 120863345, + "step": 5618, + "time_per_iteration": 2.66094708442688 + }, + { + "auxiliary_loss_clip": 0.01198682, + "auxiliary_loss_mlp": 0.01378743, + "balance_loss_clip": 1.04075193, + "balance_loss_mlp": 1.00020123, + "epoch": 0.6756448025010521, + "flos": 21248631279360.0, + "grad_norm": 2.9728205909375136, + "language_loss": 0.75177169, + "learning_rate": 1.0056270010867164e-06, + "loss": 0.77754593, + "num_input_tokens_seen": 120880915, + "step": 5619, + "time_per_iteration": 2.799128293991089 + }, + { + "auxiliary_loss_clip": 0.0119071, + "auxiliary_loss_mlp": 0.01026333, + "balance_loss_clip": 1.04360271, + "balance_loss_mlp": 1.01831603, + "epoch": 0.6757650453916912, + "flos": 21646521210240.0, + "grad_norm": 2.6873677722723723, + "language_loss": 0.77562982, + "learning_rate": 1.004951206686758e-06, + "loss": 0.79780018, + "num_input_tokens_seen": 120899190, + "step": 5620, + "time_per_iteration": 2.721615791320801 + }, + { + "auxiliary_loss_clip": 0.0117326, + "auxiliary_loss_mlp": 0.01030043, + "balance_loss_clip": 1.04542696, + "balance_loss_mlp": 1.02233267, + "epoch": 0.6758852882823303, + "flos": 21795658479360.0, + "grad_norm": 2.088830115200652, + "language_loss": 0.71686721, + "learning_rate": 1.0042755632349087e-06, + "loss": 0.73890018, + "num_input_tokens_seen": 120916080, + "step": 5621, + "time_per_iteration": 2.6302096843719482 + }, + { + "auxiliary_loss_clip": 0.01201761, + "auxiliary_loss_mlp": 0.0102421, + "balance_loss_clip": 1.0469017, + "balance_loss_mlp": 1.01672363, + "epoch": 0.6760055311729694, + "flos": 27088783580160.0, + "grad_norm": 1.9371161974255546, + "language_loss": 0.62955821, + "learning_rate": 1.0036000708336653e-06, + "loss": 0.65181792, + "num_input_tokens_seen": 120935210, + "step": 5622, + "time_per_iteration": 2.740593671798706 + }, + { + "auxiliary_loss_clip": 0.01192195, + "auxiliary_loss_mlp": 0.01028191, + "balance_loss_clip": 1.04671049, + "balance_loss_mlp": 1.0208863, + "epoch": 0.6761257740636085, + "flos": 17999792922240.0, + "grad_norm": 2.5021395000283526, + "language_loss": 0.79743838, + "learning_rate": 1.0029247295854984e-06, + "loss": 0.81964225, + "num_input_tokens_seen": 120951830, + "step": 5623, + "time_per_iteration": 3.5861270427703857 + }, + { + "auxiliary_loss_clip": 0.01205972, + "auxiliary_loss_mlp": 0.01027787, + "balance_loss_clip": 1.0478344, + "balance_loss_mlp": 1.02055073, + "epoch": 0.6762460169542476, + "flos": 15121912273920.0, + "grad_norm": 1.8453471288824141, + "language_loss": 0.71549046, + "learning_rate": 1.0022495395928588e-06, + "loss": 0.73782802, + "num_input_tokens_seen": 120970310, + "step": 5624, + "time_per_iteration": 3.772711992263794 + }, + { + "auxiliary_loss_clip": 0.01066015, + "auxiliary_loss_mlp": 0.01003313, + "balance_loss_clip": 1.0117445, + "balance_loss_mlp": 1.00210321, + "epoch": 0.6763662598448866, + "flos": 67886970030720.0, + "grad_norm": 0.795883694591083, + "language_loss": 0.62359744, + "learning_rate": 1.0015745009581697e-06, + "loss": 0.64429075, + "num_input_tokens_seen": 121031915, + "step": 5625, + "time_per_iteration": 3.1977176666259766 + }, + { + "auxiliary_loss_clip": 0.01179485, + "auxiliary_loss_mlp": 0.01028216, + "balance_loss_clip": 1.04864442, + "balance_loss_mlp": 1.02059817, + "epoch": 0.6764865027355258, + "flos": 20631829910400.0, + "grad_norm": 2.018901780196003, + "language_loss": 0.66956151, + "learning_rate": 1.0008996137838343e-06, + "loss": 0.69163859, + "num_input_tokens_seen": 121050890, + "step": 5626, + "time_per_iteration": 2.5882606506347656 + }, + { + "auxiliary_loss_clip": 0.01175642, + "auxiliary_loss_mlp": 0.01028567, + "balance_loss_clip": 1.04885232, + "balance_loss_mlp": 1.01998997, + "epoch": 0.6766067456261649, + "flos": 21215809226880.0, + "grad_norm": 2.022484675487177, + "language_loss": 0.80344957, + "learning_rate": 1.000224878172234e-06, + "loss": 0.82549167, + "num_input_tokens_seen": 121070015, + "step": 5627, + "time_per_iteration": 3.629523992538452 + }, + { + "auxiliary_loss_clip": 0.01183834, + "auxiliary_loss_mlp": 0.0102351, + "balance_loss_clip": 1.04807568, + "balance_loss_mlp": 1.01591671, + "epoch": 0.6767269885168039, + "flos": 19938251220480.0, + "grad_norm": 2.3137664600440613, + "language_loss": 0.72538018, + "learning_rate": 9.99550294225724e-07, + "loss": 0.74745363, + "num_input_tokens_seen": 121089170, + "step": 5628, + "time_per_iteration": 2.6127681732177734 + }, + { + "auxiliary_loss_clip": 0.01206687, + "auxiliary_loss_mlp": 0.01031078, + "balance_loss_clip": 1.04018283, + "balance_loss_mlp": 1.02268577, + "epoch": 0.6768472314074431, + "flos": 20814076540800.0, + "grad_norm": 1.9908274060770514, + "language_loss": 0.72339857, + "learning_rate": 9.988758620466402e-07, + "loss": 0.74577618, + "num_input_tokens_seen": 121108040, + "step": 5629, + "time_per_iteration": 2.7643136978149414 + }, + { + "auxiliary_loss_clip": 0.01222989, + "auxiliary_loss_mlp": 0.01026163, + "balance_loss_clip": 1.04267251, + "balance_loss_mlp": 1.01890278, + "epoch": 0.6769674742980821, + "flos": 23186012169600.0, + "grad_norm": 1.667149461899024, + "language_loss": 0.76265502, + "learning_rate": 9.982015817372917e-07, + "loss": 0.78514659, + "num_input_tokens_seen": 121128480, + "step": 5630, + "time_per_iteration": 3.7016634941101074 + }, + { + "auxiliary_loss_clip": 0.01202677, + "auxiliary_loss_mlp": 0.01031341, + "balance_loss_clip": 1.04099298, + "balance_loss_mlp": 1.02363992, + "epoch": 0.6770877171887212, + "flos": 24242934885120.0, + "grad_norm": 1.8786959661570872, + "language_loss": 0.82170945, + "learning_rate": 9.975274533999657e-07, + "loss": 0.84404957, + "num_input_tokens_seen": 121148010, + "step": 5631, + "time_per_iteration": 2.8289647102355957 + }, + { + "auxiliary_loss_clip": 0.01172935, + "auxiliary_loss_mlp": 0.01024466, + "balance_loss_clip": 1.04846251, + "balance_loss_mlp": 1.01602626, + "epoch": 0.6772079600793603, + "flos": 18141567903360.0, + "grad_norm": 3.2708681892188083, + "language_loss": 0.84295332, + "learning_rate": 9.96853477136929e-07, + "loss": 0.86492741, + "num_input_tokens_seen": 121162755, + "step": 5632, + "time_per_iteration": 2.530921220779419 + }, + { + "auxiliary_loss_clip": 0.01187809, + "auxiliary_loss_mlp": 0.01026464, + "balance_loss_clip": 1.04107177, + "balance_loss_mlp": 1.01828671, + "epoch": 0.6773282029699994, + "flos": 22452069571200.0, + "grad_norm": 2.030161350782985, + "language_loss": 0.75165939, + "learning_rate": 9.96179653050422e-07, + "loss": 0.7738021, + "num_input_tokens_seen": 121182915, + "step": 5633, + "time_per_iteration": 2.6771397590637207 + }, + { + "auxiliary_loss_clip": 0.01191895, + "auxiliary_loss_mlp": 0.01029752, + "balance_loss_clip": 1.04554462, + "balance_loss_mlp": 1.02208054, + "epoch": 0.6774484458606385, + "flos": 18693730748160.0, + "grad_norm": 2.587360885812055, + "language_loss": 0.74068165, + "learning_rate": 9.955059812426635e-07, + "loss": 0.76289809, + "num_input_tokens_seen": 121200445, + "step": 5634, + "time_per_iteration": 2.703660011291504 + }, + { + "auxiliary_loss_clip": 0.01174052, + "auxiliary_loss_mlp": 0.01027131, + "balance_loss_clip": 1.0517621, + "balance_loss_mlp": 1.01924849, + "epoch": 0.6775686887512776, + "flos": 25994046821760.0, + "grad_norm": 2.2597679814614584, + "language_loss": 0.82867754, + "learning_rate": 9.948324618158493e-07, + "loss": 0.85068941, + "num_input_tokens_seen": 121220785, + "step": 5635, + "time_per_iteration": 2.578934907913208 + }, + { + "auxiliary_loss_clip": 0.01178666, + "auxiliary_loss_mlp": 0.01023061, + "balance_loss_clip": 1.04419875, + "balance_loss_mlp": 1.01540184, + "epoch": 0.6776889316419167, + "flos": 13587987922560.0, + "grad_norm": 5.0420157439473305, + "language_loss": 0.77170706, + "learning_rate": 9.941590948721502e-07, + "loss": 0.79372436, + "num_input_tokens_seen": 121237985, + "step": 5636, + "time_per_iteration": 2.630981683731079 + }, + { + "auxiliary_loss_clip": 0.01182849, + "auxiliary_loss_mlp": 0.01023088, + "balance_loss_clip": 1.04558325, + "balance_loss_mlp": 1.01600456, + "epoch": 0.6778091745325557, + "flos": 27601121220480.0, + "grad_norm": 1.7663422959899524, + "language_loss": 0.76294231, + "learning_rate": 9.934858805137188e-07, + "loss": 0.78500164, + "num_input_tokens_seen": 121258635, + "step": 5637, + "time_per_iteration": 2.677887201309204 + }, + { + "auxiliary_loss_clip": 0.01176816, + "auxiliary_loss_mlp": 0.01025876, + "balance_loss_clip": 1.04853868, + "balance_loss_mlp": 1.01869392, + "epoch": 0.6779294174231949, + "flos": 18734058743040.0, + "grad_norm": 1.7881773309549198, + "language_loss": 0.81141508, + "learning_rate": 9.92812818842677e-07, + "loss": 0.83344197, + "num_input_tokens_seen": 121277810, + "step": 5638, + "time_per_iteration": 2.66843318939209 + }, + { + "auxiliary_loss_clip": 0.01176286, + "auxiliary_loss_mlp": 0.01025051, + "balance_loss_clip": 1.04523182, + "balance_loss_mlp": 1.01741004, + "epoch": 0.678049660313834, + "flos": 45873797765760.0, + "grad_norm": 2.070488386358795, + "language_loss": 0.6414367, + "learning_rate": 9.921399099611306e-07, + "loss": 0.66345006, + "num_input_tokens_seen": 121298975, + "step": 5639, + "time_per_iteration": 2.847949266433716 + }, + { + "auxiliary_loss_clip": 0.01188994, + "auxiliary_loss_mlp": 0.01024028, + "balance_loss_clip": 1.04468012, + "balance_loss_mlp": 1.01709867, + "epoch": 0.678169903204473, + "flos": 19974556892160.0, + "grad_norm": 1.7357478338867354, + "language_loss": 0.68759567, + "learning_rate": 9.914671539711588e-07, + "loss": 0.70972586, + "num_input_tokens_seen": 121318495, + "step": 5640, + "time_per_iteration": 2.7062883377075195 + }, + { + "auxiliary_loss_clip": 0.01238076, + "auxiliary_loss_mlp": 0.01379158, + "balance_loss_clip": 1.04114294, + "balance_loss_mlp": 1.00016642, + "epoch": 0.6782901460951122, + "flos": 21395613732480.0, + "grad_norm": 2.043984736508797, + "language_loss": 0.78366792, + "learning_rate": 9.90794550974817e-07, + "loss": 0.80984026, + "num_input_tokens_seen": 121338890, + "step": 5641, + "time_per_iteration": 2.949956178665161 + }, + { + "auxiliary_loss_clip": 0.01193497, + "auxiliary_loss_mlp": 0.01029409, + "balance_loss_clip": 1.04503131, + "balance_loss_mlp": 1.02132618, + "epoch": 0.6784103889857512, + "flos": 21434002392960.0, + "grad_norm": 2.4479008232618136, + "language_loss": 0.81476319, + "learning_rate": 9.901221010741407e-07, + "loss": 0.83699226, + "num_input_tokens_seen": 121358210, + "step": 5642, + "time_per_iteration": 2.9055862426757812 + }, + { + "auxiliary_loss_clip": 0.01183557, + "auxiliary_loss_mlp": 0.01029976, + "balance_loss_clip": 1.04716873, + "balance_loss_mlp": 1.02177477, + "epoch": 0.6785306318763903, + "flos": 32671923091200.0, + "grad_norm": 2.274656990748534, + "language_loss": 0.75079894, + "learning_rate": 9.894498043711375e-07, + "loss": 0.7729342, + "num_input_tokens_seen": 121379955, + "step": 5643, + "time_per_iteration": 2.760272741317749 + }, + { + "auxiliary_loss_clip": 0.01187959, + "auxiliary_loss_mlp": 0.01023594, + "balance_loss_clip": 1.04496479, + "balance_loss_mlp": 1.01634359, + "epoch": 0.6786508747670293, + "flos": 25632139340160.0, + "grad_norm": 2.0040710185298507, + "language_loss": 0.68800008, + "learning_rate": 9.887776609677962e-07, + "loss": 0.71011561, + "num_input_tokens_seen": 121401325, + "step": 5644, + "time_per_iteration": 2.6752638816833496 + }, + { + "auxiliary_loss_clip": 0.01184626, + "auxiliary_loss_mlp": 0.01027395, + "balance_loss_clip": 1.03953862, + "balance_loss_mlp": 1.02025151, + "epoch": 0.6787711176576685, + "flos": 19171881619200.0, + "grad_norm": 2.1751117654455188, + "language_loss": 0.72227848, + "learning_rate": 9.88105670966079e-07, + "loss": 0.74439871, + "num_input_tokens_seen": 121419785, + "step": 5645, + "time_per_iteration": 2.7266671657562256 + }, + { + "auxiliary_loss_clip": 0.01194405, + "auxiliary_loss_mlp": 0.010272, + "balance_loss_clip": 1.04360306, + "balance_loss_mlp": 1.01978528, + "epoch": 0.6788913605483076, + "flos": 13985159581440.0, + "grad_norm": 2.0037779594102476, + "language_loss": 0.79004174, + "learning_rate": 9.874338344679283e-07, + "loss": 0.81225783, + "num_input_tokens_seen": 121435630, + "step": 5646, + "time_per_iteration": 2.691558599472046 + }, + { + "auxiliary_loss_clip": 0.01168964, + "auxiliary_loss_mlp": 0.01028042, + "balance_loss_clip": 1.04814315, + "balance_loss_mlp": 1.02096963, + "epoch": 0.6790116034389466, + "flos": 22017586659840.0, + "grad_norm": 1.8817304165094983, + "language_loss": 0.74035633, + "learning_rate": 9.86762151575259e-07, + "loss": 0.76232642, + "num_input_tokens_seen": 121455625, + "step": 5647, + "time_per_iteration": 2.62451171875 + }, + { + "auxiliary_loss_clip": 0.01209856, + "auxiliary_loss_mlp": 0.01378321, + "balance_loss_clip": 1.04583526, + "balance_loss_mlp": 1.0001241, + "epoch": 0.6791318463295858, + "flos": 20922454851840.0, + "grad_norm": 1.4818671075585732, + "language_loss": 0.80254567, + "learning_rate": 9.860906223899651e-07, + "loss": 0.82842743, + "num_input_tokens_seen": 121475020, + "step": 5648, + "time_per_iteration": 3.731391668319702 + }, + { + "auxiliary_loss_clip": 0.01196368, + "auxiliary_loss_mlp": 0.01027648, + "balance_loss_clip": 1.04752684, + "balance_loss_mlp": 1.01997709, + "epoch": 0.6792520892202248, + "flos": 28512749422080.0, + "grad_norm": 1.7128086688868056, + "language_loss": 0.75519526, + "learning_rate": 9.854192470139184e-07, + "loss": 0.77743542, + "num_input_tokens_seen": 121496500, + "step": 5649, + "time_per_iteration": 2.731224775314331 + }, + { + "auxiliary_loss_clip": 0.01191198, + "auxiliary_loss_mlp": 0.01026437, + "balance_loss_clip": 1.04966986, + "balance_loss_mlp": 1.01882529, + "epoch": 0.6793723321108639, + "flos": 20011904058240.0, + "grad_norm": 2.9773093116063123, + "language_loss": 0.71591544, + "learning_rate": 9.847480255489645e-07, + "loss": 0.73809171, + "num_input_tokens_seen": 121515525, + "step": 5650, + "time_per_iteration": 3.5794694423675537 + }, + { + "auxiliary_loss_clip": 0.01193591, + "auxiliary_loss_mlp": 0.01031278, + "balance_loss_clip": 1.0462625, + "balance_loss_mlp": 1.02406907, + "epoch": 0.6794925750015031, + "flos": 26649488246400.0, + "grad_norm": 2.197456599322004, + "language_loss": 0.69171065, + "learning_rate": 9.840769580969295e-07, + "loss": 0.71395934, + "num_input_tokens_seen": 121535965, + "step": 5651, + "time_per_iteration": 2.720395565032959 + }, + { + "auxiliary_loss_clip": 0.01169623, + "auxiliary_loss_mlp": 0.01023374, + "balance_loss_clip": 1.04439878, + "balance_loss_mlp": 1.01568532, + "epoch": 0.6796128178921421, + "flos": 21580374314880.0, + "grad_norm": 1.9051298255235003, + "language_loss": 0.79856789, + "learning_rate": 9.834060447596114e-07, + "loss": 0.82049787, + "num_input_tokens_seen": 121555235, + "step": 5652, + "time_per_iteration": 3.005675792694092 + }, + { + "auxiliary_loss_clip": 0.01182014, + "auxiliary_loss_mlp": 0.01025603, + "balance_loss_clip": 1.04635596, + "balance_loss_mlp": 1.01794934, + "epoch": 0.6797330607827812, + "flos": 22492002516480.0, + "grad_norm": 2.052999467215616, + "language_loss": 0.78408343, + "learning_rate": 9.827352856387868e-07, + "loss": 0.80615962, + "num_input_tokens_seen": 121574945, + "step": 5653, + "time_per_iteration": 4.031461000442505 + }, + { + "auxiliary_loss_clip": 0.01118608, + "auxiliary_loss_mlp": 0.01004393, + "balance_loss_clip": 1.01972222, + "balance_loss_mlp": 1.00318897, + "epoch": 0.6798533036734203, + "flos": 66306648286080.0, + "grad_norm": 0.7778884703498197, + "language_loss": 0.64243501, + "learning_rate": 9.820646808362118e-07, + "loss": 0.66366506, + "num_input_tokens_seen": 121641200, + "step": 5654, + "time_per_iteration": 3.396028757095337 + }, + { + "auxiliary_loss_clip": 0.01188988, + "auxiliary_loss_mlp": 0.01024487, + "balance_loss_clip": 1.04889679, + "balance_loss_mlp": 1.01669431, + "epoch": 0.6799735465640594, + "flos": 16180163792640.0, + "grad_norm": 2.204360519613285, + "language_loss": 0.72685945, + "learning_rate": 9.813942304536154e-07, + "loss": 0.74899423, + "num_input_tokens_seen": 121659170, + "step": 5655, + "time_per_iteration": 2.6354565620422363 + }, + { + "auxiliary_loss_clip": 0.01189401, + "auxiliary_loss_mlp": 0.01021319, + "balance_loss_clip": 1.04547191, + "balance_loss_mlp": 1.01405907, + "epoch": 0.6800937894546984, + "flos": 22125749489280.0, + "grad_norm": 2.671900831853445, + "language_loss": 0.63504839, + "learning_rate": 9.807239345927043e-07, + "loss": 0.65715563, + "num_input_tokens_seen": 121679180, + "step": 5656, + "time_per_iteration": 3.8331375122070312 + }, + { + "auxiliary_loss_clip": 0.01190325, + "auxiliary_loss_mlp": 0.0102871, + "balance_loss_clip": 1.04291761, + "balance_loss_mlp": 1.02121139, + "epoch": 0.6802140323453376, + "flos": 31612953300480.0, + "grad_norm": 2.176514057745871, + "language_loss": 0.7161237, + "learning_rate": 9.80053793355162e-07, + "loss": 0.73831403, + "num_input_tokens_seen": 121697875, + "step": 5657, + "time_per_iteration": 2.7112109661102295 + }, + { + "auxiliary_loss_clip": 0.01198816, + "auxiliary_loss_mlp": 0.01034649, + "balance_loss_clip": 1.04011595, + "balance_loss_mlp": 1.026811, + "epoch": 0.6803342752359767, + "flos": 17712938908800.0, + "grad_norm": 2.5701608326262777, + "language_loss": 0.74994212, + "learning_rate": 9.793838068426472e-07, + "loss": 0.77227676, + "num_input_tokens_seen": 121715570, + "step": 5658, + "time_per_iteration": 2.7659366130828857 + }, + { + "auxiliary_loss_clip": 0.01172476, + "auxiliary_loss_mlp": 0.01025796, + "balance_loss_clip": 1.04952621, + "balance_loss_mlp": 1.01818168, + "epoch": 0.6804545181266157, + "flos": 11326800902400.0, + "grad_norm": 2.1602872513189832, + "language_loss": 0.60939169, + "learning_rate": 9.78713975156799e-07, + "loss": 0.63137448, + "num_input_tokens_seen": 121731435, + "step": 5659, + "time_per_iteration": 2.5684120655059814 + }, + { + "auxiliary_loss_clip": 0.01200171, + "auxiliary_loss_mlp": 0.0103138, + "balance_loss_clip": 1.0473696, + "balance_loss_mlp": 1.02294922, + "epoch": 0.6805747610172549, + "flos": 29350976181120.0, + "grad_norm": 1.7318313352729604, + "language_loss": 0.72000027, + "learning_rate": 9.780442983992273e-07, + "loss": 0.74231583, + "num_input_tokens_seen": 121749950, + "step": 5660, + "time_per_iteration": 2.8334004878997803 + }, + { + "auxiliary_loss_clip": 0.01179205, + "auxiliary_loss_mlp": 0.0103059, + "balance_loss_clip": 1.04468119, + "balance_loss_mlp": 1.0223825, + "epoch": 0.680695003907894, + "flos": 37631868612480.0, + "grad_norm": 1.6342979431084312, + "language_loss": 0.71958941, + "learning_rate": 9.773747766715238e-07, + "loss": 0.74168736, + "num_input_tokens_seen": 121770770, + "step": 5661, + "time_per_iteration": 2.804366111755371 + }, + { + "auxiliary_loss_clip": 0.01189138, + "auxiliary_loss_mlp": 0.01027109, + "balance_loss_clip": 1.04364371, + "balance_loss_mlp": 1.01935482, + "epoch": 0.680815246798533, + "flos": 22127365601280.0, + "grad_norm": 1.673823114762803, + "language_loss": 0.8045814, + "learning_rate": 9.767054100752536e-07, + "loss": 0.82674384, + "num_input_tokens_seen": 121790720, + "step": 5662, + "time_per_iteration": 2.6647353172302246 + }, + { + "auxiliary_loss_clip": 0.01200289, + "auxiliary_loss_mlp": 0.01032278, + "balance_loss_clip": 1.04499805, + "balance_loss_mlp": 1.02419519, + "epoch": 0.6809354896891722, + "flos": 17201822330880.0, + "grad_norm": 2.9214747116996893, + "language_loss": 0.8192488, + "learning_rate": 9.760361987119584e-07, + "loss": 0.84157449, + "num_input_tokens_seen": 121808455, + "step": 5663, + "time_per_iteration": 2.721825361251831 + }, + { + "auxiliary_loss_clip": 0.01191155, + "auxiliary_loss_mlp": 0.01029855, + "balance_loss_clip": 1.04875624, + "balance_loss_mlp": 1.02178431, + "epoch": 0.6810557325798112, + "flos": 12458166554880.0, + "grad_norm": 2.2605061220807707, + "language_loss": 0.67418563, + "learning_rate": 9.753671426831592e-07, + "loss": 0.69639575, + "num_input_tokens_seen": 121824470, + "step": 5664, + "time_per_iteration": 2.6234750747680664 + }, + { + "auxiliary_loss_clip": 0.01171859, + "auxiliary_loss_mlp": 0.01029324, + "balance_loss_clip": 1.04368472, + "balance_loss_mlp": 1.02161407, + "epoch": 0.6811759754704503, + "flos": 22156165330560.0, + "grad_norm": 1.8509652771449014, + "language_loss": 0.79707229, + "learning_rate": 9.746982420903483e-07, + "loss": 0.81908417, + "num_input_tokens_seen": 121842665, + "step": 5665, + "time_per_iteration": 2.6061785221099854 + }, + { + "auxiliary_loss_clip": 0.01175832, + "auxiliary_loss_mlp": 0.01029982, + "balance_loss_clip": 1.0472672, + "balance_loss_mlp": 1.02317524, + "epoch": 0.6812962183610894, + "flos": 17525377065600.0, + "grad_norm": 2.4671598014519387, + "language_loss": 0.74754417, + "learning_rate": 9.740294970349993e-07, + "loss": 0.7696023, + "num_input_tokens_seen": 121859080, + "step": 5666, + "time_per_iteration": 2.5614728927612305 + }, + { + "auxiliary_loss_clip": 0.01089286, + "auxiliary_loss_mlp": 0.0099914, + "balance_loss_clip": 1.01059842, + "balance_loss_mlp": 0.99800169, + "epoch": 0.6814164612517285, + "flos": 60274480855680.0, + "grad_norm": 0.9050246606564024, + "language_loss": 0.60935557, + "learning_rate": 9.733609076185594e-07, + "loss": 0.63023984, + "num_input_tokens_seen": 121915485, + "step": 5667, + "time_per_iteration": 3.1213529109954834 + }, + { + "auxiliary_loss_clip": 0.01181179, + "auxiliary_loss_mlp": 0.01023094, + "balance_loss_clip": 1.04890537, + "balance_loss_mlp": 1.01562881, + "epoch": 0.6815367041423676, + "flos": 19317750750720.0, + "grad_norm": 1.79764040560632, + "language_loss": 0.83935058, + "learning_rate": 9.72692473942455e-07, + "loss": 0.86139333, + "num_input_tokens_seen": 121932710, + "step": 5668, + "time_per_iteration": 2.568202018737793 + }, + { + "auxiliary_loss_clip": 0.01212122, + "auxiliary_loss_mlp": 0.01030987, + "balance_loss_clip": 1.04560232, + "balance_loss_mlp": 1.02355456, + "epoch": 0.6816569470330067, + "flos": 22161696024960.0, + "grad_norm": 1.9183397195952057, + "language_loss": 0.7746774, + "learning_rate": 9.720241961080849e-07, + "loss": 0.79710841, + "num_input_tokens_seen": 121952025, + "step": 5669, + "time_per_iteration": 2.742709159851074 + }, + { + "auxiliary_loss_clip": 0.01170113, + "auxiliary_loss_mlp": 0.01027835, + "balance_loss_clip": 1.047189, + "balance_loss_mlp": 1.02067339, + "epoch": 0.6817771899236458, + "flos": 41463501137280.0, + "grad_norm": 2.5534546118127963, + "language_loss": 0.73421633, + "learning_rate": 9.713560742168259e-07, + "loss": 0.75619578, + "num_input_tokens_seen": 121974650, + "step": 5670, + "time_per_iteration": 2.7021334171295166 + }, + { + "auxiliary_loss_clip": 0.01194538, + "auxiliary_loss_mlp": 0.01023938, + "balance_loss_clip": 1.04379928, + "balance_loss_mlp": 1.01694345, + "epoch": 0.6818974328142848, + "flos": 21106138026240.0, + "grad_norm": 1.8821846598600183, + "language_loss": 0.71095705, + "learning_rate": 9.706881083700333e-07, + "loss": 0.73314178, + "num_input_tokens_seen": 121994335, + "step": 5671, + "time_per_iteration": 2.6853623390197754 + }, + { + "auxiliary_loss_clip": 0.0121374, + "auxiliary_loss_mlp": 0.01023266, + "balance_loss_clip": 1.04357135, + "balance_loss_mlp": 1.01526153, + "epoch": 0.682017675704924, + "flos": 20441897769600.0, + "grad_norm": 2.129906351727666, + "language_loss": 0.82869422, + "learning_rate": 9.700202986690357e-07, + "loss": 0.85106432, + "num_input_tokens_seen": 122012635, + "step": 5672, + "time_per_iteration": 2.7425990104675293 + }, + { + "auxiliary_loss_clip": 0.01178484, + "auxiliary_loss_mlp": 0.01379241, + "balance_loss_clip": 1.04645884, + "balance_loss_mlp": 1.00014305, + "epoch": 0.682137918595563, + "flos": 20044438801920.0, + "grad_norm": 2.7362867301171407, + "language_loss": 0.66703153, + "learning_rate": 9.693526452151413e-07, + "loss": 0.69260877, + "num_input_tokens_seen": 122031685, + "step": 5673, + "time_per_iteration": 2.670016050338745 + }, + { + "auxiliary_loss_clip": 0.01200532, + "auxiliary_loss_mlp": 0.01021849, + "balance_loss_clip": 1.0433774, + "balance_loss_mlp": 1.01353359, + "epoch": 0.6822581614862021, + "flos": 31684559063040.0, + "grad_norm": 1.7441556769059992, + "language_loss": 0.75509107, + "learning_rate": 9.686851481096305e-07, + "loss": 0.77731484, + "num_input_tokens_seen": 122052995, + "step": 5674, + "time_per_iteration": 3.7243075370788574 + }, + { + "auxiliary_loss_clip": 0.01211087, + "auxiliary_loss_mlp": 0.01033037, + "balance_loss_clip": 1.03996789, + "balance_loss_mlp": 1.0249784, + "epoch": 0.6823784043768413, + "flos": 23477570864640.0, + "grad_norm": 2.572684733096164, + "language_loss": 0.7204985, + "learning_rate": 9.68017807453762e-07, + "loss": 0.74293971, + "num_input_tokens_seen": 122071740, + "step": 5675, + "time_per_iteration": 2.8292689323425293 + }, + { + "auxiliary_loss_clip": 0.01196801, + "auxiliary_loss_mlp": 0.01378753, + "balance_loss_clip": 1.05058527, + "balance_loss_mlp": 1.00018406, + "epoch": 0.6824986472674803, + "flos": 14137134024960.0, + "grad_norm": 1.780159112677392, + "language_loss": 0.73301399, + "learning_rate": 9.673506233487721e-07, + "loss": 0.75876945, + "num_input_tokens_seen": 122089705, + "step": 5676, + "time_per_iteration": 3.507559061050415 + }, + { + "auxiliary_loss_clip": 0.01188606, + "auxiliary_loss_mlp": 0.01378374, + "balance_loss_clip": 1.04487741, + "balance_loss_mlp": 1.00014949, + "epoch": 0.6826188901581194, + "flos": 21504997624320.0, + "grad_norm": 1.6574001416063688, + "language_loss": 0.86066175, + "learning_rate": 9.666835958958717e-07, + "loss": 0.88633156, + "num_input_tokens_seen": 122109025, + "step": 5677, + "time_per_iteration": 2.578153610229492 + }, + { + "auxiliary_loss_clip": 0.01170001, + "auxiliary_loss_mlp": 0.01028199, + "balance_loss_clip": 1.04842103, + "balance_loss_mlp": 1.02113259, + "epoch": 0.6827391330487584, + "flos": 20810126044800.0, + "grad_norm": 1.9974928689125453, + "language_loss": 0.80584323, + "learning_rate": 9.660167251962484e-07, + "loss": 0.82782519, + "num_input_tokens_seen": 122127385, + "step": 5678, + "time_per_iteration": 2.456603527069092 + }, + { + "auxiliary_loss_clip": 0.01199441, + "auxiliary_loss_mlp": 0.01027569, + "balance_loss_clip": 1.04288268, + "balance_loss_mlp": 1.0202018, + "epoch": 0.6828593759393976, + "flos": 21688788539520.0, + "grad_norm": 2.0041682051493908, + "language_loss": 0.77721983, + "learning_rate": 9.653500113510654e-07, + "loss": 0.79948992, + "num_input_tokens_seen": 122146500, + "step": 5679, + "time_per_iteration": 3.523585319519043 + }, + { + "auxiliary_loss_clip": 0.0118452, + "auxiliary_loss_mlp": 0.01023605, + "balance_loss_clip": 1.04560137, + "balance_loss_mlp": 1.01504564, + "epoch": 0.6829796188300367, + "flos": 25337707557120.0, + "grad_norm": 2.3704380913258793, + "language_loss": 0.6700809, + "learning_rate": 9.646834544614627e-07, + "loss": 0.69216216, + "num_input_tokens_seen": 122167000, + "step": 5680, + "time_per_iteration": 2.6643495559692383 + }, + { + "auxiliary_loss_clip": 0.0117766, + "auxiliary_loss_mlp": 0.01025395, + "balance_loss_clip": 1.04469788, + "balance_loss_mlp": 1.01768851, + "epoch": 0.6830998617206757, + "flos": 20704800389760.0, + "grad_norm": 1.8243971503692027, + "language_loss": 0.76362336, + "learning_rate": 9.64017054628558e-07, + "loss": 0.78565395, + "num_input_tokens_seen": 122185825, + "step": 5681, + "time_per_iteration": 3.561070442199707 + }, + { + "auxiliary_loss_clip": 0.01209557, + "auxiliary_loss_mlp": 0.01028955, + "balance_loss_clip": 1.04190636, + "balance_loss_mlp": 1.02132821, + "epoch": 0.6832201046113149, + "flos": 21726638496000.0, + "grad_norm": 1.6170475042346932, + "language_loss": 0.79013467, + "learning_rate": 9.63350811953441e-07, + "loss": 0.81251979, + "num_input_tokens_seen": 122206200, + "step": 5682, + "time_per_iteration": 2.829533338546753 + }, + { + "auxiliary_loss_clip": 0.01198804, + "auxiliary_loss_mlp": 0.01027887, + "balance_loss_clip": 1.04478359, + "balance_loss_mlp": 1.02020383, + "epoch": 0.6833403475019539, + "flos": 19536554448000.0, + "grad_norm": 1.9229895967030401, + "language_loss": 0.70497751, + "learning_rate": 9.626847265371826e-07, + "loss": 0.72724444, + "num_input_tokens_seen": 122225520, + "step": 5683, + "time_per_iteration": 2.7012014389038086 + }, + { + "auxiliary_loss_clip": 0.01179501, + "auxiliary_loss_mlp": 0.01026352, + "balance_loss_clip": 1.04473329, + "balance_loss_mlp": 1.01915741, + "epoch": 0.683460590392593, + "flos": 19352153001600.0, + "grad_norm": 1.9378986226323447, + "language_loss": 0.78250492, + "learning_rate": 9.620187984808262e-07, + "loss": 0.8045634, + "num_input_tokens_seen": 122244320, + "step": 5684, + "time_per_iteration": 2.7122719287872314 + }, + { + "auxiliary_loss_clip": 0.01189966, + "auxiliary_loss_mlp": 0.01378668, + "balance_loss_clip": 1.04747689, + "balance_loss_mlp": 1.00009727, + "epoch": 0.6835808332832322, + "flos": 23288500650240.0, + "grad_norm": 2.43991343643253, + "language_loss": 0.8588227, + "learning_rate": 9.613530278853919e-07, + "loss": 0.88450909, + "num_input_tokens_seen": 122264295, + "step": 5685, + "time_per_iteration": 2.710033416748047 + }, + { + "auxiliary_loss_clip": 0.01178643, + "auxiliary_loss_mlp": 0.01028269, + "balance_loss_clip": 1.04708433, + "balance_loss_mlp": 1.02088344, + "epoch": 0.6837010761738712, + "flos": 21653416621440.0, + "grad_norm": 1.7246528245212438, + "language_loss": 0.73924369, + "learning_rate": 9.60687414851879e-07, + "loss": 0.76131284, + "num_input_tokens_seen": 122285300, + "step": 5686, + "time_per_iteration": 2.6502525806427 + }, + { + "auxiliary_loss_clip": 0.01190075, + "auxiliary_loss_mlp": 0.01023983, + "balance_loss_clip": 1.04576886, + "balance_loss_mlp": 1.0163238, + "epoch": 0.6838213190645103, + "flos": 17566387418880.0, + "grad_norm": 2.188345058362768, + "language_loss": 0.77480447, + "learning_rate": 9.600219594812575e-07, + "loss": 0.7969451, + "num_input_tokens_seen": 122303240, + "step": 5687, + "time_per_iteration": 2.6003479957580566 + }, + { + "auxiliary_loss_clip": 0.01170937, + "auxiliary_loss_mlp": 0.01028163, + "balance_loss_clip": 1.04935706, + "balance_loss_mlp": 1.02104592, + "epoch": 0.6839415619551494, + "flos": 23112538899840.0, + "grad_norm": 1.8196535040653237, + "language_loss": 0.72760618, + "learning_rate": 9.593566618744786e-07, + "loss": 0.74959719, + "num_input_tokens_seen": 122323390, + "step": 5688, + "time_per_iteration": 2.6870498657226562 + }, + { + "auxiliary_loss_clip": 0.0116955, + "auxiliary_loss_mlp": 0.01025144, + "balance_loss_clip": 1.04678047, + "balance_loss_mlp": 1.01782751, + "epoch": 0.6840618048457885, + "flos": 22127868391680.0, + "grad_norm": 1.6949699495023673, + "language_loss": 0.73700321, + "learning_rate": 9.58691522132466e-07, + "loss": 0.75895011, + "num_input_tokens_seen": 122342200, + "step": 5689, + "time_per_iteration": 2.6390819549560547 + }, + { + "auxiliary_loss_clip": 0.01192224, + "auxiliary_loss_mlp": 0.01028196, + "balance_loss_clip": 1.04736388, + "balance_loss_mlp": 1.02046549, + "epoch": 0.6841820477364275, + "flos": 22015898720640.0, + "grad_norm": 3.0827856369926843, + "language_loss": 0.85154283, + "learning_rate": 9.58026540356123e-07, + "loss": 0.87374699, + "num_input_tokens_seen": 122360465, + "step": 5690, + "time_per_iteration": 2.702817678451538 + }, + { + "auxiliary_loss_clip": 0.01179553, + "auxiliary_loss_mlp": 0.01027179, + "balance_loss_clip": 1.04515386, + "balance_loss_mlp": 1.0192697, + "epoch": 0.6843022906270667, + "flos": 24900531125760.0, + "grad_norm": 1.687209327866327, + "language_loss": 0.8683821, + "learning_rate": 9.573617166463246e-07, + "loss": 0.89044946, + "num_input_tokens_seen": 122381680, + "step": 5691, + "time_per_iteration": 2.6715734004974365 + }, + { + "auxiliary_loss_clip": 0.01189156, + "auxiliary_loss_mlp": 0.01021727, + "balance_loss_clip": 1.04382706, + "balance_loss_mlp": 1.01356161, + "epoch": 0.6844225335177058, + "flos": 19969924037760.0, + "grad_norm": 2.973184951077291, + "language_loss": 0.6059677, + "learning_rate": 9.56697051103924e-07, + "loss": 0.62807655, + "num_input_tokens_seen": 122399120, + "step": 5692, + "time_per_iteration": 2.6924171447753906 + }, + { + "auxiliary_loss_clip": 0.011884, + "auxiliary_loss_mlp": 0.01020819, + "balance_loss_clip": 1.04636681, + "balance_loss_mlp": 1.01331449, + "epoch": 0.6845427764083448, + "flos": 25883334126720.0, + "grad_norm": 2.0709265763280733, + "language_loss": 0.80984712, + "learning_rate": 9.560325438297522e-07, + "loss": 0.83193928, + "num_input_tokens_seen": 122417430, + "step": 5693, + "time_per_iteration": 2.655802011489868 + }, + { + "auxiliary_loss_clip": 0.01192393, + "auxiliary_loss_mlp": 0.01030139, + "balance_loss_clip": 1.05019164, + "balance_loss_mlp": 1.02261353, + "epoch": 0.684663019298984, + "flos": 18880143356160.0, + "grad_norm": 1.8378541014810494, + "language_loss": 0.86702615, + "learning_rate": 9.553681949246127e-07, + "loss": 0.88925147, + "num_input_tokens_seen": 122435055, + "step": 5694, + "time_per_iteration": 2.791532516479492 + }, + { + "auxiliary_loss_clip": 0.012019, + "auxiliary_loss_mlp": 0.01032073, + "balance_loss_clip": 1.04512167, + "balance_loss_mlp": 1.02397847, + "epoch": 0.684783262189623, + "flos": 54193725302400.0, + "grad_norm": 2.1351210719981917, + "language_loss": 0.75485903, + "learning_rate": 9.547040044892886e-07, + "loss": 0.77719879, + "num_input_tokens_seen": 122462570, + "step": 5695, + "time_per_iteration": 3.002687931060791 + }, + { + "auxiliary_loss_clip": 0.01078891, + "auxiliary_loss_mlp": 0.0099884, + "balance_loss_clip": 1.01123548, + "balance_loss_mlp": 0.99767178, + "epoch": 0.6849035050802621, + "flos": 63970264143360.0, + "grad_norm": 0.861034203804743, + "language_loss": 0.60088289, + "learning_rate": 9.540399726245354e-07, + "loss": 0.62166023, + "num_input_tokens_seen": 122519275, + "step": 5696, + "time_per_iteration": 3.0937912464141846 + }, + { + "auxiliary_loss_clip": 0.01182704, + "auxiliary_loss_mlp": 0.0102569, + "balance_loss_clip": 1.04444361, + "balance_loss_mlp": 1.01716661, + "epoch": 0.6850237479709013, + "flos": 25224121774080.0, + "grad_norm": 1.836203375420767, + "language_loss": 0.6895265, + "learning_rate": 9.533760994310859e-07, + "loss": 0.71161044, + "num_input_tokens_seen": 122539675, + "step": 5697, + "time_per_iteration": 2.692432165145874 + }, + { + "auxiliary_loss_clip": 0.01174017, + "auxiliary_loss_mlp": 0.01025601, + "balance_loss_clip": 1.05044782, + "balance_loss_mlp": 1.01782846, + "epoch": 0.6851439908615403, + "flos": 19354128249600.0, + "grad_norm": 2.61676903460205, + "language_loss": 0.75479221, + "learning_rate": 9.527123850096508e-07, + "loss": 0.77678835, + "num_input_tokens_seen": 122558035, + "step": 5698, + "time_per_iteration": 2.5872292518615723 + }, + { + "auxiliary_loss_clip": 0.01182972, + "auxiliary_loss_mlp": 0.01027212, + "balance_loss_clip": 1.04563308, + "balance_loss_mlp": 1.01969028, + "epoch": 0.6852642337521794, + "flos": 23182133500800.0, + "grad_norm": 1.7684914174305917, + "language_loss": 0.71753931, + "learning_rate": 9.520488294609142e-07, + "loss": 0.73964113, + "num_input_tokens_seen": 122576815, + "step": 5699, + "time_per_iteration": 2.637726306915283 + }, + { + "auxiliary_loss_clip": 0.01113733, + "auxiliary_loss_mlp": 0.01002988, + "balance_loss_clip": 1.0107131, + "balance_loss_mlp": 1.00165296, + "epoch": 0.6853844766428185, + "flos": 62647206583680.0, + "grad_norm": 0.7384018187385505, + "language_loss": 0.53826606, + "learning_rate": 9.513854328855368e-07, + "loss": 0.55943334, + "num_input_tokens_seen": 122634690, + "step": 5700, + "time_per_iteration": 4.109625577926636 + }, + { + "auxiliary_loss_clip": 0.01167979, + "auxiliary_loss_mlp": 0.01023278, + "balance_loss_clip": 1.04767168, + "balance_loss_mlp": 1.01594102, + "epoch": 0.6855047195334576, + "flos": 23437242869760.0, + "grad_norm": 1.8632541675522185, + "language_loss": 0.80995226, + "learning_rate": 9.507221953841558e-07, + "loss": 0.83186483, + "num_input_tokens_seen": 122652320, + "step": 5701, + "time_per_iteration": 2.546649217605591 + }, + { + "auxiliary_loss_clip": 0.01183053, + "auxiliary_loss_mlp": 0.01026344, + "balance_loss_clip": 1.05039644, + "balance_loss_mlp": 1.01862502, + "epoch": 0.6856249624240967, + "flos": 20664831530880.0, + "grad_norm": 1.6309085640816559, + "language_loss": 0.77804375, + "learning_rate": 9.500591170573824e-07, + "loss": 0.80013776, + "num_input_tokens_seen": 122672340, + "step": 5702, + "time_per_iteration": 3.540978193283081 + }, + { + "auxiliary_loss_clip": 0.01199801, + "auxiliary_loss_mlp": 0.01027532, + "balance_loss_clip": 1.04095531, + "balance_loss_mlp": 1.01963758, + "epoch": 0.6857452053147358, + "flos": 17087302794240.0, + "grad_norm": 1.9476649960391723, + "language_loss": 0.74249828, + "learning_rate": 9.493961980058078e-07, + "loss": 0.76477164, + "num_input_tokens_seen": 122689935, + "step": 5703, + "time_per_iteration": 2.632645606994629 + }, + { + "auxiliary_loss_clip": 0.01194927, + "auxiliary_loss_mlp": 0.01022089, + "balance_loss_clip": 1.03728437, + "balance_loss_mlp": 1.01507902, + "epoch": 0.6858654482053749, + "flos": 30847266057600.0, + "grad_norm": 2.3232697400526616, + "language_loss": 0.67391694, + "learning_rate": 9.48733438329993e-07, + "loss": 0.69608712, + "num_input_tokens_seen": 122710200, + "step": 5704, + "time_per_iteration": 3.7317075729370117 + }, + { + "auxiliary_loss_clip": 0.01169308, + "auxiliary_loss_mlp": 0.01378589, + "balance_loss_clip": 1.04909849, + "balance_loss_mlp": 1.00014043, + "epoch": 0.6859856910960139, + "flos": 28877314510080.0, + "grad_norm": 1.700537111962925, + "language_loss": 0.74265802, + "learning_rate": 9.480708381304807e-07, + "loss": 0.76813698, + "num_input_tokens_seen": 122731495, + "step": 5705, + "time_per_iteration": 2.7203681468963623 + }, + { + "auxiliary_loss_clip": 0.0120726, + "auxiliary_loss_mlp": 0.01035077, + "balance_loss_clip": 1.04821396, + "balance_loss_mlp": 1.02715564, + "epoch": 0.6861059339866531, + "flos": 19354523299200.0, + "grad_norm": 2.2713993074131054, + "language_loss": 0.83573669, + "learning_rate": 9.474083975077858e-07, + "loss": 0.85816014, + "num_input_tokens_seen": 122748620, + "step": 5706, + "time_per_iteration": 2.7254772186279297 + }, + { + "auxiliary_loss_clip": 0.01170232, + "auxiliary_loss_mlp": 0.01027529, + "balance_loss_clip": 1.04522848, + "balance_loss_mlp": 1.01983976, + "epoch": 0.6862261768772921, + "flos": 22199976944640.0, + "grad_norm": 2.0261393104504863, + "language_loss": 0.80243552, + "learning_rate": 9.467461165623994e-07, + "loss": 0.82441312, + "num_input_tokens_seen": 122767670, + "step": 5707, + "time_per_iteration": 3.5256028175354004 + }, + { + "auxiliary_loss_clip": 0.01178255, + "auxiliary_loss_mlp": 0.01025269, + "balance_loss_clip": 1.04410291, + "balance_loss_mlp": 1.01757431, + "epoch": 0.6863464197679312, + "flos": 26285677344000.0, + "grad_norm": 2.072763047212287, + "language_loss": 0.7925747, + "learning_rate": 9.46083995394791e-07, + "loss": 0.81460989, + "num_input_tokens_seen": 122785480, + "step": 5708, + "time_per_iteration": 2.6110570430755615 + }, + { + "auxiliary_loss_clip": 0.0118046, + "auxiliary_loss_mlp": 0.01378544, + "balance_loss_clip": 1.04682136, + "balance_loss_mlp": 1.00019979, + "epoch": 0.6864666626585703, + "flos": 37815228564480.0, + "grad_norm": 1.8022198386680013, + "language_loss": 0.63549936, + "learning_rate": 9.454220341054012e-07, + "loss": 0.66108936, + "num_input_tokens_seen": 122810265, + "step": 5709, + "time_per_iteration": 2.7650554180145264 + }, + { + "auxiliary_loss_clip": 0.01196088, + "auxiliary_loss_mlp": 0.01027492, + "balance_loss_clip": 1.04425621, + "balance_loss_mlp": 1.02009475, + "epoch": 0.6865869055492094, + "flos": 19391152193280.0, + "grad_norm": 1.95714707006156, + "language_loss": 0.80403721, + "learning_rate": 9.447602327946512e-07, + "loss": 0.82627296, + "num_input_tokens_seen": 122828905, + "step": 5710, + "time_per_iteration": 2.686561107635498 + }, + { + "auxiliary_loss_clip": 0.01182076, + "auxiliary_loss_mlp": 0.01029402, + "balance_loss_clip": 1.04251909, + "balance_loss_mlp": 1.02136707, + "epoch": 0.6867071484398485, + "flos": 20375966355840.0, + "grad_norm": 1.826732407800208, + "language_loss": 0.76586604, + "learning_rate": 9.440985915629338e-07, + "loss": 0.78798079, + "num_input_tokens_seen": 122846235, + "step": 5711, + "time_per_iteration": 2.5920748710632324 + }, + { + "auxiliary_loss_clip": 0.01170458, + "auxiliary_loss_mlp": 0.01022867, + "balance_loss_clip": 1.04958272, + "balance_loss_mlp": 1.01578057, + "epoch": 0.6868273913304875, + "flos": 15889143801600.0, + "grad_norm": 1.8851790497831475, + "language_loss": 0.73444676, + "learning_rate": 9.434371105106223e-07, + "loss": 0.75638002, + "num_input_tokens_seen": 122863835, + "step": 5712, + "time_per_iteration": 2.5633957386016846 + }, + { + "auxiliary_loss_clip": 0.01193291, + "auxiliary_loss_mlp": 0.01027092, + "balance_loss_clip": 1.04401755, + "balance_loss_mlp": 1.01892042, + "epoch": 0.6869476342211267, + "flos": 24462492768000.0, + "grad_norm": 1.7519207683049849, + "language_loss": 0.70440555, + "learning_rate": 9.427757897380602e-07, + "loss": 0.72660941, + "num_input_tokens_seen": 122883235, + "step": 5713, + "time_per_iteration": 2.6927850246429443 + }, + { + "auxiliary_loss_clip": 0.01194953, + "auxiliary_loss_mlp": 0.01027625, + "balance_loss_clip": 1.04665661, + "balance_loss_mlp": 1.0196259, + "epoch": 0.6870678771117658, + "flos": 18442571875200.0, + "grad_norm": 2.6082536886808207, + "language_loss": 0.84665304, + "learning_rate": 9.421146293455695e-07, + "loss": 0.86887872, + "num_input_tokens_seen": 122898975, + "step": 5714, + "time_per_iteration": 2.6909732818603516 + }, + { + "auxiliary_loss_clip": 0.01184278, + "auxiliary_loss_mlp": 0.01025315, + "balance_loss_clip": 1.04461432, + "balance_loss_mlp": 1.01776862, + "epoch": 0.6871881200024048, + "flos": 22200371994240.0, + "grad_norm": 1.9528578736281508, + "language_loss": 0.68500078, + "learning_rate": 9.414536294334489e-07, + "loss": 0.7070967, + "num_input_tokens_seen": 122918995, + "step": 5715, + "time_per_iteration": 2.6483559608459473 + }, + { + "auxiliary_loss_clip": 0.0118772, + "auxiliary_loss_mlp": 0.01021177, + "balance_loss_clip": 1.04169989, + "balance_loss_mlp": 1.01417971, + "epoch": 0.687308362893044, + "flos": 22127724737280.0, + "grad_norm": 1.8261449489125843, + "language_loss": 0.69387746, + "learning_rate": 9.407927901019708e-07, + "loss": 0.71596646, + "num_input_tokens_seen": 122938125, + "step": 5716, + "time_per_iteration": 2.7323191165924072 + }, + { + "auxiliary_loss_clip": 0.01180475, + "auxiliary_loss_mlp": 0.01028341, + "balance_loss_clip": 1.04729939, + "balance_loss_mlp": 1.02061057, + "epoch": 0.687428605783683, + "flos": 25040546340480.0, + "grad_norm": 2.0732564885357734, + "language_loss": 0.77033514, + "learning_rate": 9.401321114513854e-07, + "loss": 0.79242331, + "num_input_tokens_seen": 122957020, + "step": 5717, + "time_per_iteration": 2.625098943710327 + }, + { + "auxiliary_loss_clip": 0.0117209, + "auxiliary_loss_mlp": 0.01022667, + "balance_loss_clip": 1.04919338, + "balance_loss_mlp": 1.01509714, + "epoch": 0.6875488486743221, + "flos": 23770063313280.0, + "grad_norm": 1.7446377041015344, + "language_loss": 0.75401717, + "learning_rate": 9.394715935819155e-07, + "loss": 0.77596474, + "num_input_tokens_seen": 122977410, + "step": 5718, + "time_per_iteration": 2.599289655685425 + }, + { + "auxiliary_loss_clip": 0.01181486, + "auxiliary_loss_mlp": 0.01021442, + "balance_loss_clip": 1.04602599, + "balance_loss_mlp": 1.01398563, + "epoch": 0.6876690915649613, + "flos": 25516937445120.0, + "grad_norm": 2.06266252778582, + "language_loss": 0.63024241, + "learning_rate": 9.388112365937608e-07, + "loss": 0.65227163, + "num_input_tokens_seen": 122996875, + "step": 5719, + "time_per_iteration": 2.6135599613189697 + }, + { + "auxiliary_loss_clip": 0.01196949, + "auxiliary_loss_mlp": 0.01029016, + "balance_loss_clip": 1.04447806, + "balance_loss_mlp": 1.02124333, + "epoch": 0.6877893344556003, + "flos": 19427996568960.0, + "grad_norm": 1.9788420286175696, + "language_loss": 0.82500345, + "learning_rate": 9.381510405870985e-07, + "loss": 0.8472631, + "num_input_tokens_seen": 123015890, + "step": 5720, + "time_per_iteration": 2.667630672454834 + }, + { + "auxiliary_loss_clip": 0.01177875, + "auxiliary_loss_mlp": 0.01022016, + "balance_loss_clip": 1.04711056, + "balance_loss_mlp": 1.01386774, + "epoch": 0.6879095773462394, + "flos": 18661303745280.0, + "grad_norm": 2.4526119399585915, + "language_loss": 0.77178776, + "learning_rate": 9.374910056620791e-07, + "loss": 0.7937867, + "num_input_tokens_seen": 123034955, + "step": 5721, + "time_per_iteration": 2.5569186210632324 + }, + { + "auxiliary_loss_clip": 0.01183462, + "auxiliary_loss_mlp": 0.01024415, + "balance_loss_clip": 1.04891658, + "balance_loss_mlp": 1.01693487, + "epoch": 0.6880298202368785, + "flos": 20883132437760.0, + "grad_norm": 1.670627821304979, + "language_loss": 0.8103615, + "learning_rate": 9.368311319188293e-07, + "loss": 0.83244026, + "num_input_tokens_seen": 123052770, + "step": 5722, + "time_per_iteration": 2.6061151027679443 + }, + { + "auxiliary_loss_clip": 0.01196143, + "auxiliary_loss_mlp": 0.01026967, + "balance_loss_clip": 1.04377973, + "balance_loss_mlp": 1.01906395, + "epoch": 0.6881500631275176, + "flos": 30153292318080.0, + "grad_norm": 1.8250052794499394, + "language_loss": 0.79083502, + "learning_rate": 9.361714194574515e-07, + "loss": 0.81306612, + "num_input_tokens_seen": 123075105, + "step": 5723, + "time_per_iteration": 2.70670223236084 + }, + { + "auxiliary_loss_clip": 0.01064787, + "auxiliary_loss_mlp": 0.01004008, + "balance_loss_clip": 1.01131916, + "balance_loss_mlp": 1.00284004, + "epoch": 0.6882703060181566, + "flos": 66181537215360.0, + "grad_norm": 0.7305798246510726, + "language_loss": 0.58207345, + "learning_rate": 9.355118683780228e-07, + "loss": 0.60276139, + "num_input_tokens_seen": 123145175, + "step": 5724, + "time_per_iteration": 3.264754295349121 + }, + { + "auxiliary_loss_clip": 0.01169889, + "auxiliary_loss_mlp": 0.01023588, + "balance_loss_clip": 1.04674363, + "balance_loss_mlp": 1.01619685, + "epoch": 0.6883905489087958, + "flos": 18214646123520.0, + "grad_norm": 2.1785104789898595, + "language_loss": 0.79218936, + "learning_rate": 9.348524787805987e-07, + "loss": 0.81412411, + "num_input_tokens_seen": 123160365, + "step": 5725, + "time_per_iteration": 2.524766445159912 + }, + { + "auxiliary_loss_clip": 0.01196215, + "auxiliary_loss_mlp": 0.01027732, + "balance_loss_clip": 1.0405314, + "balance_loss_mlp": 1.0204308, + "epoch": 0.6885107917994349, + "flos": 14056262553600.0, + "grad_norm": 2.4281336384013907, + "language_loss": 0.85218906, + "learning_rate": 9.341932507652053e-07, + "loss": 0.87442851, + "num_input_tokens_seen": 123174855, + "step": 5726, + "time_per_iteration": 3.5504300594329834 + }, + { + "auxiliary_loss_clip": 0.01181771, + "auxiliary_loss_mlp": 0.0102878, + "balance_loss_clip": 1.04065788, + "balance_loss_mlp": 1.02048349, + "epoch": 0.6886310346900739, + "flos": 28690722334080.0, + "grad_norm": 1.6754545777837662, + "language_loss": 0.78588796, + "learning_rate": 9.335341844318489e-07, + "loss": 0.80799347, + "num_input_tokens_seen": 123194995, + "step": 5727, + "time_per_iteration": 2.7311415672302246 + }, + { + "auxiliary_loss_clip": 0.01188403, + "auxiliary_loss_mlp": 0.01025864, + "balance_loss_clip": 1.04551864, + "balance_loss_mlp": 1.01853275, + "epoch": 0.6887512775807131, + "flos": 24535319592960.0, + "grad_norm": 1.815577523349892, + "language_loss": 0.73470396, + "learning_rate": 9.328752798805091e-07, + "loss": 0.75684661, + "num_input_tokens_seen": 123213465, + "step": 5728, + "time_per_iteration": 3.7468831539154053 + }, + { + "auxiliary_loss_clip": 0.01181814, + "auxiliary_loss_mlp": 0.01024361, + "balance_loss_clip": 1.04788446, + "balance_loss_mlp": 1.01660037, + "epoch": 0.6888715204713521, + "flos": 22414363269120.0, + "grad_norm": 4.786339328351845, + "language_loss": 0.75835133, + "learning_rate": 9.322165372111399e-07, + "loss": 0.78041309, + "num_input_tokens_seen": 123231610, + "step": 5729, + "time_per_iteration": 2.611288547515869 + }, + { + "auxiliary_loss_clip": 0.0119263, + "auxiliary_loss_mlp": 0.01031199, + "balance_loss_clip": 1.04562044, + "balance_loss_mlp": 1.02413535, + "epoch": 0.6889917633619912, + "flos": 22054323294720.0, + "grad_norm": 1.83247959734845, + "language_loss": 0.75543666, + "learning_rate": 9.315579565236747e-07, + "loss": 0.77767491, + "num_input_tokens_seen": 123250715, + "step": 5730, + "time_per_iteration": 3.473548412322998 + }, + { + "auxiliary_loss_clip": 0.01186973, + "auxiliary_loss_mlp": 0.01024943, + "balance_loss_clip": 1.04861093, + "balance_loss_mlp": 1.0165565, + "epoch": 0.6891120062526304, + "flos": 23949724164480.0, + "grad_norm": 1.88210726520132, + "language_loss": 0.74394381, + "learning_rate": 9.308995379180162e-07, + "loss": 0.76606297, + "num_input_tokens_seen": 123270270, + "step": 5731, + "time_per_iteration": 2.657813310623169 + }, + { + "auxiliary_loss_clip": 0.0107955, + "auxiliary_loss_mlp": 0.01005776, + "balance_loss_clip": 1.01203465, + "balance_loss_mlp": 1.00458419, + "epoch": 0.6892322491432694, + "flos": 64117354337280.0, + "grad_norm": 0.7410989810455061, + "language_loss": 0.5953449, + "learning_rate": 9.302412814940488e-07, + "loss": 0.61619806, + "num_input_tokens_seen": 123333045, + "step": 5732, + "time_per_iteration": 3.242539405822754 + }, + { + "auxiliary_loss_clip": 0.0118474, + "auxiliary_loss_mlp": 0.01030209, + "balance_loss_clip": 1.04254413, + "balance_loss_mlp": 1.02225137, + "epoch": 0.6893524920339085, + "flos": 23002436736000.0, + "grad_norm": 2.1079435946033893, + "language_loss": 0.71052039, + "learning_rate": 9.295831873516276e-07, + "loss": 0.73266983, + "num_input_tokens_seen": 123352320, + "step": 5733, + "time_per_iteration": 3.6245498657226562 + }, + { + "auxiliary_loss_clip": 0.01170488, + "auxiliary_loss_mlp": 0.01028987, + "balance_loss_clip": 1.04981565, + "balance_loss_mlp": 1.02163196, + "epoch": 0.6894727349245476, + "flos": 21396260177280.0, + "grad_norm": 1.6466807911877341, + "language_loss": 0.76102841, + "learning_rate": 9.289252555905873e-07, + "loss": 0.78302318, + "num_input_tokens_seen": 123372400, + "step": 5734, + "time_per_iteration": 2.6088390350341797 + }, + { + "auxiliary_loss_clip": 0.01183023, + "auxiliary_loss_mlp": 0.01033491, + "balance_loss_clip": 1.0505085, + "balance_loss_mlp": 1.02547967, + "epoch": 0.6895929778151867, + "flos": 19865316654720.0, + "grad_norm": 2.2303068904062036, + "language_loss": 0.75880301, + "learning_rate": 9.282674863107334e-07, + "loss": 0.78096819, + "num_input_tokens_seen": 123390215, + "step": 5735, + "time_per_iteration": 2.62056565284729 + }, + { + "auxiliary_loss_clip": 0.01174854, + "auxiliary_loss_mlp": 0.01026487, + "balance_loss_clip": 1.04644787, + "balance_loss_mlp": 1.01905465, + "epoch": 0.6897132207058257, + "flos": 18179166464640.0, + "grad_norm": 2.2161672087688014, + "language_loss": 0.75742221, + "learning_rate": 9.276098796118488e-07, + "loss": 0.77943563, + "num_input_tokens_seen": 123406700, + "step": 5736, + "time_per_iteration": 2.5496363639831543 + }, + { + "auxiliary_loss_clip": 0.011882, + "auxiliary_loss_mlp": 0.01023965, + "balance_loss_clip": 1.04648006, + "balance_loss_mlp": 1.01701486, + "epoch": 0.6898334635964649, + "flos": 32561641359360.0, + "grad_norm": 1.8763745296721877, + "language_loss": 0.66359198, + "learning_rate": 9.269524355936938e-07, + "loss": 0.68571365, + "num_input_tokens_seen": 123429880, + "step": 5737, + "time_per_iteration": 2.746918201446533 + }, + { + "auxiliary_loss_clip": 0.01180366, + "auxiliary_loss_mlp": 0.01028784, + "balance_loss_clip": 1.04233468, + "balance_loss_mlp": 1.02145267, + "epoch": 0.689953706487104, + "flos": 22819004956800.0, + "grad_norm": 1.6496893977998177, + "language_loss": 0.85060632, + "learning_rate": 9.262951543560002e-07, + "loss": 0.87269789, + "num_input_tokens_seen": 123449105, + "step": 5738, + "time_per_iteration": 2.618983507156372 + }, + { + "auxiliary_loss_clip": 0.01188882, + "auxiliary_loss_mlp": 0.01022855, + "balance_loss_clip": 1.04817975, + "balance_loss_mlp": 1.01515388, + "epoch": 0.690073949377743, + "flos": 18515362786560.0, + "grad_norm": 2.3786742607855893, + "language_loss": 0.86058468, + "learning_rate": 9.256380359984795e-07, + "loss": 0.88270205, + "num_input_tokens_seen": 123466215, + "step": 5739, + "time_per_iteration": 2.590975046157837 + }, + { + "auxiliary_loss_clip": 0.01206922, + "auxiliary_loss_mlp": 0.01028723, + "balance_loss_clip": 1.0380497, + "balance_loss_mlp": 1.02120709, + "epoch": 0.6901941922683821, + "flos": 34857194716800.0, + "grad_norm": 1.8786773907821521, + "language_loss": 0.74964428, + "learning_rate": 9.249810806208139e-07, + "loss": 0.77200067, + "num_input_tokens_seen": 123485480, + "step": 5740, + "time_per_iteration": 2.774841547012329 + }, + { + "auxiliary_loss_clip": 0.01201482, + "auxiliary_loss_mlp": 0.01378453, + "balance_loss_clip": 1.03991151, + "balance_loss_mlp": 1.00013518, + "epoch": 0.6903144351590212, + "flos": 16253672976000.0, + "grad_norm": 1.8741823662647268, + "language_loss": 0.80480647, + "learning_rate": 9.243242883226627e-07, + "loss": 0.83060586, + "num_input_tokens_seen": 123504575, + "step": 5741, + "time_per_iteration": 2.7203221321105957 + }, + { + "auxiliary_loss_clip": 0.0118049, + "auxiliary_loss_mlp": 0.01029213, + "balance_loss_clip": 1.04302585, + "balance_loss_mlp": 1.0212853, + "epoch": 0.6904346780496603, + "flos": 28035137255040.0, + "grad_norm": 2.354885146253267, + "language_loss": 0.69753921, + "learning_rate": 9.236676592036628e-07, + "loss": 0.7196362, + "num_input_tokens_seen": 123524250, + "step": 5742, + "time_per_iteration": 2.648561477661133 + }, + { + "auxiliary_loss_clip": 0.01188307, + "auxiliary_loss_mlp": 0.01030361, + "balance_loss_clip": 1.05070567, + "balance_loss_mlp": 1.02266645, + "epoch": 0.6905549209402994, + "flos": 23624266008960.0, + "grad_norm": 1.6500231533608334, + "language_loss": 0.73500693, + "learning_rate": 9.230111933634228e-07, + "loss": 0.75719362, + "num_input_tokens_seen": 123545845, + "step": 5743, + "time_per_iteration": 2.701848268508911 + }, + { + "auxiliary_loss_clip": 0.01185135, + "auxiliary_loss_mlp": 0.01025751, + "balance_loss_clip": 1.05044615, + "balance_loss_mlp": 1.01807356, + "epoch": 0.6906751638309385, + "flos": 23114945111040.0, + "grad_norm": 1.597770163811172, + "language_loss": 0.80738366, + "learning_rate": 9.223548909015288e-07, + "loss": 0.82949257, + "num_input_tokens_seen": 123567535, + "step": 5744, + "time_per_iteration": 2.7558999061584473 + }, + { + "auxiliary_loss_clip": 0.01198518, + "auxiliary_loss_mlp": 0.01025959, + "balance_loss_clip": 1.04136109, + "balance_loss_mlp": 1.01839793, + "epoch": 0.6907954067215776, + "flos": 27305468375040.0, + "grad_norm": 2.213967606492681, + "language_loss": 0.7195276, + "learning_rate": 9.216987519175407e-07, + "loss": 0.74177241, + "num_input_tokens_seen": 123587710, + "step": 5745, + "time_per_iteration": 2.769428253173828 + }, + { + "auxiliary_loss_clip": 0.01175161, + "auxiliary_loss_mlp": 0.01024924, + "balance_loss_clip": 1.04748476, + "balance_loss_mlp": 1.01763737, + "epoch": 0.6909156496122166, + "flos": 21689399070720.0, + "grad_norm": 1.7356606257105398, + "language_loss": 0.68252426, + "learning_rate": 9.210427765109942e-07, + "loss": 0.70452511, + "num_input_tokens_seen": 123607385, + "step": 5746, + "time_per_iteration": 2.722115993499756 + }, + { + "auxiliary_loss_clip": 0.01186735, + "auxiliary_loss_mlp": 0.01026831, + "balance_loss_clip": 1.04290795, + "balance_loss_mlp": 1.01789045, + "epoch": 0.6910358925028558, + "flos": 22561453463040.0, + "grad_norm": 1.874437361263131, + "language_loss": 0.81474447, + "learning_rate": 9.20386964781402e-07, + "loss": 0.83688009, + "num_input_tokens_seen": 123625405, + "step": 5747, + "time_per_iteration": 2.655055046081543 + }, + { + "auxiliary_loss_clip": 0.01186864, + "auxiliary_loss_mlp": 0.01028393, + "balance_loss_clip": 1.04532731, + "balance_loss_mlp": 1.02121067, + "epoch": 0.6911561353934949, + "flos": 22054107813120.0, + "grad_norm": 2.2432704235780165, + "language_loss": 0.84384841, + "learning_rate": 9.197313168282472e-07, + "loss": 0.86600101, + "num_input_tokens_seen": 123642850, + "step": 5748, + "time_per_iteration": 2.648186445236206 + }, + { + "auxiliary_loss_clip": 0.01174734, + "auxiliary_loss_mlp": 0.01023245, + "balance_loss_clip": 1.0447638, + "balance_loss_mlp": 1.01553774, + "epoch": 0.6912763782841339, + "flos": 24206557386240.0, + "grad_norm": 2.2221076009068916, + "language_loss": 0.71659076, + "learning_rate": 9.190758327509935e-07, + "loss": 0.73857057, + "num_input_tokens_seen": 123661595, + "step": 5749, + "time_per_iteration": 2.6353273391723633 + }, + { + "auxiliary_loss_clip": 0.01117744, + "auxiliary_loss_mlp": 0.01374086, + "balance_loss_clip": 1.01178765, + "balance_loss_mlp": 0.99982548, + "epoch": 0.6913966211747731, + "flos": 52329641091840.0, + "grad_norm": 0.9352454430928929, + "language_loss": 0.64437699, + "learning_rate": 9.184205126490767e-07, + "loss": 0.66929525, + "num_input_tokens_seen": 123710490, + "step": 5750, + "time_per_iteration": 3.210223436355591 + }, + { + "auxiliary_loss_clip": 0.01101792, + "auxiliary_loss_mlp": 0.01373922, + "balance_loss_clip": 1.01077127, + "balance_loss_mlp": 0.99968511, + "epoch": 0.6915168640654121, + "flos": 66741274851840.0, + "grad_norm": 1.1128947688084025, + "language_loss": 0.59652418, + "learning_rate": 9.177653566219075e-07, + "loss": 0.62128133, + "num_input_tokens_seen": 123765215, + "step": 5751, + "time_per_iteration": 4.043115854263306 + }, + { + "auxiliary_loss_clip": 0.01200743, + "auxiliary_loss_mlp": 0.01026256, + "balance_loss_clip": 1.04437959, + "balance_loss_mlp": 1.01850176, + "epoch": 0.6916371069560512, + "flos": 18296523175680.0, + "grad_norm": 3.9104791313426914, + "language_loss": 0.76240742, + "learning_rate": 9.171103647688744e-07, + "loss": 0.78467739, + "num_input_tokens_seen": 123783955, + "step": 5752, + "time_per_iteration": 2.6262898445129395 + }, + { + "auxiliary_loss_clip": 0.0121037, + "auxiliary_loss_mlp": 0.01030176, + "balance_loss_clip": 1.03765678, + "balance_loss_mlp": 1.02348542, + "epoch": 0.6917573498466904, + "flos": 19645794685440.0, + "grad_norm": 1.8665647403677148, + "language_loss": 0.68946016, + "learning_rate": 9.164555371893367e-07, + "loss": 0.7118656, + "num_input_tokens_seen": 123803885, + "step": 5753, + "time_per_iteration": 3.699620485305786 + }, + { + "auxiliary_loss_clip": 0.01182911, + "auxiliary_loss_mlp": 0.01378747, + "balance_loss_clip": 1.04967546, + "balance_loss_mlp": 1.00011528, + "epoch": 0.6918775927373294, + "flos": 14210319985920.0, + "grad_norm": 2.2117776415521044, + "language_loss": 0.75121593, + "learning_rate": 9.158008739826333e-07, + "loss": 0.77683252, + "num_input_tokens_seen": 123821485, + "step": 5754, + "time_per_iteration": 2.5638694763183594 + }, + { + "auxiliary_loss_clip": 0.01187496, + "auxiliary_loss_mlp": 0.01035809, + "balance_loss_clip": 1.04849577, + "balance_loss_mlp": 1.02850151, + "epoch": 0.6919978356279685, + "flos": 23985455218560.0, + "grad_norm": 1.5669049191065385, + "language_loss": 0.86499107, + "learning_rate": 9.151463752480744e-07, + "loss": 0.88722414, + "num_input_tokens_seen": 123840215, + "step": 5755, + "time_per_iteration": 2.657747745513916 + }, + { + "auxiliary_loss_clip": 0.01188014, + "auxiliary_loss_mlp": 0.01025346, + "balance_loss_clip": 1.04418015, + "balance_loss_mlp": 1.01809216, + "epoch": 0.6921180785186076, + "flos": 23622937205760.0, + "grad_norm": 1.4343897266492758, + "language_loss": 0.80429894, + "learning_rate": 9.144920410849493e-07, + "loss": 0.82643259, + "num_input_tokens_seen": 123861450, + "step": 5756, + "time_per_iteration": 3.6530935764312744 + }, + { + "auxiliary_loss_clip": 0.01193061, + "auxiliary_loss_mlp": 0.0103082, + "balance_loss_clip": 1.04722309, + "balance_loss_mlp": 1.02316082, + "epoch": 0.6922383214092467, + "flos": 21142623265920.0, + "grad_norm": 1.805333380095378, + "language_loss": 0.8063947, + "learning_rate": 9.138378715925176e-07, + "loss": 0.82863349, + "num_input_tokens_seen": 123880545, + "step": 5757, + "time_per_iteration": 2.6717095375061035 + }, + { + "auxiliary_loss_clip": 0.01177082, + "auxiliary_loss_mlp": 0.01027655, + "balance_loss_clip": 1.04119134, + "balance_loss_mlp": 1.0198946, + "epoch": 0.6923585642998857, + "flos": 21470667200640.0, + "grad_norm": 1.5857011633854659, + "language_loss": 0.81022245, + "learning_rate": 9.131838668700167e-07, + "loss": 0.83226979, + "num_input_tokens_seen": 123900615, + "step": 5758, + "time_per_iteration": 2.732893228530884 + }, + { + "auxiliary_loss_clip": 0.01193567, + "auxiliary_loss_mlp": 0.01023619, + "balance_loss_clip": 1.04274511, + "balance_loss_mlp": 1.01661575, + "epoch": 0.6924788071905249, + "flos": 21105204272640.0, + "grad_norm": 1.724461373608356, + "language_loss": 0.86349082, + "learning_rate": 9.125300270166598e-07, + "loss": 0.88566267, + "num_input_tokens_seen": 123921220, + "step": 5759, + "time_per_iteration": 3.6358234882354736 + }, + { + "auxiliary_loss_clip": 0.01203167, + "auxiliary_loss_mlp": 0.01021791, + "balance_loss_clip": 1.04283917, + "balance_loss_mlp": 1.01398849, + "epoch": 0.692599050081164, + "flos": 26250018117120.0, + "grad_norm": 1.8085002358804694, + "language_loss": 0.85677177, + "learning_rate": 9.118763521316324e-07, + "loss": 0.87902135, + "num_input_tokens_seen": 123941795, + "step": 5760, + "time_per_iteration": 2.810692548751831 + }, + { + "auxiliary_loss_clip": 0.01171059, + "auxiliary_loss_mlp": 0.01378803, + "balance_loss_clip": 1.04832399, + "balance_loss_mlp": 1.00016141, + "epoch": 0.692719292971803, + "flos": 20885215426560.0, + "grad_norm": 2.2041964478820697, + "language_loss": 0.76356524, + "learning_rate": 9.112228423140987e-07, + "loss": 0.78906387, + "num_input_tokens_seen": 123960715, + "step": 5761, + "time_per_iteration": 2.6525590419769287 + }, + { + "auxiliary_loss_clip": 0.01191518, + "auxiliary_loss_mlp": 0.01027245, + "balance_loss_clip": 1.04534984, + "balance_loss_mlp": 1.01970541, + "epoch": 0.6928395358624422, + "flos": 25921938268800.0, + "grad_norm": 2.548571727749434, + "language_loss": 0.86084157, + "learning_rate": 9.105694976631932e-07, + "loss": 0.88302922, + "num_input_tokens_seen": 123978625, + "step": 5762, + "time_per_iteration": 2.666243314743042 + }, + { + "auxiliary_loss_clip": 0.01181335, + "auxiliary_loss_mlp": 0.01026096, + "balance_loss_clip": 1.04819989, + "balance_loss_mlp": 1.01742363, + "epoch": 0.6929597787530812, + "flos": 23586559706880.0, + "grad_norm": 2.1503963310840524, + "language_loss": 0.72468042, + "learning_rate": 9.099163182780283e-07, + "loss": 0.74675477, + "num_input_tokens_seen": 123996780, + "step": 5763, + "time_per_iteration": 2.6657497882843018 + }, + { + "auxiliary_loss_clip": 0.01185345, + "auxiliary_loss_mlp": 0.01025198, + "balance_loss_clip": 1.04637563, + "balance_loss_mlp": 1.01700866, + "epoch": 0.6930800216437203, + "flos": 18255656476800.0, + "grad_norm": 3.3945358217414445, + "language_loss": 0.48732072, + "learning_rate": 9.092633042576916e-07, + "loss": 0.50942618, + "num_input_tokens_seen": 124014045, + "step": 5764, + "time_per_iteration": 2.571592330932617 + }, + { + "auxiliary_loss_clip": 0.01186495, + "auxiliary_loss_mlp": 0.01023157, + "balance_loss_clip": 1.04826057, + "balance_loss_mlp": 1.01567698, + "epoch": 0.6932002645343595, + "flos": 29168621809920.0, + "grad_norm": 2.5736125425694842, + "language_loss": 0.55866194, + "learning_rate": 9.086104557012446e-07, + "loss": 0.58075845, + "num_input_tokens_seen": 124034615, + "step": 5765, + "time_per_iteration": 2.723022222518921 + }, + { + "auxiliary_loss_clip": 0.01166935, + "auxiliary_loss_mlp": 0.01023341, + "balance_loss_clip": 1.04425454, + "balance_loss_mlp": 1.01553833, + "epoch": 0.6933205074249985, + "flos": 23842746483840.0, + "grad_norm": 1.7391499581035155, + "language_loss": 0.65894151, + "learning_rate": 9.079577727077239e-07, + "loss": 0.68084431, + "num_input_tokens_seen": 124053445, + "step": 5766, + "time_per_iteration": 2.6258625984191895 + }, + { + "auxiliary_loss_clip": 0.01180307, + "auxiliary_loss_mlp": 0.01027549, + "balance_loss_clip": 1.04797411, + "balance_loss_mlp": 1.02009225, + "epoch": 0.6934407503156376, + "flos": 24166696268160.0, + "grad_norm": 2.6345778703903626, + "language_loss": 0.72255719, + "learning_rate": 9.073052553761404e-07, + "loss": 0.74463576, + "num_input_tokens_seen": 124072810, + "step": 5767, + "time_per_iteration": 2.6911823749542236 + }, + { + "auxiliary_loss_clip": 0.01203548, + "auxiliary_loss_mlp": 0.01027923, + "balance_loss_clip": 1.04133177, + "balance_loss_mlp": 1.02019787, + "epoch": 0.6935609932062767, + "flos": 20631327120000.0, + "grad_norm": 1.8083069894025845, + "language_loss": 0.77970791, + "learning_rate": 9.066529038054805e-07, + "loss": 0.80202264, + "num_input_tokens_seen": 124092875, + "step": 5768, + "time_per_iteration": 2.7526535987854004 + }, + { + "auxiliary_loss_clip": 0.01185494, + "auxiliary_loss_mlp": 0.01025022, + "balance_loss_clip": 1.0460844, + "balance_loss_mlp": 1.01758933, + "epoch": 0.6936812360969158, + "flos": 18254184019200.0, + "grad_norm": 1.834464808858292, + "language_loss": 0.74148041, + "learning_rate": 9.060007180947071e-07, + "loss": 0.76358557, + "num_input_tokens_seen": 124110930, + "step": 5769, + "time_per_iteration": 2.706660032272339 + }, + { + "auxiliary_loss_clip": 0.01204705, + "auxiliary_loss_mlp": 0.01028208, + "balance_loss_clip": 1.0391953, + "balance_loss_mlp": 1.02065039, + "epoch": 0.6938014789875548, + "flos": 31317336368640.0, + "grad_norm": 2.3665666437814146, + "language_loss": 0.73094594, + "learning_rate": 9.053486983427534e-07, + "loss": 0.7532751, + "num_input_tokens_seen": 124132180, + "step": 5770, + "time_per_iteration": 2.7448723316192627 + }, + { + "auxiliary_loss_clip": 0.01190512, + "auxiliary_loss_mlp": 0.0103167, + "balance_loss_clip": 1.04375398, + "balance_loss_mlp": 1.02434158, + "epoch": 0.6939217218781939, + "flos": 17528429721600.0, + "grad_norm": 1.984948189607988, + "language_loss": 0.70675898, + "learning_rate": 9.046968446485326e-07, + "loss": 0.72898078, + "num_input_tokens_seen": 124150585, + "step": 5771, + "time_per_iteration": 2.6897757053375244 + }, + { + "auxiliary_loss_clip": 0.01183516, + "auxiliary_loss_mlp": 0.01027722, + "balance_loss_clip": 1.04866695, + "balance_loss_mlp": 1.01976454, + "epoch": 0.6940419647688331, + "flos": 18551776199040.0, + "grad_norm": 2.403483621939092, + "language_loss": 0.70771182, + "learning_rate": 9.040451571109295e-07, + "loss": 0.72982424, + "num_input_tokens_seen": 124166205, + "step": 5772, + "time_per_iteration": 2.5457797050476074 + }, + { + "auxiliary_loss_clip": 0.01104518, + "auxiliary_loss_mlp": 0.01001753, + "balance_loss_clip": 1.01753449, + "balance_loss_mlp": 1.00068653, + "epoch": 0.6941622076594721, + "flos": 66926286829440.0, + "grad_norm": 0.8321425855740172, + "language_loss": 0.60471392, + "learning_rate": 9.033936358288042e-07, + "loss": 0.62577659, + "num_input_tokens_seen": 124219940, + "step": 5773, + "time_per_iteration": 3.2880539894104004 + }, + { + "auxiliary_loss_clip": 0.01173086, + "auxiliary_loss_mlp": 0.01024763, + "balance_loss_clip": 1.04978514, + "balance_loss_mlp": 1.01736021, + "epoch": 0.6942824505501112, + "flos": 26578062051840.0, + "grad_norm": 1.6445958225976447, + "language_loss": 0.82274938, + "learning_rate": 9.027422809009937e-07, + "loss": 0.84472787, + "num_input_tokens_seen": 124239885, + "step": 5774, + "time_per_iteration": 2.6726090908050537 + }, + { + "auxiliary_loss_clip": 0.0118169, + "auxiliary_loss_mlp": 0.01027071, + "balance_loss_clip": 1.04564762, + "balance_loss_mlp": 1.01974273, + "epoch": 0.6944026934407503, + "flos": 21248308056960.0, + "grad_norm": 1.671991412372872, + "language_loss": 0.83168763, + "learning_rate": 9.020910924263054e-07, + "loss": 0.85377526, + "num_input_tokens_seen": 124258410, + "step": 5775, + "time_per_iteration": 2.5840086936950684 + }, + { + "auxiliary_loss_clip": 0.01102455, + "auxiliary_loss_mlp": 0.01001784, + "balance_loss_clip": 1.01651239, + "balance_loss_mlp": 1.00065136, + "epoch": 0.6945229363313894, + "flos": 70677191537280.0, + "grad_norm": 0.8957942899908437, + "language_loss": 0.58101159, + "learning_rate": 9.014400705035261e-07, + "loss": 0.60205394, + "num_input_tokens_seen": 124315315, + "step": 5776, + "time_per_iteration": 3.287336826324463 + }, + { + "auxiliary_loss_clip": 0.01172924, + "auxiliary_loss_mlp": 0.01025738, + "balance_loss_clip": 1.05116689, + "balance_loss_mlp": 1.01848459, + "epoch": 0.6946431792220285, + "flos": 18952934267520.0, + "grad_norm": 2.0672947882441433, + "language_loss": 0.76604497, + "learning_rate": 9.00789215231414e-07, + "loss": 0.78803158, + "num_input_tokens_seen": 124333710, + "step": 5777, + "time_per_iteration": 3.4412283897399902 + }, + { + "auxiliary_loss_clip": 0.01193393, + "auxiliary_loss_mlp": 0.01378888, + "balance_loss_clip": 1.04143453, + "balance_loss_mlp": 1.0001595, + "epoch": 0.6947634221126676, + "flos": 20338834671360.0, + "grad_norm": 2.4760827450762686, + "language_loss": 0.82019019, + "learning_rate": 9.001385267087056e-07, + "loss": 0.84591299, + "num_input_tokens_seen": 124352855, + "step": 5778, + "time_per_iteration": 2.667177677154541 + }, + { + "auxiliary_loss_clip": 0.01182752, + "auxiliary_loss_mlp": 0.01021388, + "balance_loss_clip": 1.04859841, + "balance_loss_mlp": 1.0141753, + "epoch": 0.6948836650033067, + "flos": 21833723917440.0, + "grad_norm": 1.527427058271586, + "language_loss": 0.70139432, + "learning_rate": 8.994880050341072e-07, + "loss": 0.72343576, + "num_input_tokens_seen": 124372960, + "step": 5779, + "time_per_iteration": 3.521782398223877 + }, + { + "auxiliary_loss_clip": 0.01178649, + "auxiliary_loss_mlp": 0.01033111, + "balance_loss_clip": 1.0445869, + "balance_loss_mlp": 1.0254066, + "epoch": 0.6950039078939457, + "flos": 23657519024640.0, + "grad_norm": 3.103155071162721, + "language_loss": 0.77499604, + "learning_rate": 8.988376503063026e-07, + "loss": 0.79711366, + "num_input_tokens_seen": 124394220, + "step": 5780, + "time_per_iteration": 2.725458860397339 + }, + { + "auxiliary_loss_clip": 0.01211494, + "auxiliary_loss_mlp": 0.01029755, + "balance_loss_clip": 1.0420568, + "balance_loss_mlp": 1.02217293, + "epoch": 0.6951241507845849, + "flos": 21792462168960.0, + "grad_norm": 2.2117646752466253, + "language_loss": 0.81556785, + "learning_rate": 8.981874626239521e-07, + "loss": 0.83798039, + "num_input_tokens_seen": 124412795, + "step": 5781, + "time_per_iteration": 3.7128312587738037 + }, + { + "auxiliary_loss_clip": 0.01183535, + "auxiliary_loss_mlp": 0.01027367, + "balance_loss_clip": 1.05142808, + "balance_loss_mlp": 1.01965415, + "epoch": 0.695244393675224, + "flos": 14647568244480.0, + "grad_norm": 1.9139582072748613, + "language_loss": 0.88429171, + "learning_rate": 8.975374420856872e-07, + "loss": 0.90640068, + "num_input_tokens_seen": 124429690, + "step": 5782, + "time_per_iteration": 2.630291700363159 + }, + { + "auxiliary_loss_clip": 0.01182889, + "auxiliary_loss_mlp": 0.01028703, + "balance_loss_clip": 1.0402447, + "balance_loss_mlp": 1.0210762, + "epoch": 0.695364636565863, + "flos": 16873203778560.0, + "grad_norm": 2.571231665715132, + "language_loss": 0.7248131, + "learning_rate": 8.968875887901157e-07, + "loss": 0.74692893, + "num_input_tokens_seen": 124447070, + "step": 5783, + "time_per_iteration": 2.6341466903686523 + }, + { + "auxiliary_loss_clip": 0.01184774, + "auxiliary_loss_mlp": 0.01021555, + "balance_loss_clip": 1.04310846, + "balance_loss_mlp": 1.01382482, + "epoch": 0.6954848794565022, + "flos": 19354523299200.0, + "grad_norm": 1.8830654474924589, + "language_loss": 0.62665039, + "learning_rate": 8.9623790283582e-07, + "loss": 0.64871371, + "num_input_tokens_seen": 124464950, + "step": 5784, + "time_per_iteration": 2.6442463397979736 + }, + { + "auxiliary_loss_clip": 0.01195996, + "auxiliary_loss_mlp": 0.01032527, + "balance_loss_clip": 1.04496801, + "balance_loss_mlp": 1.02488577, + "epoch": 0.6956051223471412, + "flos": 18990209606400.0, + "grad_norm": 2.988881236856714, + "language_loss": 0.76607317, + "learning_rate": 8.955883843213561e-07, + "loss": 0.78835839, + "num_input_tokens_seen": 124483965, + "step": 5785, + "time_per_iteration": 3.510504961013794 + }, + { + "auxiliary_loss_clip": 0.01186663, + "auxiliary_loss_mlp": 0.01030573, + "balance_loss_clip": 1.04707086, + "balance_loss_mlp": 1.02246618, + "epoch": 0.6957253652377803, + "flos": 16107229226880.0, + "grad_norm": 1.8678869022118114, + "language_loss": 0.87034506, + "learning_rate": 8.949390333452569e-07, + "loss": 0.89251745, + "num_input_tokens_seen": 124501910, + "step": 5786, + "time_per_iteration": 2.675696849822998 + }, + { + "auxiliary_loss_clip": 0.01169629, + "auxiliary_loss_mlp": 0.01024238, + "balance_loss_clip": 1.04972279, + "balance_loss_mlp": 1.01667392, + "epoch": 0.6958456081284194, + "flos": 29388646569600.0, + "grad_norm": 2.0730629426228284, + "language_loss": 0.67890418, + "learning_rate": 8.942898500060279e-07, + "loss": 0.70084286, + "num_input_tokens_seen": 124521625, + "step": 5787, + "time_per_iteration": 2.656614303588867 + }, + { + "auxiliary_loss_clip": 0.01211725, + "auxiliary_loss_mlp": 0.01026273, + "balance_loss_clip": 1.0439539, + "balance_loss_mlp": 1.01792872, + "epoch": 0.6959658510190585, + "flos": 25154850395520.0, + "grad_norm": 2.670235714828811, + "language_loss": 0.71748745, + "learning_rate": 8.936408344021493e-07, + "loss": 0.73986745, + "num_input_tokens_seen": 124538540, + "step": 5788, + "time_per_iteration": 2.722905397415161 + }, + { + "auxiliary_loss_clip": 0.01200256, + "auxiliary_loss_mlp": 0.01028789, + "balance_loss_clip": 1.05100417, + "balance_loss_mlp": 1.02047992, + "epoch": 0.6960860939096976, + "flos": 42814388759040.0, + "grad_norm": 2.192663852331382, + "language_loss": 0.71143198, + "learning_rate": 8.929919866320765e-07, + "loss": 0.73372245, + "num_input_tokens_seen": 124559355, + "step": 5789, + "time_per_iteration": 2.9360642433166504 + }, + { + "auxiliary_loss_clip": 0.01203629, + "auxiliary_loss_mlp": 0.01378958, + "balance_loss_clip": 1.04456711, + "balance_loss_mlp": 1.00014007, + "epoch": 0.6962063368003367, + "flos": 17566566986880.0, + "grad_norm": 1.884943340544005, + "language_loss": 0.81431895, + "learning_rate": 8.923433067942385e-07, + "loss": 0.84014481, + "num_input_tokens_seen": 124577920, + "step": 5790, + "time_per_iteration": 2.7257370948791504 + }, + { + "auxiliary_loss_clip": 0.01205125, + "auxiliary_loss_mlp": 0.01028572, + "balance_loss_clip": 1.04639232, + "balance_loss_mlp": 1.02110338, + "epoch": 0.6963265796909758, + "flos": 21251648021760.0, + "grad_norm": 1.8944897261632634, + "language_loss": 0.6871109, + "learning_rate": 8.916947949870417e-07, + "loss": 0.70944786, + "num_input_tokens_seen": 124597585, + "step": 5791, + "time_per_iteration": 2.7329111099243164 + }, + { + "auxiliary_loss_clip": 0.01076958, + "auxiliary_loss_mlp": 0.01001926, + "balance_loss_clip": 1.00986016, + "balance_loss_mlp": 1.00071621, + "epoch": 0.6964468225816148, + "flos": 68828295801600.0, + "grad_norm": 0.7447594815349099, + "language_loss": 0.58164036, + "learning_rate": 8.910464513088615e-07, + "loss": 0.60242921, + "num_input_tokens_seen": 124661625, + "step": 5792, + "time_per_iteration": 3.25252628326416 + }, + { + "auxiliary_loss_clip": 0.01181076, + "auxiliary_loss_mlp": 0.01023497, + "balance_loss_clip": 1.04411352, + "balance_loss_mlp": 1.01570702, + "epoch": 0.696567065472254, + "flos": 18950887192320.0, + "grad_norm": 1.9549787190600043, + "language_loss": 0.78079963, + "learning_rate": 8.903982758580542e-07, + "loss": 0.80284536, + "num_input_tokens_seen": 124680565, + "step": 5793, + "time_per_iteration": 2.706817626953125 + }, + { + "auxiliary_loss_clip": 0.01186272, + "auxiliary_loss_mlp": 0.01027692, + "balance_loss_clip": 1.04670334, + "balance_loss_mlp": 1.02016401, + "epoch": 0.696687308362893, + "flos": 22856675345280.0, + "grad_norm": 1.852745814315225, + "language_loss": 0.80406231, + "learning_rate": 8.897502687329457e-07, + "loss": 0.82620198, + "num_input_tokens_seen": 124700365, + "step": 5794, + "time_per_iteration": 2.66623854637146 + }, + { + "auxiliary_loss_clip": 0.01190314, + "auxiliary_loss_mlp": 0.01027545, + "balance_loss_clip": 1.04261565, + "balance_loss_mlp": 1.02037179, + "epoch": 0.6968075512535321, + "flos": 24972926987520.0, + "grad_norm": 2.0481781636517913, + "language_loss": 0.79770434, + "learning_rate": 8.891024300318382e-07, + "loss": 0.81988299, + "num_input_tokens_seen": 124718935, + "step": 5795, + "time_per_iteration": 2.829373836517334 + }, + { + "auxiliary_loss_clip": 0.01184708, + "auxiliary_loss_mlp": 0.01025382, + "balance_loss_clip": 1.04158163, + "balance_loss_mlp": 1.01827669, + "epoch": 0.6969277941441713, + "flos": 21030438113280.0, + "grad_norm": 1.50700902860077, + "language_loss": 0.75647318, + "learning_rate": 8.884547598530103e-07, + "loss": 0.77857411, + "num_input_tokens_seen": 124739505, + "step": 5796, + "time_per_iteration": 2.6951351165771484 + }, + { + "auxiliary_loss_clip": 0.0121082, + "auxiliary_loss_mlp": 0.01025823, + "balance_loss_clip": 1.03839374, + "balance_loss_mlp": 1.01803207, + "epoch": 0.6970480370348103, + "flos": 21579404647680.0, + "grad_norm": 1.7929142873323742, + "language_loss": 0.74917746, + "learning_rate": 8.8780725829471e-07, + "loss": 0.77154392, + "num_input_tokens_seen": 124757410, + "step": 5797, + "time_per_iteration": 2.7870547771453857 + }, + { + "auxiliary_loss_clip": 0.01171792, + "auxiliary_loss_mlp": 0.01026013, + "balance_loss_clip": 1.04807043, + "balance_loss_mlp": 1.01847327, + "epoch": 0.6971682799254494, + "flos": 22419175691520.0, + "grad_norm": 1.9108152434616992, + "language_loss": 0.77783376, + "learning_rate": 8.87159925455165e-07, + "loss": 0.79981184, + "num_input_tokens_seen": 124777240, + "step": 5798, + "time_per_iteration": 2.645925283432007 + }, + { + "auxiliary_loss_clip": 0.01194073, + "auxiliary_loss_mlp": 0.01026959, + "balance_loss_clip": 1.044011, + "balance_loss_mlp": 1.01903486, + "epoch": 0.6972885228160886, + "flos": 20005834659840.0, + "grad_norm": 1.8824817921991595, + "language_loss": 0.73070198, + "learning_rate": 8.865127614325738e-07, + "loss": 0.75291234, + "num_input_tokens_seen": 124795670, + "step": 5799, + "time_per_iteration": 2.6690726280212402 + }, + { + "auxiliary_loss_clip": 0.01178244, + "auxiliary_loss_mlp": 0.01027493, + "balance_loss_clip": 1.04306912, + "balance_loss_mlp": 1.01938105, + "epoch": 0.6974087657067276, + "flos": 37853437656960.0, + "grad_norm": 2.1832448215717273, + "language_loss": 0.66586697, + "learning_rate": 8.85865766325113e-07, + "loss": 0.68792439, + "num_input_tokens_seen": 124819600, + "step": 5800, + "time_per_iteration": 2.819302558898926 + }, + { + "auxiliary_loss_clip": 0.01183942, + "auxiliary_loss_mlp": 0.01025774, + "balance_loss_clip": 1.04622757, + "balance_loss_mlp": 1.01765013, + "epoch": 0.6975290085973667, + "flos": 29489267543040.0, + "grad_norm": 3.0193403708948963, + "language_loss": 0.71904612, + "learning_rate": 8.852189402309287e-07, + "loss": 0.74114323, + "num_input_tokens_seen": 124838785, + "step": 5801, + "time_per_iteration": 2.70170259475708 + }, + { + "auxiliary_loss_clip": 0.01179273, + "auxiliary_loss_mlp": 0.01024619, + "balance_loss_clip": 1.04729128, + "balance_loss_mlp": 1.01650143, + "epoch": 0.6976492514880057, + "flos": 12895630295040.0, + "grad_norm": 2.6126541755303663, + "language_loss": 0.73915011, + "learning_rate": 8.845722832481441e-07, + "loss": 0.76118904, + "num_input_tokens_seen": 124854215, + "step": 5802, + "time_per_iteration": 2.6235435009002686 + }, + { + "auxiliary_loss_clip": 0.01179805, + "auxiliary_loss_mlp": 0.01022304, + "balance_loss_clip": 1.04781163, + "balance_loss_mlp": 1.01384306, + "epoch": 0.6977694943786449, + "flos": 24352929308160.0, + "grad_norm": 2.0762989509586487, + "language_loss": 0.77297723, + "learning_rate": 8.83925795474858e-07, + "loss": 0.79499829, + "num_input_tokens_seen": 124874340, + "step": 5803, + "time_per_iteration": 3.5581414699554443 + }, + { + "auxiliary_loss_clip": 0.01192653, + "auxiliary_loss_mlp": 0.01033288, + "balance_loss_clip": 1.04518449, + "balance_loss_mlp": 1.02496696, + "epoch": 0.6978897372692839, + "flos": 29898470257920.0, + "grad_norm": 2.3395043950268466, + "language_loss": 0.5950647, + "learning_rate": 8.832794770091414e-07, + "loss": 0.61732411, + "num_input_tokens_seen": 124895175, + "step": 5804, + "time_per_iteration": 2.758894920349121 + }, + { + "auxiliary_loss_clip": 0.01193109, + "auxiliary_loss_mlp": 0.01024477, + "balance_loss_clip": 1.04505563, + "balance_loss_mlp": 1.01693439, + "epoch": 0.698009980159923, + "flos": 21761579450880.0, + "grad_norm": 2.3855879028658356, + "language_loss": 0.82702863, + "learning_rate": 8.826333279490401e-07, + "loss": 0.84920448, + "num_input_tokens_seen": 124915810, + "step": 5805, + "time_per_iteration": 3.5702321529388428 + }, + { + "auxiliary_loss_clip": 0.01190116, + "auxiliary_loss_mlp": 0.01023375, + "balance_loss_clip": 1.04659593, + "balance_loss_mlp": 1.01620734, + "epoch": 0.6981302230505622, + "flos": 19857164267520.0, + "grad_norm": 2.2058646768102372, + "language_loss": 0.68108165, + "learning_rate": 8.819873483925748e-07, + "loss": 0.70321655, + "num_input_tokens_seen": 124932930, + "step": 5806, + "time_per_iteration": 2.642123222351074 + }, + { + "auxiliary_loss_clip": 0.01203901, + "auxiliary_loss_mlp": 0.01378965, + "balance_loss_clip": 1.0471859, + "balance_loss_mlp": 1.00012922, + "epoch": 0.6982504659412012, + "flos": 22198648141440.0, + "grad_norm": 3.140482323846704, + "language_loss": 0.74637961, + "learning_rate": 8.81341538437739e-07, + "loss": 0.77220833, + "num_input_tokens_seen": 124951220, + "step": 5807, + "time_per_iteration": 2.7018752098083496 + }, + { + "auxiliary_loss_clip": 0.01193456, + "auxiliary_loss_mlp": 0.01022141, + "balance_loss_clip": 1.04514265, + "balance_loss_mlp": 1.01471996, + "epoch": 0.6983707088318403, + "flos": 35588479708800.0, + "grad_norm": 2.1398992429750106, + "language_loss": 0.68210512, + "learning_rate": 8.80695898182503e-07, + "loss": 0.70426106, + "num_input_tokens_seen": 124972200, + "step": 5808, + "time_per_iteration": 3.6303610801696777 + }, + { + "auxiliary_loss_clip": 0.01076604, + "auxiliary_loss_mlp": 0.01000281, + "balance_loss_clip": 1.01637888, + "balance_loss_mlp": 0.99921399, + "epoch": 0.6984909517224794, + "flos": 65440052760960.0, + "grad_norm": 0.8326945993472599, + "language_loss": 0.65163457, + "learning_rate": 8.800504277248093e-07, + "loss": 0.6724034, + "num_input_tokens_seen": 125036950, + "step": 5809, + "time_per_iteration": 3.2407472133636475 + }, + { + "auxiliary_loss_clip": 0.01196486, + "auxiliary_loss_mlp": 0.01378775, + "balance_loss_clip": 1.05053318, + "balance_loss_mlp": 1.00020218, + "epoch": 0.6986111946131185, + "flos": 18546927863040.0, + "grad_norm": 1.737070760557196, + "language_loss": 0.75178576, + "learning_rate": 8.794051271625753e-07, + "loss": 0.77753836, + "num_input_tokens_seen": 125054585, + "step": 5810, + "time_per_iteration": 2.718165397644043 + }, + { + "auxiliary_loss_clip": 0.01187851, + "auxiliary_loss_mlp": 0.01027692, + "balance_loss_clip": 1.04541504, + "balance_loss_mlp": 1.0202055, + "epoch": 0.6987314375037575, + "flos": 23039173370880.0, + "grad_norm": 1.6054108405734855, + "language_loss": 0.83292472, + "learning_rate": 8.787599965936925e-07, + "loss": 0.85508013, + "num_input_tokens_seen": 125075515, + "step": 5811, + "time_per_iteration": 3.585991144180298 + }, + { + "auxiliary_loss_clip": 0.01189175, + "auxiliary_loss_mlp": 0.01024892, + "balance_loss_clip": 1.04570532, + "balance_loss_mlp": 1.01710761, + "epoch": 0.6988516803943967, + "flos": 38400393029760.0, + "grad_norm": 1.633934373379828, + "language_loss": 0.71895611, + "learning_rate": 8.781150361160261e-07, + "loss": 0.74109685, + "num_input_tokens_seen": 125097425, + "step": 5812, + "time_per_iteration": 2.7999696731567383 + }, + { + "auxiliary_loss_clip": 0.01199593, + "auxiliary_loss_mlp": 0.01027689, + "balance_loss_clip": 1.04523706, + "balance_loss_mlp": 1.0199523, + "epoch": 0.6989719232850358, + "flos": 24096993926400.0, + "grad_norm": 1.8101889339703061, + "language_loss": 0.7355026, + "learning_rate": 8.774702458274181e-07, + "loss": 0.75777543, + "num_input_tokens_seen": 125117830, + "step": 5813, + "time_per_iteration": 2.6757314205169678 + }, + { + "auxiliary_loss_clip": 0.01177339, + "auxiliary_loss_mlp": 0.01022115, + "balance_loss_clip": 1.04560161, + "balance_loss_mlp": 1.0139792, + "epoch": 0.6990921661756748, + "flos": 14866838818560.0, + "grad_norm": 2.9205925094417258, + "language_loss": 0.70963848, + "learning_rate": 8.768256258256799e-07, + "loss": 0.73163307, + "num_input_tokens_seen": 125134455, + "step": 5814, + "time_per_iteration": 2.6136276721954346 + }, + { + "auxiliary_loss_clip": 0.01183463, + "auxiliary_loss_mlp": 0.01026569, + "balance_loss_clip": 1.04798865, + "balance_loss_mlp": 1.01904953, + "epoch": 0.699212409066314, + "flos": 20193719725440.0, + "grad_norm": 2.05173258471981, + "language_loss": 0.7354033, + "learning_rate": 8.76181176208602e-07, + "loss": 0.75750363, + "num_input_tokens_seen": 125152555, + "step": 5815, + "time_per_iteration": 2.691338539123535 + }, + { + "auxiliary_loss_clip": 0.01190167, + "auxiliary_loss_mlp": 0.01026167, + "balance_loss_clip": 1.03778601, + "balance_loss_mlp": 1.0177629, + "epoch": 0.699332651956953, + "flos": 19427888828160.0, + "grad_norm": 2.090994450269299, + "language_loss": 0.73494911, + "learning_rate": 8.755368970739461e-07, + "loss": 0.75711244, + "num_input_tokens_seen": 125171915, + "step": 5816, + "time_per_iteration": 2.7059056758880615 + }, + { + "auxiliary_loss_clip": 0.01202675, + "auxiliary_loss_mlp": 0.01027911, + "balance_loss_clip": 1.04349196, + "balance_loss_mlp": 1.01937521, + "epoch": 0.6994528948475921, + "flos": 16143714466560.0, + "grad_norm": 2.643286977629477, + "language_loss": 0.61406994, + "learning_rate": 8.748927885194479e-07, + "loss": 0.63637573, + "num_input_tokens_seen": 125190220, + "step": 5817, + "time_per_iteration": 2.7001430988311768 + }, + { + "auxiliary_loss_clip": 0.01096469, + "auxiliary_loss_mlp": 0.01000322, + "balance_loss_clip": 1.01408792, + "balance_loss_mlp": 0.99895686, + "epoch": 0.6995731377382313, + "flos": 64952420699520.0, + "grad_norm": 0.875229528841476, + "language_loss": 0.57336378, + "learning_rate": 8.742488506428209e-07, + "loss": 0.59433174, + "num_input_tokens_seen": 125249310, + "step": 5818, + "time_per_iteration": 3.1734561920166016 + }, + { + "auxiliary_loss_clip": 0.01192575, + "auxiliary_loss_mlp": 0.01378609, + "balance_loss_clip": 1.04544926, + "balance_loss_mlp": 1.00010228, + "epoch": 0.6996933806288703, + "flos": 24900136076160.0, + "grad_norm": 1.8037548794432667, + "language_loss": 0.78195828, + "learning_rate": 8.736050835417466e-07, + "loss": 0.80767006, + "num_input_tokens_seen": 125269350, + "step": 5819, + "time_per_iteration": 2.775616407394409 + }, + { + "auxiliary_loss_clip": 0.01184678, + "auxiliary_loss_mlp": 0.01029712, + "balance_loss_clip": 1.04855561, + "balance_loss_mlp": 1.02283096, + "epoch": 0.6998136235195094, + "flos": 20777806782720.0, + "grad_norm": 2.0961643757708686, + "language_loss": 0.60916889, + "learning_rate": 8.729614873138862e-07, + "loss": 0.63131279, + "num_input_tokens_seen": 125286985, + "step": 5820, + "time_per_iteration": 2.5634241104125977 + }, + { + "auxiliary_loss_clip": 0.0121562, + "auxiliary_loss_mlp": 0.01031573, + "balance_loss_clip": 1.0463295, + "balance_loss_mlp": 1.02334177, + "epoch": 0.6999338664101485, + "flos": 23733470332800.0, + "grad_norm": 2.0079641671603006, + "language_loss": 0.77905482, + "learning_rate": 8.723180620568716e-07, + "loss": 0.80152678, + "num_input_tokens_seen": 125306240, + "step": 5821, + "time_per_iteration": 2.753080129623413 + }, + { + "auxiliary_loss_clip": 0.01190434, + "auxiliary_loss_mlp": 0.01022245, + "balance_loss_clip": 1.04581094, + "balance_loss_mlp": 1.01454735, + "epoch": 0.7000541093007876, + "flos": 19864598382720.0, + "grad_norm": 1.814444831213662, + "language_loss": 0.85217142, + "learning_rate": 8.716748078683116e-07, + "loss": 0.87429821, + "num_input_tokens_seen": 125323015, + "step": 5822, + "time_per_iteration": 2.6045944690704346 + }, + { + "auxiliary_loss_clip": 0.01216661, + "auxiliary_loss_mlp": 0.0102941, + "balance_loss_clip": 1.04161942, + "balance_loss_mlp": 1.02082705, + "epoch": 0.7001743521914267, + "flos": 29679056029440.0, + "grad_norm": 1.9185041811448837, + "language_loss": 0.68566179, + "learning_rate": 8.710317248457855e-07, + "loss": 0.70812249, + "num_input_tokens_seen": 125342630, + "step": 5823, + "time_per_iteration": 2.8348262310028076 + }, + { + "auxiliary_loss_clip": 0.0118678, + "auxiliary_loss_mlp": 0.01026371, + "balance_loss_clip": 1.04922438, + "balance_loss_mlp": 1.01864004, + "epoch": 0.7002945950820658, + "flos": 27489762080640.0, + "grad_norm": 2.0447370132685214, + "language_loss": 0.71891224, + "learning_rate": 8.703888130868482e-07, + "loss": 0.74104381, + "num_input_tokens_seen": 125364480, + "step": 5824, + "time_per_iteration": 2.695521354675293 + }, + { + "auxiliary_loss_clip": 0.01195831, + "auxiliary_loss_mlp": 0.01024666, + "balance_loss_clip": 1.04351377, + "balance_loss_mlp": 1.01789474, + "epoch": 0.7004148379727049, + "flos": 22158463800960.0, + "grad_norm": 2.2132603196233536, + "language_loss": 0.81975377, + "learning_rate": 8.697460726890307e-07, + "loss": 0.84195876, + "num_input_tokens_seen": 125381625, + "step": 5825, + "time_per_iteration": 2.6741018295288086 + }, + { + "auxiliary_loss_clip": 0.01191276, + "auxiliary_loss_mlp": 0.01378791, + "balance_loss_clip": 1.04032731, + "balance_loss_mlp": 1.00017321, + "epoch": 0.7005350808633439, + "flos": 19423758764160.0, + "grad_norm": 1.8713680736506852, + "language_loss": 0.9046306, + "learning_rate": 8.691035037498354e-07, + "loss": 0.93033129, + "num_input_tokens_seen": 125397615, + "step": 5826, + "time_per_iteration": 2.7355077266693115 + }, + { + "auxiliary_loss_clip": 0.0118086, + "auxiliary_loss_mlp": 0.01025046, + "balance_loss_clip": 1.04192984, + "balance_loss_mlp": 1.01800346, + "epoch": 0.7006553237539831, + "flos": 23476708938240.0, + "grad_norm": 1.6699673832701636, + "language_loss": 0.72304273, + "learning_rate": 8.684611063667391e-07, + "loss": 0.74510181, + "num_input_tokens_seen": 125418080, + "step": 5827, + "time_per_iteration": 2.6337053775787354 + }, + { + "auxiliary_loss_clip": 0.01179256, + "auxiliary_loss_mlp": 0.01029173, + "balance_loss_clip": 1.04654574, + "balance_loss_mlp": 1.02141833, + "epoch": 0.7007755666446221, + "flos": 31212872640000.0, + "grad_norm": 1.9429392169994595, + "language_loss": 0.76818585, + "learning_rate": 8.678188806371935e-07, + "loss": 0.79027009, + "num_input_tokens_seen": 125440115, + "step": 5828, + "time_per_iteration": 2.7453272342681885 + }, + { + "auxiliary_loss_clip": 0.01177616, + "auxiliary_loss_mlp": 0.01023604, + "balance_loss_clip": 1.04574609, + "balance_loss_mlp": 1.01613605, + "epoch": 0.7008958095352612, + "flos": 18149899858560.0, + "grad_norm": 1.8232291434502068, + "language_loss": 0.85468936, + "learning_rate": 8.671768266586228e-07, + "loss": 0.87670159, + "num_input_tokens_seen": 125458240, + "step": 5829, + "time_per_iteration": 3.531663656234741 + }, + { + "auxiliary_loss_clip": 0.01194387, + "auxiliary_loss_mlp": 0.0102571, + "balance_loss_clip": 1.04315543, + "balance_loss_mlp": 1.01756215, + "epoch": 0.7010160524259004, + "flos": 27452307173760.0, + "grad_norm": 1.72090217080975, + "language_loss": 0.78022313, + "learning_rate": 8.665349445284275e-07, + "loss": 0.80242407, + "num_input_tokens_seen": 125477980, + "step": 5830, + "time_per_iteration": 2.7096517086029053 + }, + { + "auxiliary_loss_clip": 0.01198481, + "auxiliary_loss_mlp": 0.01029844, + "balance_loss_clip": 1.04542589, + "balance_loss_mlp": 1.02219129, + "epoch": 0.7011362953165394, + "flos": 23842064125440.0, + "grad_norm": 1.5077802926116397, + "language_loss": 0.80712783, + "learning_rate": 8.658932343439799e-07, + "loss": 0.82941115, + "num_input_tokens_seen": 125497765, + "step": 5831, + "time_per_iteration": 3.6310856342315674 + }, + { + "auxiliary_loss_clip": 0.01172116, + "auxiliary_loss_mlp": 0.01025265, + "balance_loss_clip": 1.04922462, + "balance_loss_mlp": 1.01773095, + "epoch": 0.7012565382071785, + "flos": 24823430582400.0, + "grad_norm": 2.428969447715841, + "language_loss": 0.77678847, + "learning_rate": 8.65251696202627e-07, + "loss": 0.79876226, + "num_input_tokens_seen": 125514145, + "step": 5832, + "time_per_iteration": 2.529078960418701 + }, + { + "auxiliary_loss_clip": 0.0120215, + "auxiliary_loss_mlp": 0.01028658, + "balance_loss_clip": 1.04747295, + "balance_loss_mlp": 1.0208497, + "epoch": 0.7013767810978175, + "flos": 21397445326080.0, + "grad_norm": 2.859105892873413, + "language_loss": 0.87662184, + "learning_rate": 8.646103302016896e-07, + "loss": 0.89892995, + "num_input_tokens_seen": 125533115, + "step": 5833, + "time_per_iteration": 3.5850865840911865 + }, + { + "auxiliary_loss_clip": 0.01217908, + "auxiliary_loss_mlp": 0.01025754, + "balance_loss_clip": 1.04415917, + "balance_loss_mlp": 1.01800513, + "epoch": 0.7014970239884567, + "flos": 16687150306560.0, + "grad_norm": 1.840096451227004, + "language_loss": 0.88586199, + "learning_rate": 8.639691364384614e-07, + "loss": 0.90829861, + "num_input_tokens_seen": 125550740, + "step": 5834, + "time_per_iteration": 2.631680727005005 + }, + { + "auxiliary_loss_clip": 0.01194389, + "auxiliary_loss_mlp": 0.01025353, + "balance_loss_clip": 1.04882407, + "balance_loss_mlp": 1.01793814, + "epoch": 0.7016172668790958, + "flos": 12568268718720.0, + "grad_norm": 2.0075118447538203, + "language_loss": 0.72375828, + "learning_rate": 8.633281150102136e-07, + "loss": 0.74595571, + "num_input_tokens_seen": 125567590, + "step": 5835, + "time_per_iteration": 2.8659729957580566 + }, + { + "auxiliary_loss_clip": 0.01186622, + "auxiliary_loss_mlp": 0.01028043, + "balance_loss_clip": 1.04573393, + "balance_loss_mlp": 1.02048802, + "epoch": 0.7017375097697348, + "flos": 17452729808640.0, + "grad_norm": 2.413468138330871, + "language_loss": 0.68037868, + "learning_rate": 8.626872660141855e-07, + "loss": 0.70252532, + "num_input_tokens_seen": 125585500, + "step": 5836, + "time_per_iteration": 2.596097230911255 + }, + { + "auxiliary_loss_clip": 0.01207157, + "auxiliary_loss_mlp": 0.01024287, + "balance_loss_clip": 1.04405665, + "balance_loss_mlp": 1.01678228, + "epoch": 0.701857752660374, + "flos": 18513028402560.0, + "grad_norm": 1.7343202628510728, + "language_loss": 0.74820077, + "learning_rate": 8.620465895475957e-07, + "loss": 0.7705152, + "num_input_tokens_seen": 125603720, + "step": 5837, + "time_per_iteration": 3.598318099975586 + }, + { + "auxiliary_loss_clip": 0.01199943, + "auxiliary_loss_mlp": 0.01026836, + "balance_loss_clip": 1.0426836, + "balance_loss_mlp": 1.0193795, + "epoch": 0.701977995551013, + "flos": 24425971614720.0, + "grad_norm": 1.6095061882175963, + "language_loss": 0.75194275, + "learning_rate": 8.614060857076333e-07, + "loss": 0.77421051, + "num_input_tokens_seen": 125624390, + "step": 5838, + "time_per_iteration": 2.74507737159729 + }, + { + "auxiliary_loss_clip": 0.01181466, + "auxiliary_loss_mlp": 0.0103306, + "balance_loss_clip": 1.04388523, + "balance_loss_mlp": 1.02549601, + "epoch": 0.7020982384416521, + "flos": 23002759958400.0, + "grad_norm": 15.304879715064029, + "language_loss": 0.74803889, + "learning_rate": 8.60765754591462e-07, + "loss": 0.77018416, + "num_input_tokens_seen": 125644085, + "step": 5839, + "time_per_iteration": 2.6340134143829346 + }, + { + "auxiliary_loss_clip": 0.01169424, + "auxiliary_loss_mlp": 0.01024879, + "balance_loss_clip": 1.04677153, + "balance_loss_mlp": 1.01748204, + "epoch": 0.7022184813322913, + "flos": 20449080489600.0, + "grad_norm": 2.035814328617099, + "language_loss": 0.72760975, + "learning_rate": 8.601255962962211e-07, + "loss": 0.74955279, + "num_input_tokens_seen": 125663095, + "step": 5840, + "time_per_iteration": 2.6114776134490967 + }, + { + "auxiliary_loss_clip": 0.01189149, + "auxiliary_loss_mlp": 0.01033545, + "balance_loss_clip": 1.04967809, + "balance_loss_mlp": 1.02518797, + "epoch": 0.7023387242229303, + "flos": 19790514581760.0, + "grad_norm": 2.670403832321855, + "language_loss": 0.72673327, + "learning_rate": 8.594856109190194e-07, + "loss": 0.7489602, + "num_input_tokens_seen": 125680125, + "step": 5841, + "time_per_iteration": 2.581066846847534 + }, + { + "auxiliary_loss_clip": 0.01170854, + "auxiliary_loss_mlp": 0.01026517, + "balance_loss_clip": 1.04737735, + "balance_loss_mlp": 1.01894689, + "epoch": 0.7024589671135694, + "flos": 33259278286080.0, + "grad_norm": 5.723791373430322, + "language_loss": 0.69293851, + "learning_rate": 8.588457985569446e-07, + "loss": 0.71491218, + "num_input_tokens_seen": 125703035, + "step": 5842, + "time_per_iteration": 2.716371536254883 + }, + { + "auxiliary_loss_clip": 0.01172791, + "auxiliary_loss_mlp": 0.01033323, + "balance_loss_clip": 1.04908824, + "balance_loss_mlp": 1.02491271, + "epoch": 0.7025792100042085, + "flos": 19098982967040.0, + "grad_norm": 2.136417386184433, + "language_loss": 0.72076368, + "learning_rate": 8.582061593070542e-07, + "loss": 0.74282479, + "num_input_tokens_seen": 125723765, + "step": 5843, + "time_per_iteration": 2.5609469413757324 + }, + { + "auxiliary_loss_clip": 0.01171425, + "auxiliary_loss_mlp": 0.01379071, + "balance_loss_clip": 1.04917395, + "balance_loss_mlp": 1.00011373, + "epoch": 0.7026994528948476, + "flos": 18952611045120.0, + "grad_norm": 2.621447281144109, + "language_loss": 0.76607579, + "learning_rate": 8.57566693266383e-07, + "loss": 0.7915808, + "num_input_tokens_seen": 125741455, + "step": 5844, + "time_per_iteration": 2.5611445903778076 + }, + { + "auxiliary_loss_clip": 0.01192816, + "auxiliary_loss_mlp": 0.01378706, + "balance_loss_clip": 1.04439974, + "balance_loss_mlp": 1.00012612, + "epoch": 0.7028196957854866, + "flos": 19536662188800.0, + "grad_norm": 8.535531987961148, + "language_loss": 0.69520831, + "learning_rate": 8.569274005319354e-07, + "loss": 0.72092348, + "num_input_tokens_seen": 125759855, + "step": 5845, + "time_per_iteration": 2.630706787109375 + }, + { + "auxiliary_loss_clip": 0.01174061, + "auxiliary_loss_mlp": 0.01028475, + "balance_loss_clip": 1.04527378, + "balance_loss_mlp": 1.0211879, + "epoch": 0.7029399386761258, + "flos": 20845318394880.0, + "grad_norm": 1.799560843203349, + "language_loss": 0.79594815, + "learning_rate": 8.562882812006913e-07, + "loss": 0.81797349, + "num_input_tokens_seen": 125777345, + "step": 5846, + "time_per_iteration": 2.6239535808563232 + }, + { + "auxiliary_loss_clip": 0.01170429, + "auxiliary_loss_mlp": 0.01028808, + "balance_loss_clip": 1.04849052, + "balance_loss_mlp": 1.0213635, + "epoch": 0.7030601815667649, + "flos": 22055005653120.0, + "grad_norm": 1.7196562696021747, + "language_loss": 0.77735198, + "learning_rate": 8.556493353696066e-07, + "loss": 0.79934436, + "num_input_tokens_seen": 125796345, + "step": 5847, + "time_per_iteration": 2.582354784011841 + }, + { + "auxiliary_loss_clip": 0.01183545, + "auxiliary_loss_mlp": 0.01379169, + "balance_loss_clip": 1.04826725, + "balance_loss_mlp": 1.0001334, + "epoch": 0.7031804244574039, + "flos": 27198742089600.0, + "grad_norm": 2.151796043568363, + "language_loss": 0.68004632, + "learning_rate": 8.550105631356077e-07, + "loss": 0.70567346, + "num_input_tokens_seen": 125816070, + "step": 5848, + "time_per_iteration": 2.685948371887207 + }, + { + "auxiliary_loss_clip": 0.01187504, + "auxiliary_loss_mlp": 0.01024557, + "balance_loss_clip": 1.04122806, + "balance_loss_mlp": 1.01688027, + "epoch": 0.7033006673480431, + "flos": 22379853277440.0, + "grad_norm": 2.106117203443555, + "language_loss": 0.77009732, + "learning_rate": 8.543719645955961e-07, + "loss": 0.79221785, + "num_input_tokens_seen": 125834400, + "step": 5849, + "time_per_iteration": 2.7232091426849365 + }, + { + "auxiliary_loss_clip": 0.01189892, + "auxiliary_loss_mlp": 0.0103082, + "balance_loss_clip": 1.04471731, + "balance_loss_mlp": 1.02316642, + "epoch": 0.7034209102386821, + "flos": 24715986024960.0, + "grad_norm": 1.6783630515605104, + "language_loss": 0.74652493, + "learning_rate": 8.537335398464467e-07, + "loss": 0.76873201, + "num_input_tokens_seen": 125854720, + "step": 5850, + "time_per_iteration": 2.650259017944336 + }, + { + "auxiliary_loss_clip": 0.01186251, + "auxiliary_loss_mlp": 0.0102654, + "balance_loss_clip": 1.04287803, + "balance_loss_mlp": 1.01833272, + "epoch": 0.7035411531293212, + "flos": 22556174163840.0, + "grad_norm": 2.785797224224517, + "language_loss": 0.8538965, + "learning_rate": 8.53095288985007e-07, + "loss": 0.87602448, + "num_input_tokens_seen": 125868455, + "step": 5851, + "time_per_iteration": 2.6826953887939453 + }, + { + "auxiliary_loss_clip": 0.01168327, + "auxiliary_loss_mlp": 0.01026315, + "balance_loss_clip": 1.04778671, + "balance_loss_mlp": 1.01927614, + "epoch": 0.7036613960199604, + "flos": 22674967418880.0, + "grad_norm": 1.6632434311506297, + "language_loss": 0.82108259, + "learning_rate": 8.524572121081009e-07, + "loss": 0.84302896, + "num_input_tokens_seen": 125888555, + "step": 5852, + "time_per_iteration": 2.5604565143585205 + }, + { + "auxiliary_loss_clip": 0.0118399, + "auxiliary_loss_mlp": 0.01025015, + "balance_loss_clip": 1.04761982, + "balance_loss_mlp": 1.01709378, + "epoch": 0.7037816389105994, + "flos": 22492146170880.0, + "grad_norm": 2.2068820115868855, + "language_loss": 0.62414879, + "learning_rate": 8.518193093125232e-07, + "loss": 0.6462388, + "num_input_tokens_seen": 125907610, + "step": 5853, + "time_per_iteration": 2.627946376800537 + }, + { + "auxiliary_loss_clip": 0.01197321, + "auxiliary_loss_mlp": 0.0102878, + "balance_loss_clip": 1.04862225, + "balance_loss_mlp": 1.02170479, + "epoch": 0.7039018818012385, + "flos": 27087490690560.0, + "grad_norm": 1.63847853358226, + "language_loss": 0.81081772, + "learning_rate": 8.511815806950436e-07, + "loss": 0.83307862, + "num_input_tokens_seen": 125928640, + "step": 5854, + "time_per_iteration": 2.692394733428955 + }, + { + "auxiliary_loss_clip": 0.01175062, + "auxiliary_loss_mlp": 0.01022334, + "balance_loss_clip": 1.04346228, + "balance_loss_mlp": 1.01521719, + "epoch": 0.7040221246918776, + "flos": 17749819198080.0, + "grad_norm": 1.5416744517335035, + "language_loss": 0.7798115, + "learning_rate": 8.505440263524044e-07, + "loss": 0.80178547, + "num_input_tokens_seen": 125947485, + "step": 5855, + "time_per_iteration": 3.5110418796539307 + }, + { + "auxiliary_loss_clip": 0.01181237, + "auxiliary_loss_mlp": 0.01030251, + "balance_loss_clip": 1.04688585, + "balance_loss_mlp": 1.02163839, + "epoch": 0.7041423675825167, + "flos": 16279851012480.0, + "grad_norm": 2.52493267829607, + "language_loss": 0.88332856, + "learning_rate": 8.49906646381322e-07, + "loss": 0.90544331, + "num_input_tokens_seen": 125960320, + "step": 5856, + "time_per_iteration": 2.589916706085205 + }, + { + "auxiliary_loss_clip": 0.01200217, + "auxiliary_loss_mlp": 0.01026559, + "balance_loss_clip": 1.04564607, + "balance_loss_mlp": 1.01946259, + "epoch": 0.7042626104731557, + "flos": 25483181639040.0, + "grad_norm": 1.650168390085659, + "language_loss": 0.72003609, + "learning_rate": 8.492694408784884e-07, + "loss": 0.74230385, + "num_input_tokens_seen": 125980575, + "step": 5857, + "time_per_iteration": 3.5809473991394043 + }, + { + "auxiliary_loss_clip": 0.0118138, + "auxiliary_loss_mlp": 0.01025061, + "balance_loss_clip": 1.0459702, + "balance_loss_mlp": 1.01729441, + "epoch": 0.7043828533637949, + "flos": 17857622891520.0, + "grad_norm": 2.237972452852121, + "language_loss": 0.6230402, + "learning_rate": 8.486324099405642e-07, + "loss": 0.64510459, + "num_input_tokens_seen": 125997420, + "step": 5858, + "time_per_iteration": 2.660468816757202 + }, + { + "auxiliary_loss_clip": 0.01180396, + "auxiliary_loss_mlp": 0.01026947, + "balance_loss_clip": 1.04713428, + "balance_loss_mlp": 1.01925802, + "epoch": 0.704503096254434, + "flos": 29494259533440.0, + "grad_norm": 1.6249449836849932, + "language_loss": 0.74890155, + "learning_rate": 8.479955536641887e-07, + "loss": 0.77097499, + "num_input_tokens_seen": 126018915, + "step": 5859, + "time_per_iteration": 2.654849052429199 + }, + { + "auxiliary_loss_clip": 0.01178666, + "auxiliary_loss_mlp": 0.01026566, + "balance_loss_clip": 1.04093993, + "balance_loss_mlp": 1.01879323, + "epoch": 0.704623339145073, + "flos": 30920739327360.0, + "grad_norm": 1.788856508968012, + "language_loss": 0.65899074, + "learning_rate": 8.473588721459716e-07, + "loss": 0.68104303, + "num_input_tokens_seen": 126038825, + "step": 5860, + "time_per_iteration": 3.604989767074585 + }, + { + "auxiliary_loss_clip": 0.01181296, + "auxiliary_loss_mlp": 0.01033458, + "balance_loss_clip": 1.04953623, + "balance_loss_mlp": 1.02476811, + "epoch": 0.7047435820357122, + "flos": 23914747296000.0, + "grad_norm": 1.9482590964450042, + "language_loss": 0.70357966, + "learning_rate": 8.467223654824967e-07, + "loss": 0.7257272, + "num_input_tokens_seen": 126058280, + "step": 5861, + "time_per_iteration": 2.619006395339966 + }, + { + "auxiliary_loss_clip": 0.01170406, + "auxiliary_loss_mlp": 0.0102736, + "balance_loss_clip": 1.04553986, + "balance_loss_mlp": 1.01973629, + "epoch": 0.7048638249263512, + "flos": 46494010926720.0, + "grad_norm": 1.7994200931450368, + "language_loss": 0.62737823, + "learning_rate": 8.460860337703233e-07, + "loss": 0.64935589, + "num_input_tokens_seen": 126078885, + "step": 5862, + "time_per_iteration": 3.737696409225464 + }, + { + "auxiliary_loss_clip": 0.01180156, + "auxiliary_loss_mlp": 0.01022775, + "balance_loss_clip": 1.04122329, + "balance_loss_mlp": 1.01453209, + "epoch": 0.7049840678169903, + "flos": 21689219502720.0, + "grad_norm": 1.951882150692134, + "language_loss": 0.70604324, + "learning_rate": 8.454498771059797e-07, + "loss": 0.72807264, + "num_input_tokens_seen": 126098260, + "step": 5863, + "time_per_iteration": 2.7100226879119873 + }, + { + "auxiliary_loss_clip": 0.01197883, + "auxiliary_loss_mlp": 0.01027351, + "balance_loss_clip": 1.0406549, + "balance_loss_mlp": 1.01949465, + "epoch": 0.7051043107076294, + "flos": 18405081054720.0, + "grad_norm": 2.136676377115449, + "language_loss": 0.83369601, + "learning_rate": 8.448138955859725e-07, + "loss": 0.85594839, + "num_input_tokens_seen": 126114845, + "step": 5864, + "time_per_iteration": 2.6083035469055176 + }, + { + "auxiliary_loss_clip": 0.01187537, + "auxiliary_loss_mlp": 0.01025124, + "balance_loss_clip": 1.04513943, + "balance_loss_mlp": 1.01736069, + "epoch": 0.7052245535982685, + "flos": 19319043640320.0, + "grad_norm": 1.7383353739500205, + "language_loss": 0.90114737, + "learning_rate": 8.44178089306778e-07, + "loss": 0.92327398, + "num_input_tokens_seen": 126132780, + "step": 5865, + "time_per_iteration": 2.6662192344665527 + }, + { + "auxiliary_loss_clip": 0.01170726, + "auxiliary_loss_mlp": 0.01025142, + "balance_loss_clip": 1.04939985, + "balance_loss_mlp": 1.01732445, + "epoch": 0.7053447964889076, + "flos": 19062138591360.0, + "grad_norm": 1.842132772451775, + "language_loss": 0.7689836, + "learning_rate": 8.4354245836485e-07, + "loss": 0.79094231, + "num_input_tokens_seen": 126151225, + "step": 5866, + "time_per_iteration": 2.5512497425079346 + }, + { + "auxiliary_loss_clip": 0.01200078, + "auxiliary_loss_mlp": 0.01032014, + "balance_loss_clip": 1.04444504, + "balance_loss_mlp": 1.02408695, + "epoch": 0.7054650393795466, + "flos": 27379228953600.0, + "grad_norm": 1.5802454565400885, + "language_loss": 0.72744519, + "learning_rate": 8.429070028566108e-07, + "loss": 0.74976611, + "num_input_tokens_seen": 126172535, + "step": 5867, + "time_per_iteration": 2.8191826343536377 + }, + { + "auxiliary_loss_clip": 0.01179715, + "auxiliary_loss_mlp": 0.01025379, + "balance_loss_clip": 1.04691803, + "balance_loss_mlp": 1.01768351, + "epoch": 0.7055852822701858, + "flos": 16102201322880.0, + "grad_norm": 1.8624790158275188, + "language_loss": 0.75137174, + "learning_rate": 8.422717228784586e-07, + "loss": 0.7734226, + "num_input_tokens_seen": 126189410, + "step": 5868, + "time_per_iteration": 2.544548273086548 + }, + { + "auxiliary_loss_clip": 0.01205443, + "auxiliary_loss_mlp": 0.01028395, + "balance_loss_clip": 1.04504037, + "balance_loss_mlp": 1.0204618, + "epoch": 0.7057055251608249, + "flos": 11692299744000.0, + "grad_norm": 2.0337857785948503, + "language_loss": 0.69191343, + "learning_rate": 8.416366185267663e-07, + "loss": 0.71425182, + "num_input_tokens_seen": 126206910, + "step": 5869, + "time_per_iteration": 2.6831860542297363 + }, + { + "auxiliary_loss_clip": 0.01178947, + "auxiliary_loss_mlp": 0.01026036, + "balance_loss_clip": 1.04548192, + "balance_loss_mlp": 1.01800418, + "epoch": 0.7058257680514639, + "flos": 22711560399360.0, + "grad_norm": 1.7217722507718218, + "language_loss": 0.77774715, + "learning_rate": 8.410016898978778e-07, + "loss": 0.799797, + "num_input_tokens_seen": 126224385, + "step": 5870, + "time_per_iteration": 2.6080410480499268 + }, + { + "auxiliary_loss_clip": 0.01204797, + "auxiliary_loss_mlp": 0.01030464, + "balance_loss_clip": 1.04403949, + "balance_loss_mlp": 1.02318025, + "epoch": 0.7059460109421031, + "flos": 17529543043200.0, + "grad_norm": 1.887812358690168, + "language_loss": 0.78945857, + "learning_rate": 8.403669370881115e-07, + "loss": 0.81181115, + "num_input_tokens_seen": 126243120, + "step": 5871, + "time_per_iteration": 2.7874114513397217 + }, + { + "auxiliary_loss_clip": 0.01170058, + "auxiliary_loss_mlp": 0.01026443, + "balance_loss_clip": 1.04864383, + "balance_loss_mlp": 1.0191772, + "epoch": 0.7060662538327421, + "flos": 23544687427200.0, + "grad_norm": 2.110763687832023, + "language_loss": 0.78823292, + "learning_rate": 8.397323601937587e-07, + "loss": 0.81019795, + "num_input_tokens_seen": 126263020, + "step": 5872, + "time_per_iteration": 2.6191744804382324 + }, + { + "auxiliary_loss_clip": 0.01191132, + "auxiliary_loss_mlp": 0.01027672, + "balance_loss_clip": 1.04428625, + "balance_loss_mlp": 1.01966119, + "epoch": 0.7061864967233812, + "flos": 30260736875520.0, + "grad_norm": 2.811927376127759, + "language_loss": 0.77126658, + "learning_rate": 8.390979593110838e-07, + "loss": 0.79345459, + "num_input_tokens_seen": 126285150, + "step": 5873, + "time_per_iteration": 2.7622969150543213 + }, + { + "auxiliary_loss_clip": 0.01195294, + "auxiliary_loss_mlp": 0.01024702, + "balance_loss_clip": 1.04802012, + "balance_loss_mlp": 1.01679587, + "epoch": 0.7063067396140204, + "flos": 20701460424960.0, + "grad_norm": 2.077352350448865, + "language_loss": 0.8169598, + "learning_rate": 8.384637345363262e-07, + "loss": 0.83915973, + "num_input_tokens_seen": 126304340, + "step": 5874, + "time_per_iteration": 2.6872291564941406 + }, + { + "auxiliary_loss_clip": 0.01175211, + "auxiliary_loss_mlp": 0.01030516, + "balance_loss_clip": 1.04026651, + "balance_loss_mlp": 1.02189684, + "epoch": 0.7064269825046594, + "flos": 32266168081920.0, + "grad_norm": 1.7358696012901864, + "language_loss": 0.76876605, + "learning_rate": 8.378296859656964e-07, + "loss": 0.79082334, + "num_input_tokens_seen": 126325495, + "step": 5875, + "time_per_iteration": 2.6639933586120605 + }, + { + "auxiliary_loss_clip": 0.01190091, + "auxiliary_loss_mlp": 0.01024126, + "balance_loss_clip": 1.046453, + "balance_loss_mlp": 1.01700282, + "epoch": 0.7065472253952985, + "flos": 30227124723840.0, + "grad_norm": 1.9157831792590314, + "language_loss": 0.68562543, + "learning_rate": 8.371958136953792e-07, + "loss": 0.70776761, + "num_input_tokens_seen": 126345525, + "step": 5876, + "time_per_iteration": 2.6350021362304688 + }, + { + "auxiliary_loss_clip": 0.01203748, + "auxiliary_loss_mlp": 0.01024385, + "balance_loss_clip": 1.04396057, + "balance_loss_mlp": 1.01571238, + "epoch": 0.7066674682859376, + "flos": 16216720859520.0, + "grad_norm": 3.6979782547066513, + "language_loss": 0.66218531, + "learning_rate": 8.365621178215326e-07, + "loss": 0.68446672, + "num_input_tokens_seen": 126361995, + "step": 5877, + "time_per_iteration": 2.5308175086975098 + }, + { + "auxiliary_loss_clip": 0.01172244, + "auxiliary_loss_mlp": 0.01023871, + "balance_loss_clip": 1.04472315, + "balance_loss_mlp": 1.01604521, + "epoch": 0.7067877111765767, + "flos": 14830461319680.0, + "grad_norm": 2.1820553429454974, + "language_loss": 0.7468468, + "learning_rate": 8.359285984402871e-07, + "loss": 0.76880789, + "num_input_tokens_seen": 126379260, + "step": 5878, + "time_per_iteration": 2.4919803142547607 + }, + { + "auxiliary_loss_clip": 0.01184571, + "auxiliary_loss_mlp": 0.01021313, + "balance_loss_clip": 1.04617655, + "balance_loss_mlp": 1.01384759, + "epoch": 0.7069079540672157, + "flos": 25440196037760.0, + "grad_norm": 2.0128399619417556, + "language_loss": 0.74386859, + "learning_rate": 8.352952556477489e-07, + "loss": 0.76592743, + "num_input_tokens_seen": 126397170, + "step": 5879, + "time_per_iteration": 2.5586228370666504 + }, + { + "auxiliary_loss_clip": 0.01179353, + "auxiliary_loss_mlp": 0.01025345, + "balance_loss_clip": 1.04781199, + "balance_loss_mlp": 1.01805854, + "epoch": 0.7070281969578549, + "flos": 24607751368320.0, + "grad_norm": 1.8144791020441267, + "language_loss": 0.76542974, + "learning_rate": 8.34662089539993e-07, + "loss": 0.78747666, + "num_input_tokens_seen": 126416680, + "step": 5880, + "time_per_iteration": 2.522188901901245 + }, + { + "auxiliary_loss_clip": 0.01169585, + "auxiliary_loss_mlp": 0.01027414, + "balance_loss_clip": 1.04940188, + "balance_loss_mlp": 1.01977241, + "epoch": 0.707148439848494, + "flos": 26724469887360.0, + "grad_norm": 2.058007961903343, + "language_loss": 0.79148936, + "learning_rate": 8.340291002130722e-07, + "loss": 0.81345934, + "num_input_tokens_seen": 126435870, + "step": 5881, + "time_per_iteration": 3.3855974674224854 + }, + { + "auxiliary_loss_clip": 0.01174331, + "auxiliary_loss_mlp": 0.01030845, + "balance_loss_clip": 1.05063403, + "balance_loss_mlp": 1.02314401, + "epoch": 0.707268682739133, + "flos": 15085750256640.0, + "grad_norm": 2.5209969189356807, + "language_loss": 0.79750472, + "learning_rate": 8.3339628776301e-07, + "loss": 0.81955647, + "num_input_tokens_seen": 126454010, + "step": 5882, + "time_per_iteration": 2.593395233154297 + }, + { + "auxiliary_loss_clip": 0.01168953, + "auxiliary_loss_mlp": 0.01026426, + "balance_loss_clip": 1.0477109, + "balance_loss_mlp": 1.01903844, + "epoch": 0.7073889256297722, + "flos": 34313148345600.0, + "grad_norm": 2.2016533815485193, + "language_loss": 0.56906927, + "learning_rate": 8.327636522858033e-07, + "loss": 0.59102303, + "num_input_tokens_seen": 126473615, + "step": 5883, + "time_per_iteration": 3.2656350135803223 + }, + { + "auxiliary_loss_clip": 0.01208381, + "auxiliary_loss_mlp": 0.01025873, + "balance_loss_clip": 1.04627347, + "balance_loss_mlp": 1.0180409, + "epoch": 0.7075091685204112, + "flos": 20083940784000.0, + "grad_norm": 3.1818475198117677, + "language_loss": 0.7714799, + "learning_rate": 8.321311938774225e-07, + "loss": 0.79382241, + "num_input_tokens_seen": 126492705, + "step": 5884, + "time_per_iteration": 2.6101033687591553 + }, + { + "auxiliary_loss_clip": 0.01172289, + "auxiliary_loss_mlp": 0.01029082, + "balance_loss_clip": 1.0486244, + "balance_loss_mlp": 1.02137566, + "epoch": 0.7076294114110503, + "flos": 20777124424320.0, + "grad_norm": 2.0295311435614383, + "language_loss": 0.79521441, + "learning_rate": 8.314989126338104e-07, + "loss": 0.81722808, + "num_input_tokens_seen": 126512715, + "step": 5885, + "time_per_iteration": 2.511949062347412 + }, + { + "auxiliary_loss_clip": 0.01180566, + "auxiliary_loss_mlp": 0.01027976, + "balance_loss_clip": 1.04604316, + "balance_loss_mlp": 1.0198102, + "epoch": 0.7077496543016895, + "flos": 17967689141760.0, + "grad_norm": 1.7116824615249984, + "language_loss": 0.84519839, + "learning_rate": 8.308668086508847e-07, + "loss": 0.86728382, + "num_input_tokens_seen": 126530795, + "step": 5886, + "time_per_iteration": 3.4403631687164307 + }, + { + "auxiliary_loss_clip": 0.0119784, + "auxiliary_loss_mlp": 0.01028797, + "balance_loss_clip": 1.04158139, + "balance_loss_mlp": 1.02056003, + "epoch": 0.7078698971923285, + "flos": 45478098564480.0, + "grad_norm": 1.725899743126666, + "language_loss": 0.73643947, + "learning_rate": 8.302348820245342e-07, + "loss": 0.75870585, + "num_input_tokens_seen": 126553360, + "step": 5887, + "time_per_iteration": 2.8278071880340576 + }, + { + "auxiliary_loss_clip": 0.01194177, + "auxiliary_loss_mlp": 0.01030144, + "balance_loss_clip": 1.04182792, + "balance_loss_mlp": 1.02231753, + "epoch": 0.7079901400829676, + "flos": 26943704547840.0, + "grad_norm": 4.5744062941429275, + "language_loss": 0.7026403, + "learning_rate": 8.296031328506232e-07, + "loss": 0.72488356, + "num_input_tokens_seen": 126573110, + "step": 5888, + "time_per_iteration": 3.4872782230377197 + }, + { + "auxiliary_loss_clip": 0.01189799, + "auxiliary_loss_mlp": 0.01024877, + "balance_loss_clip": 1.04608488, + "balance_loss_mlp": 1.01650858, + "epoch": 0.7081103829736067, + "flos": 24423206267520.0, + "grad_norm": 1.840881392927448, + "language_loss": 0.75388443, + "learning_rate": 8.289715612249857e-07, + "loss": 0.7760312, + "num_input_tokens_seen": 126593725, + "step": 5889, + "time_per_iteration": 2.6822776794433594 + }, + { + "auxiliary_loss_clip": 0.01183311, + "auxiliary_loss_mlp": 0.01032193, + "balance_loss_clip": 1.0457381, + "balance_loss_mlp": 1.02450085, + "epoch": 0.7082306258642458, + "flos": 18543300589440.0, + "grad_norm": 3.317872124167484, + "language_loss": 0.77977604, + "learning_rate": 8.283401672434305e-07, + "loss": 0.80193102, + "num_input_tokens_seen": 126608950, + "step": 5890, + "time_per_iteration": 2.609628915786743 + }, + { + "auxiliary_loss_clip": 0.01185931, + "auxiliary_loss_mlp": 0.01024517, + "balance_loss_clip": 1.04659212, + "balance_loss_mlp": 1.0175252, + "epoch": 0.7083508687548848, + "flos": 23477534951040.0, + "grad_norm": 1.945010613172775, + "language_loss": 0.70178533, + "learning_rate": 8.277089510017412e-07, + "loss": 0.72388983, + "num_input_tokens_seen": 126629755, + "step": 5891, + "time_per_iteration": 2.6619465351104736 + }, + { + "auxiliary_loss_clip": 0.01187156, + "auxiliary_loss_mlp": 0.01024226, + "balance_loss_clip": 1.04721141, + "balance_loss_mlp": 1.0166502, + "epoch": 0.708471111645524, + "flos": 22419463000320.0, + "grad_norm": 1.796611282182915, + "language_loss": 0.82455903, + "learning_rate": 8.270779125956719e-07, + "loss": 0.84667289, + "num_input_tokens_seen": 126650135, + "step": 5892, + "time_per_iteration": 2.7469987869262695 + }, + { + "auxiliary_loss_clip": 0.01198796, + "auxiliary_loss_mlp": 0.01028373, + "balance_loss_clip": 1.04111409, + "balance_loss_mlp": 1.02017713, + "epoch": 0.7085913545361631, + "flos": 20922885815040.0, + "grad_norm": 2.7119440077890613, + "language_loss": 0.79925621, + "learning_rate": 8.264470521209505e-07, + "loss": 0.82152784, + "num_input_tokens_seen": 126668500, + "step": 5893, + "time_per_iteration": 2.7212939262390137 + }, + { + "auxiliary_loss_clip": 0.01167271, + "auxiliary_loss_mlp": 0.01024491, + "balance_loss_clip": 1.04363191, + "balance_loss_mlp": 1.01683819, + "epoch": 0.7087115974268021, + "flos": 15012384727680.0, + "grad_norm": 2.3743652362239804, + "language_loss": 0.76453388, + "learning_rate": 8.258163696732785e-07, + "loss": 0.78645146, + "num_input_tokens_seen": 126686090, + "step": 5894, + "time_per_iteration": 2.7676148414611816 + }, + { + "auxiliary_loss_clip": 0.01172678, + "auxiliary_loss_mlp": 0.01025732, + "balance_loss_clip": 1.04503322, + "balance_loss_mlp": 1.01811719, + "epoch": 0.7088318403174413, + "flos": 21539040739200.0, + "grad_norm": 2.3276419979541694, + "language_loss": 0.77028513, + "learning_rate": 8.251858653483288e-07, + "loss": 0.79226923, + "num_input_tokens_seen": 126704255, + "step": 5895, + "time_per_iteration": 2.603710174560547 + }, + { + "auxiliary_loss_clip": 0.01182823, + "auxiliary_loss_mlp": 0.01025109, + "balance_loss_clip": 1.04844522, + "balance_loss_mlp": 1.01683021, + "epoch": 0.7089520832080803, + "flos": 15516785462400.0, + "grad_norm": 2.5101950814477116, + "language_loss": 0.85859692, + "learning_rate": 8.245555392417501e-07, + "loss": 0.88067627, + "num_input_tokens_seen": 126718910, + "step": 5896, + "time_per_iteration": 2.567660093307495 + }, + { + "auxiliary_loss_clip": 0.01187247, + "auxiliary_loss_mlp": 0.01025708, + "balance_loss_clip": 1.03899336, + "balance_loss_mlp": 1.01794171, + "epoch": 0.7090723260987194, + "flos": 20412667077120.0, + "grad_norm": 2.245565194428913, + "language_loss": 0.7898888, + "learning_rate": 8.239253914491613e-07, + "loss": 0.81201839, + "num_input_tokens_seen": 126737235, + "step": 5897, + "time_per_iteration": 2.6895227432250977 + }, + { + "auxiliary_loss_clip": 0.01189267, + "auxiliary_loss_mlp": 0.01024602, + "balance_loss_clip": 1.04506135, + "balance_loss_mlp": 1.01669288, + "epoch": 0.7091925689893585, + "flos": 25668337271040.0, + "grad_norm": 2.39577025777193, + "language_loss": 0.75134718, + "learning_rate": 8.232954220661556e-07, + "loss": 0.7734859, + "num_input_tokens_seen": 126759970, + "step": 5898, + "time_per_iteration": 2.7729620933532715 + }, + { + "auxiliary_loss_clip": 0.01171362, + "auxiliary_loss_mlp": 0.01027439, + "balance_loss_clip": 1.05102706, + "balance_loss_mlp": 1.01985765, + "epoch": 0.7093128118799976, + "flos": 24206629213440.0, + "grad_norm": 2.5382496368243532, + "language_loss": 0.69817162, + "learning_rate": 8.226656311882989e-07, + "loss": 0.72015965, + "num_input_tokens_seen": 126779280, + "step": 5899, + "time_per_iteration": 2.6175742149353027 + }, + { + "auxiliary_loss_clip": 0.01178966, + "auxiliary_loss_mlp": 0.01021972, + "balance_loss_clip": 1.04909039, + "balance_loss_mlp": 1.0150342, + "epoch": 0.7094330547706367, + "flos": 16646786398080.0, + "grad_norm": 2.1121713559089517, + "language_loss": 0.77110744, + "learning_rate": 8.22036018911129e-07, + "loss": 0.79311681, + "num_input_tokens_seen": 126797310, + "step": 5900, + "time_per_iteration": 2.627607822418213 + }, + { + "auxiliary_loss_clip": 0.01173708, + "auxiliary_loss_mlp": 0.01022369, + "balance_loss_clip": 1.04727602, + "balance_loss_mlp": 1.01424456, + "epoch": 0.7095532976612757, + "flos": 16283370545280.0, + "grad_norm": 2.2942733326250817, + "language_loss": 0.8036598, + "learning_rate": 8.214065853301599e-07, + "loss": 0.82562053, + "num_input_tokens_seen": 126812840, + "step": 5901, + "time_per_iteration": 2.6785361766815186 + }, + { + "auxiliary_loss_clip": 0.01078387, + "auxiliary_loss_mlp": 0.01002927, + "balance_loss_clip": 1.01103759, + "balance_loss_mlp": 1.00174642, + "epoch": 0.7096735405519149, + "flos": 70722080559360.0, + "grad_norm": 0.8188276401170796, + "language_loss": 0.58197331, + "learning_rate": 8.207773305408734e-07, + "loss": 0.60278642, + "num_input_tokens_seen": 126880060, + "step": 5902, + "time_per_iteration": 3.4031410217285156 + }, + { + "auxiliary_loss_clip": 0.0121326, + "auxiliary_loss_mlp": 0.01030808, + "balance_loss_clip": 1.04290295, + "balance_loss_mlp": 1.02284491, + "epoch": 0.709793783442554, + "flos": 23621500661760.0, + "grad_norm": 2.0061702850862373, + "language_loss": 0.79844689, + "learning_rate": 8.201482546387288e-07, + "loss": 0.82088757, + "num_input_tokens_seen": 126899535, + "step": 5903, + "time_per_iteration": 2.744624376296997 + }, + { + "auxiliary_loss_clip": 0.01179473, + "auxiliary_loss_mlp": 0.01026602, + "balance_loss_clip": 1.04809332, + "balance_loss_mlp": 1.01904464, + "epoch": 0.709914026333193, + "flos": 25993472204160.0, + "grad_norm": 1.6150394089073656, + "language_loss": 0.91817474, + "learning_rate": 8.195193577191553e-07, + "loss": 0.9402355, + "num_input_tokens_seen": 126921365, + "step": 5904, + "time_per_iteration": 2.660336971282959 + }, + { + "auxiliary_loss_clip": 0.01194836, + "auxiliary_loss_mlp": 0.01378782, + "balance_loss_clip": 1.04433012, + "balance_loss_mlp": 1.00012267, + "epoch": 0.7100342692238322, + "flos": 24861531934080.0, + "grad_norm": 1.8719998949460042, + "language_loss": 0.84654248, + "learning_rate": 8.188906398775579e-07, + "loss": 0.87227869, + "num_input_tokens_seen": 126941910, + "step": 5905, + "time_per_iteration": 2.8635647296905518 + }, + { + "auxiliary_loss_clip": 0.01172128, + "auxiliary_loss_mlp": 0.01378826, + "balance_loss_clip": 1.04806638, + "balance_loss_mlp": 1.0001893, + "epoch": 0.7101545121144712, + "flos": 24932203943040.0, + "grad_norm": 1.8160368374505782, + "language_loss": 0.68769854, + "learning_rate": 8.18262101209311e-07, + "loss": 0.71320808, + "num_input_tokens_seen": 126961120, + "step": 5906, + "time_per_iteration": 2.551218032836914 + }, + { + "auxiliary_loss_clip": 0.01182947, + "auxiliary_loss_mlp": 0.01023714, + "balance_loss_clip": 1.04703283, + "balance_loss_mlp": 1.01637077, + "epoch": 0.7102747550051103, + "flos": 23768842250880.0, + "grad_norm": 1.8307180137298773, + "language_loss": 0.69901025, + "learning_rate": 8.176337418097626e-07, + "loss": 0.72107685, + "num_input_tokens_seen": 126981590, + "step": 5907, + "time_per_iteration": 3.449807643890381 + }, + { + "auxiliary_loss_clip": 0.01178336, + "auxiliary_loss_mlp": 0.01378629, + "balance_loss_clip": 1.04830861, + "balance_loss_mlp": 1.00017452, + "epoch": 0.7103949978957494, + "flos": 15303907509120.0, + "grad_norm": 2.1892572608976564, + "language_loss": 0.79750264, + "learning_rate": 8.170055617742364e-07, + "loss": 0.82307231, + "num_input_tokens_seen": 126998870, + "step": 5908, + "time_per_iteration": 2.574227809906006 + }, + { + "auxiliary_loss_clip": 0.01181211, + "auxiliary_loss_mlp": 0.01027685, + "balance_loss_clip": 1.04404569, + "balance_loss_mlp": 1.02049696, + "epoch": 0.7105152407863885, + "flos": 22638805401600.0, + "grad_norm": 1.9313657321534756, + "language_loss": 0.70621181, + "learning_rate": 8.163775611980252e-07, + "loss": 0.72830081, + "num_input_tokens_seen": 127017980, + "step": 5909, + "time_per_iteration": 3.5074291229248047 + }, + { + "auxiliary_loss_clip": 0.01188776, + "auxiliary_loss_mlp": 0.01023643, + "balance_loss_clip": 1.04711843, + "balance_loss_mlp": 1.01594174, + "epoch": 0.7106354836770276, + "flos": 17238594879360.0, + "grad_norm": 1.736730784045532, + "language_loss": 0.78882813, + "learning_rate": 8.157497401763982e-07, + "loss": 0.81095231, + "num_input_tokens_seen": 127035645, + "step": 5910, + "time_per_iteration": 2.677865982055664 + }, + { + "auxiliary_loss_clip": 0.01178082, + "auxiliary_loss_mlp": 0.01027851, + "balance_loss_clip": 1.04619241, + "balance_loss_mlp": 1.02062654, + "epoch": 0.7107557265676667, + "flos": 20193647898240.0, + "grad_norm": 2.2738935911173974, + "language_loss": 0.77726585, + "learning_rate": 8.151220988045935e-07, + "loss": 0.79932511, + "num_input_tokens_seen": 127054900, + "step": 5911, + "time_per_iteration": 2.557213306427002 + }, + { + "auxiliary_loss_clip": 0.01178973, + "auxiliary_loss_mlp": 0.01024166, + "balance_loss_clip": 1.04618061, + "balance_loss_mlp": 1.01727581, + "epoch": 0.7108759694583058, + "flos": 21507080613120.0, + "grad_norm": 1.8051333207286127, + "language_loss": 0.82897449, + "learning_rate": 8.144946371778234e-07, + "loss": 0.85100591, + "num_input_tokens_seen": 127075010, + "step": 5912, + "time_per_iteration": 3.5332095623016357 + }, + { + "auxiliary_loss_clip": 0.01188187, + "auxiliary_loss_mlp": 0.01379066, + "balance_loss_clip": 1.04668355, + "balance_loss_mlp": 1.00017977, + "epoch": 0.7109962123489448, + "flos": 24061909317120.0, + "grad_norm": 1.9849744032681633, + "language_loss": 0.78009272, + "learning_rate": 8.138673553912751e-07, + "loss": 0.80576521, + "num_input_tokens_seen": 127095570, + "step": 5913, + "time_per_iteration": 2.6395058631896973 + }, + { + "auxiliary_loss_clip": 0.0120224, + "auxiliary_loss_mlp": 0.0103011, + "balance_loss_clip": 1.04139757, + "balance_loss_mlp": 1.02277303, + "epoch": 0.711116455239584, + "flos": 30480474326400.0, + "grad_norm": 2.243959337102154, + "language_loss": 0.56463146, + "learning_rate": 8.132402535401059e-07, + "loss": 0.58695495, + "num_input_tokens_seen": 127116825, + "step": 5914, + "time_per_iteration": 2.7578296661376953 + }, + { + "auxiliary_loss_clip": 0.01175589, + "auxiliary_loss_mlp": 0.01021973, + "balance_loss_clip": 1.04665947, + "balance_loss_mlp": 1.01459992, + "epoch": 0.711236698130223, + "flos": 25045610158080.0, + "grad_norm": 1.832160597242845, + "language_loss": 0.74149686, + "learning_rate": 8.126133317194465e-07, + "loss": 0.7634725, + "num_input_tokens_seen": 127137015, + "step": 5915, + "time_per_iteration": 3.7418015003204346 + }, + { + "auxiliary_loss_clip": 0.01219267, + "auxiliary_loss_mlp": 0.01025766, + "balance_loss_clip": 1.03834927, + "balance_loss_mlp": 1.01762962, + "epoch": 0.7113569410208621, + "flos": 24206701040640.0, + "grad_norm": 2.7715199461541444, + "language_loss": 0.7412433, + "learning_rate": 8.11986590024401e-07, + "loss": 0.76369357, + "num_input_tokens_seen": 127156755, + "step": 5916, + "time_per_iteration": 2.726177453994751 + }, + { + "auxiliary_loss_clip": 0.01190203, + "auxiliary_loss_mlp": 0.01022669, + "balance_loss_clip": 1.04606414, + "balance_loss_mlp": 1.01523066, + "epoch": 0.7114771839115013, + "flos": 35439306526080.0, + "grad_norm": 1.5791612829683415, + "language_loss": 0.69055635, + "learning_rate": 8.113600285500442e-07, + "loss": 0.71268505, + "num_input_tokens_seen": 127176965, + "step": 5917, + "time_per_iteration": 2.7654428482055664 + }, + { + "auxiliary_loss_clip": 0.01170181, + "auxiliary_loss_mlp": 0.01022617, + "balance_loss_clip": 1.04728532, + "balance_loss_mlp": 1.01491916, + "epoch": 0.7115974268021403, + "flos": 21099458096640.0, + "grad_norm": 1.741992304994068, + "language_loss": 0.73896837, + "learning_rate": 8.107336473914268e-07, + "loss": 0.76089638, + "num_input_tokens_seen": 127195595, + "step": 5918, + "time_per_iteration": 2.5435173511505127 + }, + { + "auxiliary_loss_clip": 0.01087877, + "auxiliary_loss_mlp": 0.01003655, + "balance_loss_clip": 1.01247048, + "balance_loss_mlp": 1.00243878, + "epoch": 0.7117176696927794, + "flos": 56752866616320.0, + "grad_norm": 0.76850573751256, + "language_loss": 0.55702877, + "learning_rate": 8.101074466435694e-07, + "loss": 0.57794404, + "num_input_tokens_seen": 127255070, + "step": 5919, + "time_per_iteration": 3.1409120559692383 + }, + { + "auxiliary_loss_clip": 0.01170998, + "auxiliary_loss_mlp": 0.01030778, + "balance_loss_clip": 1.04476047, + "balance_loss_mlp": 1.02334261, + "epoch": 0.7118379125834186, + "flos": 15925269905280.0, + "grad_norm": 1.6264014265452067, + "language_loss": 0.68040812, + "learning_rate": 8.094814264014662e-07, + "loss": 0.7024259, + "num_input_tokens_seen": 127273825, + "step": 5920, + "time_per_iteration": 2.567075490951538 + }, + { + "auxiliary_loss_clip": 0.01172823, + "auxiliary_loss_mlp": 0.01027426, + "balance_loss_clip": 1.0479517, + "balance_loss_mlp": 1.01947451, + "epoch": 0.7119581554740576, + "flos": 20193360589440.0, + "grad_norm": 2.0591307086977686, + "language_loss": 0.81264478, + "learning_rate": 8.088555867600844e-07, + "loss": 0.8346473, + "num_input_tokens_seen": 127289990, + "step": 5921, + "time_per_iteration": 2.560896873474121 + }, + { + "auxiliary_loss_clip": 0.01191023, + "auxiliary_loss_mlp": 0.01028297, + "balance_loss_clip": 1.04152393, + "balance_loss_mlp": 1.02118874, + "epoch": 0.7120783983646967, + "flos": 34715383822080.0, + "grad_norm": 2.053038960156224, + "language_loss": 0.60759437, + "learning_rate": 8.08229927814362e-07, + "loss": 0.6297875, + "num_input_tokens_seen": 127312880, + "step": 5922, + "time_per_iteration": 2.8134946823120117 + }, + { + "auxiliary_loss_clip": 0.01195045, + "auxiliary_loss_mlp": 0.0102642, + "balance_loss_clip": 1.04310369, + "balance_loss_mlp": 1.01891303, + "epoch": 0.7121986412553358, + "flos": 26359114700160.0, + "grad_norm": 1.6855208551517313, + "language_loss": 0.6485185, + "learning_rate": 8.076044496592134e-07, + "loss": 0.67073315, + "num_input_tokens_seen": 127334730, + "step": 5923, + "time_per_iteration": 2.721517324447632 + }, + { + "auxiliary_loss_clip": 0.01186403, + "auxiliary_loss_mlp": 0.01026305, + "balance_loss_clip": 1.04568934, + "balance_loss_mlp": 1.01845515, + "epoch": 0.7123188841459749, + "flos": 11145344371200.0, + "grad_norm": 2.6685457834874104, + "language_loss": 0.78170872, + "learning_rate": 8.069791523895204e-07, + "loss": 0.80383581, + "num_input_tokens_seen": 127351180, + "step": 5924, + "time_per_iteration": 2.645853042602539 + }, + { + "auxiliary_loss_clip": 0.01182538, + "auxiliary_loss_mlp": 0.01023953, + "balance_loss_clip": 1.04054964, + "balance_loss_mlp": 1.01609683, + "epoch": 0.7124391270366139, + "flos": 20811670329600.0, + "grad_norm": 2.096155345349674, + "language_loss": 0.77491212, + "learning_rate": 8.063540361001422e-07, + "loss": 0.79697704, + "num_input_tokens_seen": 127369750, + "step": 5925, + "time_per_iteration": 2.6571848392486572 + }, + { + "auxiliary_loss_clip": 0.01188772, + "auxiliary_loss_mlp": 0.01029863, + "balance_loss_clip": 1.04297543, + "balance_loss_mlp": 1.02230477, + "epoch": 0.7125593699272531, + "flos": 17603734584960.0, + "grad_norm": 2.464008523875618, + "language_loss": 0.7984584, + "learning_rate": 8.057291008859069e-07, + "loss": 0.82064474, + "num_input_tokens_seen": 127387910, + "step": 5926, + "time_per_iteration": 2.749760866165161 + }, + { + "auxiliary_loss_clip": 0.01172239, + "auxiliary_loss_mlp": 0.01020985, + "balance_loss_clip": 1.04340887, + "balance_loss_mlp": 1.01384401, + "epoch": 0.7126796128178922, + "flos": 28654057526400.0, + "grad_norm": 2.5044473741352453, + "language_loss": 0.68360949, + "learning_rate": 8.051043468416187e-07, + "loss": 0.70554179, + "num_input_tokens_seen": 127409160, + "step": 5927, + "time_per_iteration": 2.7715225219726562 + }, + { + "auxiliary_loss_clip": 0.01168303, + "auxiliary_loss_mlp": 0.01024159, + "balance_loss_clip": 1.04856515, + "balance_loss_mlp": 1.01723611, + "epoch": 0.7127998557085312, + "flos": 16034438315520.0, + "grad_norm": 2.0093075025832445, + "language_loss": 0.8222568, + "learning_rate": 8.044797740620506e-07, + "loss": 0.84418142, + "num_input_tokens_seen": 127427765, + "step": 5928, + "time_per_iteration": 2.63439679145813 + }, + { + "auxiliary_loss_clip": 0.0120071, + "auxiliary_loss_mlp": 0.01026214, + "balance_loss_clip": 1.04417276, + "balance_loss_mlp": 1.01898074, + "epoch": 0.7129200985991703, + "flos": 23403271582080.0, + "grad_norm": 2.956054021360803, + "language_loss": 0.78540373, + "learning_rate": 8.038553826419494e-07, + "loss": 0.80767304, + "num_input_tokens_seen": 127446475, + "step": 5929, + "time_per_iteration": 2.77992582321167 + }, + { + "auxiliary_loss_clip": 0.01166051, + "auxiliary_loss_mlp": 0.01024452, + "balance_loss_clip": 1.04405391, + "balance_loss_mlp": 1.01716805, + "epoch": 0.7130403414898094, + "flos": 21397445326080.0, + "grad_norm": 2.049225809014158, + "language_loss": 0.81073594, + "learning_rate": 8.032311726760364e-07, + "loss": 0.83264089, + "num_input_tokens_seen": 127467695, + "step": 5930, + "time_per_iteration": 2.636725664138794 + }, + { + "auxiliary_loss_clip": 0.01184297, + "auxiliary_loss_mlp": 0.01024508, + "balance_loss_clip": 1.04232335, + "balance_loss_mlp": 1.01611018, + "epoch": 0.7131605843804485, + "flos": 74739045306240.0, + "grad_norm": 1.8132469563925044, + "language_loss": 0.69133127, + "learning_rate": 8.026071442590022e-07, + "loss": 0.71341932, + "num_input_tokens_seen": 127494590, + "step": 5931, + "time_per_iteration": 3.108529567718506 + }, + { + "auxiliary_loss_clip": 0.01178855, + "auxiliary_loss_mlp": 0.01021878, + "balance_loss_clip": 1.0492115, + "balance_loss_mlp": 1.0149405, + "epoch": 0.7132808272710875, + "flos": 18368739469440.0, + "grad_norm": 2.572028346445952, + "language_loss": 0.80864644, + "learning_rate": 8.019832974855134e-07, + "loss": 0.83065379, + "num_input_tokens_seen": 127512550, + "step": 5932, + "time_per_iteration": 2.6153745651245117 + }, + { + "auxiliary_loss_clip": 0.01193838, + "auxiliary_loss_mlp": 0.01028615, + "balance_loss_clip": 1.04466403, + "balance_loss_mlp": 1.02094984, + "epoch": 0.7134010701617267, + "flos": 23253380127360.0, + "grad_norm": 2.3638513449776855, + "language_loss": 0.82453942, + "learning_rate": 8.013596324502052e-07, + "loss": 0.84676397, + "num_input_tokens_seen": 127531015, + "step": 5933, + "time_per_iteration": 3.6909611225128174 + }, + { + "auxiliary_loss_clip": 0.01172352, + "auxiliary_loss_mlp": 0.01028769, + "balance_loss_clip": 1.04739571, + "balance_loss_mlp": 1.0218097, + "epoch": 0.7135213130523658, + "flos": 23653137565440.0, + "grad_norm": 1.8773648755317331, + "language_loss": 0.78572744, + "learning_rate": 8.007361492476872e-07, + "loss": 0.80773866, + "num_input_tokens_seen": 127550340, + "step": 5934, + "time_per_iteration": 2.5950865745544434 + }, + { + "auxiliary_loss_clip": 0.01202827, + "auxiliary_loss_mlp": 0.01025714, + "balance_loss_clip": 1.04204535, + "balance_loss_mlp": 1.01834941, + "epoch": 0.7136415559430048, + "flos": 24790644443520.0, + "grad_norm": 2.211018291445426, + "language_loss": 0.78966773, + "learning_rate": 8.001128479725426e-07, + "loss": 0.81195307, + "num_input_tokens_seen": 127572245, + "step": 5935, + "time_per_iteration": 3.704190254211426 + }, + { + "auxiliary_loss_clip": 0.01192913, + "auxiliary_loss_mlp": 0.01030906, + "balance_loss_clip": 1.03748941, + "balance_loss_mlp": 1.02347338, + "epoch": 0.713761798833644, + "flos": 18296954138880.0, + "grad_norm": 1.6259491951578886, + "language_loss": 0.80762112, + "learning_rate": 7.994897287193248e-07, + "loss": 0.82985938, + "num_input_tokens_seen": 127591625, + "step": 5936, + "time_per_iteration": 2.7198610305786133 + }, + { + "auxiliary_loss_clip": 0.01183426, + "auxiliary_loss_mlp": 0.01024543, + "balance_loss_clip": 1.04717386, + "balance_loss_mlp": 1.0169729, + "epoch": 0.713882041724283, + "flos": 15558262692480.0, + "grad_norm": 2.9752581638890825, + "language_loss": 0.84011739, + "learning_rate": 7.988667915825605e-07, + "loss": 0.86219704, + "num_input_tokens_seen": 127608690, + "step": 5937, + "time_per_iteration": 2.634960412979126 + }, + { + "auxiliary_loss_clip": 0.01186583, + "auxiliary_loss_mlp": 0.01023723, + "balance_loss_clip": 1.04571927, + "balance_loss_mlp": 1.01618314, + "epoch": 0.7140022846149221, + "flos": 24061011477120.0, + "grad_norm": 2.4001141962418244, + "language_loss": 0.75826508, + "learning_rate": 7.982440366567491e-07, + "loss": 0.78036809, + "num_input_tokens_seen": 127627180, + "step": 5938, + "time_per_iteration": 3.6344637870788574 + }, + { + "auxiliary_loss_clip": 0.01169454, + "auxiliary_loss_mlp": 0.01024982, + "balance_loss_clip": 1.04278314, + "balance_loss_mlp": 1.0181278, + "epoch": 0.7141225275055613, + "flos": 27891710248320.0, + "grad_norm": 1.664141813863324, + "language_loss": 0.752828, + "learning_rate": 7.97621464036361e-07, + "loss": 0.77477241, + "num_input_tokens_seen": 127648940, + "step": 5939, + "time_per_iteration": 2.6147353649139404 + }, + { + "auxiliary_loss_clip": 0.01181103, + "auxiliary_loss_mlp": 0.01021541, + "balance_loss_clip": 1.04629004, + "balance_loss_mlp": 1.0135777, + "epoch": 0.7142427703962003, + "flos": 19682603147520.0, + "grad_norm": 1.62466467468335, + "language_loss": 0.67928094, + "learning_rate": 7.969990738158417e-07, + "loss": 0.70130742, + "num_input_tokens_seen": 127667350, + "step": 5940, + "time_per_iteration": 3.6169745922088623 + }, + { + "auxiliary_loss_clip": 0.01182263, + "auxiliary_loss_mlp": 0.01023448, + "balance_loss_clip": 1.04871154, + "balance_loss_mlp": 1.01522803, + "epoch": 0.7143630132868394, + "flos": 21032377447680.0, + "grad_norm": 2.024404646765013, + "language_loss": 0.84916985, + "learning_rate": 7.963768660896062e-07, + "loss": 0.87122697, + "num_input_tokens_seen": 127685760, + "step": 5941, + "time_per_iteration": 2.5901713371276855 + }, + { + "auxiliary_loss_clip": 0.01181985, + "auxiliary_loss_mlp": 0.01025589, + "balance_loss_clip": 1.04709148, + "balance_loss_mlp": 1.01779866, + "epoch": 0.7144832561774785, + "flos": 24129923719680.0, + "grad_norm": 2.174402215925629, + "language_loss": 0.82377219, + "learning_rate": 7.957548409520432e-07, + "loss": 0.84584796, + "num_input_tokens_seen": 127704985, + "step": 5942, + "time_per_iteration": 2.6196937561035156 + }, + { + "auxiliary_loss_clip": 0.01195197, + "auxiliary_loss_mlp": 0.01028256, + "balance_loss_clip": 1.0419569, + "balance_loss_mlp": 1.02070999, + "epoch": 0.7146034990681176, + "flos": 16325817442560.0, + "grad_norm": 2.145544160975461, + "language_loss": 0.84114015, + "learning_rate": 7.951329984975135e-07, + "loss": 0.86337471, + "num_input_tokens_seen": 127721925, + "step": 5943, + "time_per_iteration": 2.7009644508361816 + }, + { + "auxiliary_loss_clip": 0.0110659, + "auxiliary_loss_mlp": 0.01002979, + "balance_loss_clip": 1.01227665, + "balance_loss_mlp": 1.00177479, + "epoch": 0.7147237419587567, + "flos": 69627164232960.0, + "grad_norm": 0.7208581616827764, + "language_loss": 0.54293358, + "learning_rate": 7.94511338820349e-07, + "loss": 0.56402934, + "num_input_tokens_seen": 127784230, + "step": 5944, + "time_per_iteration": 3.2408323287963867 + }, + { + "auxiliary_loss_clip": 0.01187926, + "auxiliary_loss_mlp": 0.01379225, + "balance_loss_clip": 1.04493403, + "balance_loss_mlp": 1.00014853, + "epoch": 0.7148439848493958, + "flos": 22266806198400.0, + "grad_norm": 2.181208984774594, + "language_loss": 0.78338242, + "learning_rate": 7.938898620148575e-07, + "loss": 0.80905396, + "num_input_tokens_seen": 127801990, + "step": 5945, + "time_per_iteration": 2.658159017562866 + }, + { + "auxiliary_loss_clip": 0.0118689, + "auxiliary_loss_mlp": 0.0102542, + "balance_loss_clip": 1.04594612, + "balance_loss_mlp": 1.01770687, + "epoch": 0.7149642277400349, + "flos": 17931383470080.0, + "grad_norm": 1.9061228211570795, + "language_loss": 0.70815098, + "learning_rate": 7.932685681753135e-07, + "loss": 0.73027402, + "num_input_tokens_seen": 127819270, + "step": 5946, + "time_per_iteration": 2.60591721534729 + }, + { + "auxiliary_loss_clip": 0.01167013, + "auxiliary_loss_mlp": 0.01022144, + "balance_loss_clip": 1.04749739, + "balance_loss_mlp": 1.01519418, + "epoch": 0.7150844706306739, + "flos": 31681937370240.0, + "grad_norm": 2.055086707287311, + "language_loss": 0.62587118, + "learning_rate": 7.92647457395969e-07, + "loss": 0.64776278, + "num_input_tokens_seen": 127841095, + "step": 5947, + "time_per_iteration": 2.6348166465759277 + }, + { + "auxiliary_loss_clip": 0.0121606, + "auxiliary_loss_mlp": 0.0103156, + "balance_loss_clip": 1.03743839, + "balance_loss_mlp": 1.02322173, + "epoch": 0.7152047135213131, + "flos": 10926217451520.0, + "grad_norm": 1.9400838351443177, + "language_loss": 0.7390126, + "learning_rate": 7.920265297710444e-07, + "loss": 0.76148874, + "num_input_tokens_seen": 127858485, + "step": 5948, + "time_per_iteration": 2.75508975982666 + }, + { + "auxiliary_loss_clip": 0.01180851, + "auxiliary_loss_mlp": 0.01025357, + "balance_loss_clip": 1.04735208, + "balance_loss_mlp": 1.01833892, + "epoch": 0.7153249564119522, + "flos": 20995640812800.0, + "grad_norm": 1.8839040054252292, + "language_loss": 0.7325986, + "learning_rate": 7.914057853947363e-07, + "loss": 0.75466073, + "num_input_tokens_seen": 127877665, + "step": 5949, + "time_per_iteration": 2.6317453384399414 + }, + { + "auxiliary_loss_clip": 0.01192068, + "auxiliary_loss_mlp": 0.0102303, + "balance_loss_clip": 1.04280508, + "balance_loss_mlp": 1.01511765, + "epoch": 0.7154451993025912, + "flos": 24243114453120.0, + "grad_norm": 1.8385221323602994, + "language_loss": 0.62601715, + "learning_rate": 7.907852243612089e-07, + "loss": 0.64816809, + "num_input_tokens_seen": 127898070, + "step": 5950, + "time_per_iteration": 2.7212533950805664 + }, + { + "auxiliary_loss_clip": 0.01184078, + "auxiliary_loss_mlp": 0.01027274, + "balance_loss_clip": 1.04357708, + "balance_loss_mlp": 1.0202738, + "epoch": 0.7155654421932304, + "flos": 23330947547520.0, + "grad_norm": 2.182176091444964, + "language_loss": 0.72421587, + "learning_rate": 7.901648467646009e-07, + "loss": 0.74632943, + "num_input_tokens_seen": 127917010, + "step": 5951, + "time_per_iteration": 2.636441946029663 + }, + { + "auxiliary_loss_clip": 0.01172626, + "auxiliary_loss_mlp": 0.01019645, + "balance_loss_clip": 1.04954541, + "balance_loss_mlp": 1.01231933, + "epoch": 0.7156856850838694, + "flos": 22711883621760.0, + "grad_norm": 1.8155841599044822, + "language_loss": 0.72594672, + "learning_rate": 7.895446526990244e-07, + "loss": 0.74786937, + "num_input_tokens_seen": 127937025, + "step": 5952, + "time_per_iteration": 2.5812973976135254 + }, + { + "auxiliary_loss_clip": 0.01214461, + "auxiliary_loss_mlp": 0.01025673, + "balance_loss_clip": 1.0431273, + "balance_loss_mlp": 1.01827002, + "epoch": 0.7158059279745085, + "flos": 19865424395520.0, + "grad_norm": 1.5623336050207208, + "language_loss": 0.75380677, + "learning_rate": 7.889246422585609e-07, + "loss": 0.77620816, + "num_input_tokens_seen": 127956410, + "step": 5953, + "time_per_iteration": 2.662458658218384 + }, + { + "auxiliary_loss_clip": 0.01169015, + "auxiliary_loss_mlp": 0.01027426, + "balance_loss_clip": 1.0476265, + "balance_loss_mlp": 1.02014518, + "epoch": 0.7159261708651476, + "flos": 24134772055680.0, + "grad_norm": 1.63694968014957, + "language_loss": 0.73343372, + "learning_rate": 7.883048155372675e-07, + "loss": 0.75539815, + "num_input_tokens_seen": 127974925, + "step": 5954, + "time_per_iteration": 2.5926291942596436 + }, + { + "auxiliary_loss_clip": 0.0119255, + "auxiliary_loss_mlp": 0.01023572, + "balance_loss_clip": 1.04661179, + "balance_loss_mlp": 1.01597857, + "epoch": 0.7160464137557867, + "flos": 16983198201600.0, + "grad_norm": 2.338800790615999, + "language_loss": 0.7175557, + "learning_rate": 7.876851726291698e-07, + "loss": 0.73971689, + "num_input_tokens_seen": 127993225, + "step": 5955, + "time_per_iteration": 2.5764846801757812 + }, + { + "auxiliary_loss_clip": 0.01196554, + "auxiliary_loss_mlp": 0.01026066, + "balance_loss_clip": 1.04130936, + "balance_loss_mlp": 1.01912236, + "epoch": 0.7161666566464258, + "flos": 25228251838080.0, + "grad_norm": 1.8133585216422399, + "language_loss": 0.78486037, + "learning_rate": 7.870657136282666e-07, + "loss": 0.80708659, + "num_input_tokens_seen": 128012085, + "step": 5956, + "time_per_iteration": 2.691943645477295 + }, + { + "auxiliary_loss_clip": 0.01170275, + "auxiliary_loss_mlp": 0.01024267, + "balance_loss_clip": 1.04384303, + "balance_loss_mlp": 1.01726294, + "epoch": 0.7162868995370649, + "flos": 26468390851200.0, + "grad_norm": 1.5078730936497202, + "language_loss": 0.82058072, + "learning_rate": 7.86446438628531e-07, + "loss": 0.84252608, + "num_input_tokens_seen": 128033155, + "step": 5957, + "time_per_iteration": 2.6241157054901123 + }, + { + "auxiliary_loss_clip": 0.01064869, + "auxiliary_loss_mlp": 0.01000596, + "balance_loss_clip": 1.01175928, + "balance_loss_mlp": 0.99935609, + "epoch": 0.716407142427704, + "flos": 69998912040960.0, + "grad_norm": 0.7678269287062676, + "language_loss": 0.56880343, + "learning_rate": 7.858273477239059e-07, + "loss": 0.58945805, + "num_input_tokens_seen": 128101575, + "step": 5958, + "time_per_iteration": 3.229729652404785 + }, + { + "auxiliary_loss_clip": 0.01187928, + "auxiliary_loss_mlp": 0.0102669, + "balance_loss_clip": 1.0404408, + "balance_loss_mlp": 1.01904845, + "epoch": 0.716527385318343, + "flos": 20740459616640.0, + "grad_norm": 1.65743557672041, + "language_loss": 0.71674287, + "learning_rate": 7.852084410083067e-07, + "loss": 0.73888904, + "num_input_tokens_seen": 128120395, + "step": 5959, + "time_per_iteration": 3.5787391662597656 + }, + { + "auxiliary_loss_clip": 0.01182024, + "auxiliary_loss_mlp": 0.01028742, + "balance_loss_clip": 1.04496503, + "balance_loss_mlp": 1.02152395, + "epoch": 0.7166476282089821, + "flos": 25371966153600.0, + "grad_norm": 1.5353768983992948, + "language_loss": 0.63804448, + "learning_rate": 7.84589718575621e-07, + "loss": 0.66015208, + "num_input_tokens_seen": 128140840, + "step": 5960, + "time_per_iteration": 2.6715800762176514 + }, + { + "auxiliary_loss_clip": 0.01184253, + "auxiliary_loss_mlp": 0.01020143, + "balance_loss_clip": 1.04004049, + "balance_loss_mlp": 1.01243043, + "epoch": 0.7167678710996213, + "flos": 24133730561280.0, + "grad_norm": 2.0902386611187165, + "language_loss": 0.69262159, + "learning_rate": 7.83971180519708e-07, + "loss": 0.71466553, + "num_input_tokens_seen": 128159695, + "step": 5961, + "time_per_iteration": 3.9413232803344727 + }, + { + "auxiliary_loss_clip": 0.01172787, + "auxiliary_loss_mlp": 0.01028646, + "balance_loss_clip": 1.04928541, + "balance_loss_mlp": 1.02081418, + "epoch": 0.7168881139902603, + "flos": 30226586019840.0, + "grad_norm": 2.073073528522134, + "language_loss": 0.75554937, + "learning_rate": 7.833528269344008e-07, + "loss": 0.77756369, + "num_input_tokens_seen": 128179600, + "step": 5962, + "time_per_iteration": 2.727482318878174 + }, + { + "auxiliary_loss_clip": 0.01195041, + "auxiliary_loss_mlp": 0.01027863, + "balance_loss_clip": 1.04548931, + "balance_loss_mlp": 1.01959634, + "epoch": 0.7170083568808994, + "flos": 14606414236800.0, + "grad_norm": 2.2553912177426336, + "language_loss": 0.77480924, + "learning_rate": 7.827346579135023e-07, + "loss": 0.79703832, + "num_input_tokens_seen": 128196940, + "step": 5963, + "time_per_iteration": 2.7096307277679443 + }, + { + "auxiliary_loss_clip": 0.01184032, + "auxiliary_loss_mlp": 0.0102841, + "balance_loss_clip": 1.04465699, + "balance_loss_mlp": 1.02064621, + "epoch": 0.7171285997715385, + "flos": 23331091201920.0, + "grad_norm": 1.9036655772435176, + "language_loss": 0.83200771, + "learning_rate": 7.821166735507885e-07, + "loss": 0.85413212, + "num_input_tokens_seen": 128215970, + "step": 5964, + "time_per_iteration": 3.668306350708008 + }, + { + "auxiliary_loss_clip": 0.01165035, + "auxiliary_loss_mlp": 0.01023154, + "balance_loss_clip": 1.04579926, + "balance_loss_mlp": 1.01601291, + "epoch": 0.7172488426621776, + "flos": 16543543731840.0, + "grad_norm": 1.7794397897608745, + "language_loss": 0.68635404, + "learning_rate": 7.81498873940007e-07, + "loss": 0.70823598, + "num_input_tokens_seen": 128233185, + "step": 5965, + "time_per_iteration": 2.511361837387085 + }, + { + "auxiliary_loss_clip": 0.01182181, + "auxiliary_loss_mlp": 0.01022792, + "balance_loss_clip": 1.04434395, + "balance_loss_mlp": 1.01456702, + "epoch": 0.7173690855528166, + "flos": 26541612725760.0, + "grad_norm": 2.360642034817984, + "language_loss": 0.77167237, + "learning_rate": 7.808812591748768e-07, + "loss": 0.79372215, + "num_input_tokens_seen": 128253565, + "step": 5966, + "time_per_iteration": 3.6157872676849365 + }, + { + "auxiliary_loss_clip": 0.01191806, + "auxiliary_loss_mlp": 0.01024902, + "balance_loss_clip": 1.04347372, + "balance_loss_mlp": 1.01678371, + "epoch": 0.7174893284434558, + "flos": 22784099915520.0, + "grad_norm": 1.9048135815021376, + "language_loss": 0.6455214, + "learning_rate": 7.802638293490915e-07, + "loss": 0.66768849, + "num_input_tokens_seen": 128273210, + "step": 5967, + "time_per_iteration": 2.674233913421631 + }, + { + "auxiliary_loss_clip": 0.01191123, + "auxiliary_loss_mlp": 0.01021492, + "balance_loss_clip": 1.04570484, + "balance_loss_mlp": 1.01446509, + "epoch": 0.7176095713340949, + "flos": 23293564467840.0, + "grad_norm": 1.7042491617976254, + "language_loss": 0.76939994, + "learning_rate": 7.796465845563123e-07, + "loss": 0.79152608, + "num_input_tokens_seen": 128292085, + "step": 5968, + "time_per_iteration": 2.6299211978912354 + }, + { + "auxiliary_loss_clip": 0.01181118, + "auxiliary_loss_mlp": 0.01378655, + "balance_loss_clip": 1.04386473, + "balance_loss_mlp": 1.00017953, + "epoch": 0.7177298142247339, + "flos": 25591631777280.0, + "grad_norm": 1.862837289927579, + "language_loss": 0.79564297, + "learning_rate": 7.790295248901766e-07, + "loss": 0.82124066, + "num_input_tokens_seen": 128313215, + "step": 5969, + "time_per_iteration": 2.7008392810821533 + }, + { + "auxiliary_loss_clip": 0.01177661, + "auxiliary_loss_mlp": 0.0102777, + "balance_loss_clip": 1.04730439, + "balance_loss_mlp": 1.01986647, + "epoch": 0.7178500571153731, + "flos": 31652778504960.0, + "grad_norm": 1.6051479384729481, + "language_loss": 0.61874902, + "learning_rate": 7.784126504442902e-07, + "loss": 0.64080334, + "num_input_tokens_seen": 128336445, + "step": 5970, + "time_per_iteration": 2.7763872146606445 + }, + { + "auxiliary_loss_clip": 0.01187619, + "auxiliary_loss_mlp": 0.0102527, + "balance_loss_clip": 1.04348028, + "balance_loss_mlp": 1.01778388, + "epoch": 0.7179703000060121, + "flos": 19427242383360.0, + "grad_norm": 1.3821510890081778, + "language_loss": 0.67803293, + "learning_rate": 7.777959613122351e-07, + "loss": 0.70016181, + "num_input_tokens_seen": 128356270, + "step": 5971, + "time_per_iteration": 2.6907262802124023 + }, + { + "auxiliary_loss_clip": 0.01181002, + "auxiliary_loss_mlp": 0.01022803, + "balance_loss_clip": 1.04575014, + "balance_loss_mlp": 1.01573384, + "epoch": 0.7180905428966512, + "flos": 28839249072000.0, + "grad_norm": 1.7101385147615382, + "language_loss": 0.77852684, + "learning_rate": 7.771794575875604e-07, + "loss": 0.80056489, + "num_input_tokens_seen": 128378140, + "step": 5972, + "time_per_iteration": 2.639883518218994 + }, + { + "auxiliary_loss_clip": 0.01180122, + "auxiliary_loss_mlp": 0.01028019, + "balance_loss_clip": 1.04856193, + "balance_loss_mlp": 1.02026451, + "epoch": 0.7182107857872904, + "flos": 20047563285120.0, + "grad_norm": 2.3044092667759473, + "language_loss": 0.77783084, + "learning_rate": 7.765631393637888e-07, + "loss": 0.79991221, + "num_input_tokens_seen": 128396335, + "step": 5973, + "time_per_iteration": 2.6538500785827637 + }, + { + "auxiliary_loss_clip": 0.01171125, + "auxiliary_loss_mlp": 0.01027629, + "balance_loss_clip": 1.04247606, + "balance_loss_mlp": 1.01989865, + "epoch": 0.7183310286779294, + "flos": 22747686503040.0, + "grad_norm": 3.5540917441418407, + "language_loss": 0.4865225, + "learning_rate": 7.75947006734417e-07, + "loss": 0.50850999, + "num_input_tokens_seen": 128414115, + "step": 5974, + "time_per_iteration": 2.5833702087402344 + }, + { + "auxiliary_loss_clip": 0.01167619, + "auxiliary_loss_mlp": 0.01024747, + "balance_loss_clip": 1.04630077, + "balance_loss_mlp": 1.01702261, + "epoch": 0.7184512715685685, + "flos": 17158262112000.0, + "grad_norm": 2.0024441991131483, + "language_loss": 0.82554942, + "learning_rate": 7.753310597929101e-07, + "loss": 0.84747303, + "num_input_tokens_seen": 128430755, + "step": 5975, + "time_per_iteration": 2.554486036300659 + }, + { + "auxiliary_loss_clip": 0.01065243, + "auxiliary_loss_mlp": 0.01000556, + "balance_loss_clip": 1.01202345, + "balance_loss_mlp": 0.99932212, + "epoch": 0.7185715144592076, + "flos": 65509611448320.0, + "grad_norm": 0.7534569578009698, + "language_loss": 0.5510639, + "learning_rate": 7.747152986327095e-07, + "loss": 0.57172185, + "num_input_tokens_seen": 128491300, + "step": 5976, + "time_per_iteration": 3.0320661067962646 + }, + { + "auxiliary_loss_clip": 0.01208165, + "auxiliary_loss_mlp": 0.0102766, + "balance_loss_clip": 1.04269385, + "balance_loss_mlp": 1.02077019, + "epoch": 0.7186917573498467, + "flos": 16180522928640.0, + "grad_norm": 1.6794856984632418, + "language_loss": 0.68103504, + "learning_rate": 7.740997233472228e-07, + "loss": 0.70339334, + "num_input_tokens_seen": 128508920, + "step": 5977, + "time_per_iteration": 2.7278330326080322 + }, + { + "auxiliary_loss_clip": 0.01189298, + "auxiliary_loss_mlp": 0.01022724, + "balance_loss_clip": 1.0460434, + "balance_loss_mlp": 1.01614356, + "epoch": 0.7188120002404857, + "flos": 29242274647680.0, + "grad_norm": 2.906840566400786, + "language_loss": 0.7104972, + "learning_rate": 7.734843340298329e-07, + "loss": 0.73261738, + "num_input_tokens_seen": 128528745, + "step": 5978, + "time_per_iteration": 2.653477191925049 + }, + { + "auxiliary_loss_clip": 0.01191167, + "auxiliary_loss_mlp": 0.01025527, + "balance_loss_clip": 1.04316068, + "balance_loss_mlp": 1.01745617, + "epoch": 0.7189322431311249, + "flos": 33401161008000.0, + "grad_norm": 1.845412954346436, + "language_loss": 0.7522614, + "learning_rate": 7.72869130773895e-07, + "loss": 0.77442837, + "num_input_tokens_seen": 128549345, + "step": 5979, + "time_per_iteration": 2.713088035583496 + }, + { + "auxiliary_loss_clip": 0.01078999, + "auxiliary_loss_mlp": 0.01000997, + "balance_loss_clip": 1.01203942, + "balance_loss_mlp": 0.99978155, + "epoch": 0.719052486021764, + "flos": 61351263792000.0, + "grad_norm": 0.8237684505242788, + "language_loss": 0.59380579, + "learning_rate": 7.722541136727343e-07, + "loss": 0.61460572, + "num_input_tokens_seen": 128605360, + "step": 5980, + "time_per_iteration": 3.070472240447998 + }, + { + "auxiliary_loss_clip": 0.01178645, + "auxiliary_loss_mlp": 0.01023127, + "balance_loss_clip": 1.0466342, + "balance_loss_mlp": 1.01558065, + "epoch": 0.719172728912403, + "flos": 15596795007360.0, + "grad_norm": 2.1699040038361397, + "language_loss": 0.80661464, + "learning_rate": 7.716392828196483e-07, + "loss": 0.82863235, + "num_input_tokens_seen": 128623160, + "step": 5981, + "time_per_iteration": 2.553788661956787 + }, + { + "auxiliary_loss_clip": 0.01179631, + "auxiliary_loss_mlp": 0.01026375, + "balance_loss_clip": 1.04802227, + "balance_loss_mlp": 1.01879346, + "epoch": 0.7192929718030422, + "flos": 15553162961280.0, + "grad_norm": 2.448929016036425, + "language_loss": 0.77101266, + "learning_rate": 7.710246383079064e-07, + "loss": 0.79307276, + "num_input_tokens_seen": 128638545, + "step": 5982, + "time_per_iteration": 2.6809515953063965 + }, + { + "auxiliary_loss_clip": 0.01184345, + "auxiliary_loss_mlp": 0.0102702, + "balance_loss_clip": 1.04102397, + "balance_loss_mlp": 1.01958084, + "epoch": 0.7194132146936812, + "flos": 21862487733120.0, + "grad_norm": 2.6219018043175546, + "language_loss": 0.91989893, + "learning_rate": 7.704101802307492e-07, + "loss": 0.94201255, + "num_input_tokens_seen": 128650845, + "step": 5983, + "time_per_iteration": 2.6300690174102783 + }, + { + "auxiliary_loss_clip": 0.01185703, + "auxiliary_loss_mlp": 0.0102533, + "balance_loss_clip": 1.04285097, + "balance_loss_mlp": 1.01773, + "epoch": 0.7195334575843203, + "flos": 27338900958720.0, + "grad_norm": 2.1180011158865883, + "language_loss": 0.87343693, + "learning_rate": 7.697959086813912e-07, + "loss": 0.89554727, + "num_input_tokens_seen": 128667010, + "step": 5984, + "time_per_iteration": 2.6793007850646973 + }, + { + "auxiliary_loss_clip": 0.01185882, + "auxiliary_loss_mlp": 0.01027654, + "balance_loss_clip": 1.04148531, + "balance_loss_mlp": 1.01942873, + "epoch": 0.7196537004749595, + "flos": 18770615809920.0, + "grad_norm": 2.290651181994944, + "language_loss": 0.79997069, + "learning_rate": 7.691818237530145e-07, + "loss": 0.82210606, + "num_input_tokens_seen": 128685870, + "step": 5985, + "time_per_iteration": 3.5730233192443848 + }, + { + "auxiliary_loss_clip": 0.01216786, + "auxiliary_loss_mlp": 0.01026185, + "balance_loss_clip": 1.04376817, + "balance_loss_mlp": 1.01875877, + "epoch": 0.7197739433655985, + "flos": 24531009960960.0, + "grad_norm": 1.7378646426094817, + "language_loss": 0.77334183, + "learning_rate": 7.685679255387774e-07, + "loss": 0.79577154, + "num_input_tokens_seen": 128704185, + "step": 5986, + "time_per_iteration": 3.691403388977051 + }, + { + "auxiliary_loss_clip": 0.01183842, + "auxiliary_loss_mlp": 0.01027031, + "balance_loss_clip": 1.04549634, + "balance_loss_mlp": 1.01991451, + "epoch": 0.7198941862562376, + "flos": 18040587793920.0, + "grad_norm": 1.9364548157104373, + "language_loss": 0.7747345, + "learning_rate": 7.679542141318065e-07, + "loss": 0.79684329, + "num_input_tokens_seen": 128721290, + "step": 5987, + "time_per_iteration": 2.666201591491699 + }, + { + "auxiliary_loss_clip": 0.01168893, + "auxiliary_loss_mlp": 0.01027393, + "balance_loss_clip": 1.03977168, + "balance_loss_mlp": 1.01995409, + "epoch": 0.7200144291468767, + "flos": 29022393542400.0, + "grad_norm": 1.8249120719165233, + "language_loss": 0.75514412, + "learning_rate": 7.673406896252013e-07, + "loss": 0.777107, + "num_input_tokens_seen": 128742665, + "step": 5988, + "time_per_iteration": 2.6683900356292725 + }, + { + "auxiliary_loss_clip": 0.01193239, + "auxiliary_loss_mlp": 0.01024747, + "balance_loss_clip": 1.04183757, + "balance_loss_mlp": 1.01673603, + "epoch": 0.7201346720375158, + "flos": 25374264624000.0, + "grad_norm": 1.5511455365234141, + "language_loss": 0.7850647, + "learning_rate": 7.667273521120347e-07, + "loss": 0.80724454, + "num_input_tokens_seen": 128762225, + "step": 5989, + "time_per_iteration": 2.715040683746338 + }, + { + "auxiliary_loss_clip": 0.01196093, + "auxiliary_loss_mlp": 0.01022495, + "balance_loss_clip": 1.04319406, + "balance_loss_mlp": 1.01531219, + "epoch": 0.7202549149281549, + "flos": 14355614499840.0, + "grad_norm": 1.9350409759328648, + "language_loss": 0.79603243, + "learning_rate": 7.661142016853468e-07, + "loss": 0.81821835, + "num_input_tokens_seen": 128779585, + "step": 5990, + "time_per_iteration": 3.479811906814575 + }, + { + "auxiliary_loss_clip": 0.01201572, + "auxiliary_loss_mlp": 0.01020217, + "balance_loss_clip": 1.04218912, + "balance_loss_mlp": 1.01288605, + "epoch": 0.7203751578187939, + "flos": 23001682550400.0, + "grad_norm": 1.7896367220709282, + "language_loss": 0.74642032, + "learning_rate": 7.655012384381543e-07, + "loss": 0.76863819, + "num_input_tokens_seen": 128799070, + "step": 5991, + "time_per_iteration": 2.7249796390533447 + }, + { + "auxiliary_loss_clip": 0.01185238, + "auxiliary_loss_mlp": 0.01024924, + "balance_loss_clip": 1.0475595, + "balance_loss_mlp": 1.01764977, + "epoch": 0.7204954007094331, + "flos": 23692424065920.0, + "grad_norm": 2.1753143254824674, + "language_loss": 0.81614137, + "learning_rate": 7.648884624634415e-07, + "loss": 0.83824295, + "num_input_tokens_seen": 128817620, + "step": 5992, + "time_per_iteration": 3.7220165729522705 + }, + { + "auxiliary_loss_clip": 0.0117766, + "auxiliary_loss_mlp": 0.01021712, + "balance_loss_clip": 1.0469147, + "balance_loss_mlp": 1.01446152, + "epoch": 0.7206156436000721, + "flos": 16253026531200.0, + "grad_norm": 1.9844324452168751, + "language_loss": 0.89097059, + "learning_rate": 7.642758738541683e-07, + "loss": 0.91296434, + "num_input_tokens_seen": 128834200, + "step": 5993, + "time_per_iteration": 2.5995545387268066 + }, + { + "auxiliary_loss_clip": 0.0107823, + "auxiliary_loss_mlp": 0.0100513, + "balance_loss_clip": 1.01192808, + "balance_loss_mlp": 1.00393188, + "epoch": 0.7207358864907112, + "flos": 54377806504320.0, + "grad_norm": 0.7571335415056228, + "language_loss": 0.60742712, + "learning_rate": 7.636634727032621e-07, + "loss": 0.62826073, + "num_input_tokens_seen": 128891305, + "step": 5994, + "time_per_iteration": 3.0612668991088867 + }, + { + "auxiliary_loss_clip": 0.01192622, + "auxiliary_loss_mlp": 0.01028487, + "balance_loss_clip": 1.03902984, + "balance_loss_mlp": 1.02039313, + "epoch": 0.7208561293813504, + "flos": 19135540033920.0, + "grad_norm": 1.9370505522312957, + "language_loss": 0.78539383, + "learning_rate": 7.630512591036231e-07, + "loss": 0.80760491, + "num_input_tokens_seen": 128910615, + "step": 5995, + "time_per_iteration": 2.6811606884002686 + }, + { + "auxiliary_loss_clip": 0.01179037, + "auxiliary_loss_mlp": 0.01023932, + "balance_loss_clip": 1.04519737, + "balance_loss_mlp": 1.01639771, + "epoch": 0.7209763722719894, + "flos": 17748526308480.0, + "grad_norm": 2.3219859808219487, + "language_loss": 0.64371854, + "learning_rate": 7.624392331481255e-07, + "loss": 0.66574824, + "num_input_tokens_seen": 128928270, + "step": 5996, + "time_per_iteration": 2.574995279312134 + }, + { + "auxiliary_loss_clip": 0.01076144, + "auxiliary_loss_mlp": 0.01001917, + "balance_loss_clip": 1.01092362, + "balance_loss_mlp": 1.00068879, + "epoch": 0.7210966151626285, + "flos": 66819488716800.0, + "grad_norm": 0.7449334714665675, + "language_loss": 0.51825595, + "learning_rate": 7.618273949296115e-07, + "loss": 0.53903657, + "num_input_tokens_seen": 128987780, + "step": 5997, + "time_per_iteration": 3.109063148498535 + }, + { + "auxiliary_loss_clip": 0.01180703, + "auxiliary_loss_mlp": 0.01023849, + "balance_loss_clip": 1.04338264, + "balance_loss_mlp": 1.01638651, + "epoch": 0.7212168580532676, + "flos": 21141869080320.0, + "grad_norm": 1.9911717647307245, + "language_loss": 0.68245006, + "learning_rate": 7.612157445408987e-07, + "loss": 0.70449555, + "num_input_tokens_seen": 129005590, + "step": 5998, + "time_per_iteration": 2.650228261947632 + }, + { + "auxiliary_loss_clip": 0.01190301, + "auxiliary_loss_mlp": 0.0102814, + "balance_loss_clip": 1.04696369, + "balance_loss_mlp": 1.01974726, + "epoch": 0.7213371009439067, + "flos": 22345738335360.0, + "grad_norm": 2.29822118044211, + "language_loss": 0.74328375, + "learning_rate": 7.606042820747716e-07, + "loss": 0.76546818, + "num_input_tokens_seen": 129021995, + "step": 5999, + "time_per_iteration": 2.6902103424072266 + }, + { + "auxiliary_loss_clip": 0.01195359, + "auxiliary_loss_mlp": 0.01026713, + "balance_loss_clip": 1.04997468, + "balance_loss_mlp": 1.01907218, + "epoch": 0.7214573438345457, + "flos": 18515901490560.0, + "grad_norm": 1.7075099925466977, + "language_loss": 0.85538781, + "learning_rate": 7.599930076239889e-07, + "loss": 0.87760854, + "num_input_tokens_seen": 129039280, + "step": 6000, + "time_per_iteration": 2.6316208839416504 + }, + { + "auxiliary_loss_clip": 0.01209579, + "auxiliary_loss_mlp": 0.01378883, + "balance_loss_clip": 1.04457402, + "balance_loss_mlp": 1.00015986, + "epoch": 0.7215775867251849, + "flos": 35736108606720.0, + "grad_norm": 1.8211477793132727, + "language_loss": 0.70530653, + "learning_rate": 7.593819212812818e-07, + "loss": 0.73119116, + "num_input_tokens_seen": 129060860, + "step": 6001, + "time_per_iteration": 2.844871997833252 + }, + { + "auxiliary_loss_clip": 0.01177701, + "auxiliary_loss_mlp": 0.01023453, + "balance_loss_clip": 1.04716623, + "balance_loss_mlp": 1.01667321, + "epoch": 0.721697829615824, + "flos": 20372410909440.0, + "grad_norm": 2.2713104642328883, + "language_loss": 0.71870768, + "learning_rate": 7.587710231393508e-07, + "loss": 0.74071926, + "num_input_tokens_seen": 129079215, + "step": 6002, + "time_per_iteration": 2.5704519748687744 + }, + { + "auxiliary_loss_clip": 0.01207362, + "auxiliary_loss_mlp": 0.01021052, + "balance_loss_clip": 1.03677273, + "balance_loss_mlp": 1.01358366, + "epoch": 0.721818072506463, + "flos": 20229809915520.0, + "grad_norm": 2.0084897389548817, + "language_loss": 0.83455431, + "learning_rate": 7.581603132908685e-07, + "loss": 0.85683846, + "num_input_tokens_seen": 129097185, + "step": 6003, + "time_per_iteration": 2.944232702255249 + }, + { + "auxiliary_loss_clip": 0.01185745, + "auxiliary_loss_mlp": 0.01024457, + "balance_loss_clip": 1.04162109, + "balance_loss_mlp": 1.01675642, + "epoch": 0.7219383153971022, + "flos": 18186887888640.0, + "grad_norm": 2.624774577998776, + "language_loss": 0.78319782, + "learning_rate": 7.575497918284795e-07, + "loss": 0.80529982, + "num_input_tokens_seen": 129114730, + "step": 6004, + "time_per_iteration": 2.774718761444092 + }, + { + "auxiliary_loss_clip": 0.0117093, + "auxiliary_loss_mlp": 0.01024922, + "balance_loss_clip": 1.04567814, + "balance_loss_mlp": 1.01692319, + "epoch": 0.7220585582877412, + "flos": 17342124854400.0, + "grad_norm": 2.0136664731537497, + "language_loss": 0.74508977, + "learning_rate": 7.569394588447984e-07, + "loss": 0.76704824, + "num_input_tokens_seen": 129131745, + "step": 6005, + "time_per_iteration": 2.6594762802124023 + }, + { + "auxiliary_loss_clip": 0.0116608, + "auxiliary_loss_mlp": 0.01027204, + "balance_loss_clip": 1.04334831, + "balance_loss_mlp": 1.01984286, + "epoch": 0.7221788011783803, + "flos": 16976338704000.0, + "grad_norm": 2.6722204310779296, + "language_loss": 0.77954268, + "learning_rate": 7.563293144324146e-07, + "loss": 0.80147558, + "num_input_tokens_seen": 129147295, + "step": 6006, + "time_per_iteration": 2.567603588104248 + }, + { + "auxiliary_loss_clip": 0.0116756, + "auxiliary_loss_mlp": 0.01025409, + "balance_loss_clip": 1.04765916, + "balance_loss_mlp": 1.01856375, + "epoch": 0.7222990440690195, + "flos": 26286359702400.0, + "grad_norm": 2.4738335499392266, + "language_loss": 0.79823411, + "learning_rate": 7.557193586838834e-07, + "loss": 0.82016379, + "num_input_tokens_seen": 129162660, + "step": 6007, + "time_per_iteration": 2.6092426776885986 + }, + { + "auxiliary_loss_clip": 0.01189359, + "auxiliary_loss_mlp": 0.01023461, + "balance_loss_clip": 1.04411459, + "balance_loss_mlp": 1.01598084, + "epoch": 0.7224192869596585, + "flos": 17601687509760.0, + "grad_norm": 2.224891229527945, + "language_loss": 0.71367788, + "learning_rate": 7.551095916917371e-07, + "loss": 0.73580611, + "num_input_tokens_seen": 129179990, + "step": 6008, + "time_per_iteration": 2.574601173400879 + }, + { + "auxiliary_loss_clip": 0.01208907, + "auxiliary_loss_mlp": 0.01026377, + "balance_loss_clip": 1.04540217, + "balance_loss_mlp": 1.01828265, + "epoch": 0.7225395298502976, + "flos": 12932331016320.0, + "grad_norm": 2.8902016280263867, + "language_loss": 0.66778785, + "learning_rate": 7.545000135484758e-07, + "loss": 0.69014072, + "num_input_tokens_seen": 129197425, + "step": 6009, + "time_per_iteration": 2.6478865146636963 + }, + { + "auxiliary_loss_clip": 0.01170282, + "auxiliary_loss_mlp": 0.01379197, + "balance_loss_clip": 1.04783869, + "balance_loss_mlp": 1.00012803, + "epoch": 0.7226597727409367, + "flos": 29643899592960.0, + "grad_norm": 1.8070855788394125, + "language_loss": 0.62707603, + "learning_rate": 7.538906243465714e-07, + "loss": 0.65257078, + "num_input_tokens_seen": 129217560, + "step": 6010, + "time_per_iteration": 2.617760181427002 + }, + { + "auxiliary_loss_clip": 0.01169708, + "auxiliary_loss_mlp": 0.01026735, + "balance_loss_clip": 1.04739153, + "balance_loss_mlp": 1.01875997, + "epoch": 0.7227800156315758, + "flos": 13771635183360.0, + "grad_norm": 2.2670317466979313, + "language_loss": 0.78432995, + "learning_rate": 7.5328142417847e-07, + "loss": 0.80629432, + "num_input_tokens_seen": 129234325, + "step": 6011, + "time_per_iteration": 3.603684902191162 + }, + { + "auxiliary_loss_clip": 0.01174887, + "auxiliary_loss_mlp": 0.01024502, + "balance_loss_clip": 1.0448308, + "balance_loss_mlp": 1.01739144, + "epoch": 0.7229002585222148, + "flos": 20301882554880.0, + "grad_norm": 1.603148631712873, + "language_loss": 0.68896616, + "learning_rate": 7.526724131365838e-07, + "loss": 0.71096003, + "num_input_tokens_seen": 129255280, + "step": 6012, + "time_per_iteration": 3.595576286315918 + }, + { + "auxiliary_loss_clip": 0.01184886, + "auxiliary_loss_mlp": 0.0102886, + "balance_loss_clip": 1.0464673, + "balance_loss_mlp": 1.02066398, + "epoch": 0.723020501412854, + "flos": 16581250033920.0, + "grad_norm": 2.1200883352653697, + "language_loss": 0.70501596, + "learning_rate": 7.520635913133017e-07, + "loss": 0.72715336, + "num_input_tokens_seen": 129273910, + "step": 6013, + "time_per_iteration": 2.61548113822937 + }, + { + "auxiliary_loss_clip": 0.01182826, + "auxiliary_loss_mlp": 0.01029441, + "balance_loss_clip": 1.04689455, + "balance_loss_mlp": 1.02188885, + "epoch": 0.7231407443034931, + "flos": 28548300908160.0, + "grad_norm": 1.8398906446902175, + "language_loss": 0.82310832, + "learning_rate": 7.514549588009798e-07, + "loss": 0.84523106, + "num_input_tokens_seen": 129294785, + "step": 6014, + "time_per_iteration": 2.7225284576416016 + }, + { + "auxiliary_loss_clip": 0.01191718, + "auxiliary_loss_mlp": 0.01027903, + "balance_loss_clip": 1.04768598, + "balance_loss_mlp": 1.02025557, + "epoch": 0.7232609871941321, + "flos": 30008536508160.0, + "grad_norm": 2.4022971970791307, + "language_loss": 0.70291483, + "learning_rate": 7.508465156919492e-07, + "loss": 0.72511101, + "num_input_tokens_seen": 129318295, + "step": 6015, + "time_per_iteration": 2.777644634246826 + }, + { + "auxiliary_loss_clip": 0.01185528, + "auxiliary_loss_mlp": 0.0103209, + "balance_loss_clip": 1.04364109, + "balance_loss_mlp": 1.02467513, + "epoch": 0.7233812300847713, + "flos": 16654005031680.0, + "grad_norm": 2.7721699043563692, + "language_loss": 0.61440778, + "learning_rate": 7.502382620785083e-07, + "loss": 0.63658392, + "num_input_tokens_seen": 129334845, + "step": 6016, + "time_per_iteration": 3.618447780609131 + }, + { + "auxiliary_loss_clip": 0.01095785, + "auxiliary_loss_mlp": 0.01001076, + "balance_loss_clip": 1.01335895, + "balance_loss_mlp": 0.99979478, + "epoch": 0.7235014729754103, + "flos": 67258784050560.0, + "grad_norm": 0.8066029067940362, + "language_loss": 0.62521511, + "learning_rate": 7.496301980529289e-07, + "loss": 0.64618373, + "num_input_tokens_seen": 129398055, + "step": 6017, + "time_per_iteration": 3.348708391189575 + }, + { + "auxiliary_loss_clip": 0.01170236, + "auxiliary_loss_mlp": 0.01030712, + "balance_loss_clip": 1.04786777, + "balance_loss_mlp": 1.02311265, + "epoch": 0.7236217158660494, + "flos": 26943237671040.0, + "grad_norm": 1.963624452534612, + "language_loss": 0.74287575, + "learning_rate": 7.490223237074547e-07, + "loss": 0.76488525, + "num_input_tokens_seen": 129417765, + "step": 6018, + "time_per_iteration": 3.5759942531585693 + }, + { + "auxiliary_loss_clip": 0.01193714, + "auxiliary_loss_mlp": 0.01021003, + "balance_loss_clip": 1.04289401, + "balance_loss_mlp": 1.01337409, + "epoch": 0.7237419587566886, + "flos": 29423372042880.0, + "grad_norm": 2.1705347311002767, + "language_loss": 0.6609652, + "learning_rate": 7.484146391342989e-07, + "loss": 0.68311238, + "num_input_tokens_seen": 129437560, + "step": 6019, + "time_per_iteration": 2.7033705711364746 + }, + { + "auxiliary_loss_clip": 0.01176073, + "auxiliary_loss_mlp": 0.01026054, + "balance_loss_clip": 1.04140782, + "balance_loss_mlp": 1.01859713, + "epoch": 0.7238622016473276, + "flos": 17821496787840.0, + "grad_norm": 3.6077757911775996, + "language_loss": 0.57047784, + "learning_rate": 7.478071444256484e-07, + "loss": 0.59249914, + "num_input_tokens_seen": 129455320, + "step": 6020, + "time_per_iteration": 2.719559907913208 + }, + { + "auxiliary_loss_clip": 0.01204786, + "auxiliary_loss_mlp": 0.01025888, + "balance_loss_clip": 1.04317617, + "balance_loss_mlp": 1.01852417, + "epoch": 0.7239824445379667, + "flos": 25739117020800.0, + "grad_norm": 1.8039420227900653, + "language_loss": 0.79336917, + "learning_rate": 7.471998396736579e-07, + "loss": 0.81567591, + "num_input_tokens_seen": 129475700, + "step": 6021, + "time_per_iteration": 2.8339710235595703 + }, + { + "auxiliary_loss_clip": 0.01197419, + "auxiliary_loss_mlp": 0.01024217, + "balance_loss_clip": 1.04496419, + "balance_loss_mlp": 1.01665878, + "epoch": 0.7241026874286057, + "flos": 23148916398720.0, + "grad_norm": 1.9002293390759666, + "language_loss": 0.75849676, + "learning_rate": 7.465927249704549e-07, + "loss": 0.78071314, + "num_input_tokens_seen": 129493585, + "step": 6022, + "time_per_iteration": 2.6962532997131348 + }, + { + "auxiliary_loss_clip": 0.01177832, + "auxiliary_loss_mlp": 0.01023884, + "balance_loss_clip": 1.04520428, + "balance_loss_mlp": 1.0162015, + "epoch": 0.7242229303192449, + "flos": 20266905686400.0, + "grad_norm": 1.79083991731087, + "language_loss": 0.77627343, + "learning_rate": 7.459858004081398e-07, + "loss": 0.79829061, + "num_input_tokens_seen": 129511555, + "step": 6023, + "time_per_iteration": 2.6521506309509277 + }, + { + "auxiliary_loss_clip": 0.01097474, + "auxiliary_loss_mlp": 0.01000279, + "balance_loss_clip": 1.01311278, + "balance_loss_mlp": 0.99906296, + "epoch": 0.724343173209884, + "flos": 62311659684480.0, + "grad_norm": 0.6530533750018467, + "language_loss": 0.57962096, + "learning_rate": 7.453790660787815e-07, + "loss": 0.60059851, + "num_input_tokens_seen": 129579650, + "step": 6024, + "time_per_iteration": 3.3745663166046143 + }, + { + "auxiliary_loss_clip": 0.01192024, + "auxiliary_loss_mlp": 0.01027952, + "balance_loss_clip": 1.04679704, + "balance_loss_mlp": 1.01961958, + "epoch": 0.724463416100523, + "flos": 35006403813120.0, + "grad_norm": 3.461443583066096, + "language_loss": 0.63539785, + "learning_rate": 7.447725220744214e-07, + "loss": 0.65759754, + "num_input_tokens_seen": 129601895, + "step": 6025, + "time_per_iteration": 2.796407461166382 + }, + { + "auxiliary_loss_clip": 0.01173936, + "auxiliary_loss_mlp": 0.01027443, + "balance_loss_clip": 1.04895711, + "balance_loss_mlp": 1.01929176, + "epoch": 0.7245836589911622, + "flos": 21871968923520.0, + "grad_norm": 1.9386419512287385, + "language_loss": 0.76861525, + "learning_rate": 7.441661684870717e-07, + "loss": 0.79062903, + "num_input_tokens_seen": 129622150, + "step": 6026, + "time_per_iteration": 2.669790267944336 + }, + { + "auxiliary_loss_clip": 0.01171234, + "auxiliary_loss_mlp": 0.01023211, + "balance_loss_clip": 1.04820991, + "balance_loss_mlp": 1.01586771, + "epoch": 0.7247039018818012, + "flos": 23006494972800.0, + "grad_norm": 1.7878018690383917, + "language_loss": 0.81513953, + "learning_rate": 7.435600054087152e-07, + "loss": 0.83708405, + "num_input_tokens_seen": 129644315, + "step": 6027, + "time_per_iteration": 2.6881041526794434 + }, + { + "auxiliary_loss_clip": 0.01170192, + "auxiliary_loss_mlp": 0.01026229, + "balance_loss_clip": 1.04826593, + "balance_loss_mlp": 1.01837301, + "epoch": 0.7248241447724403, + "flos": 31722588587520.0, + "grad_norm": 2.0146159864825424, + "language_loss": 0.74045348, + "learning_rate": 7.42954032931308e-07, + "loss": 0.76241773, + "num_input_tokens_seen": 129665355, + "step": 6028, + "time_per_iteration": 2.5907537937164307 + }, + { + "auxiliary_loss_clip": 0.01189819, + "auxiliary_loss_mlp": 0.01023785, + "balance_loss_clip": 1.0460341, + "balance_loss_mlp": 1.01606631, + "epoch": 0.7249443876630794, + "flos": 34896984007680.0, + "grad_norm": 3.022961142230916, + "language_loss": 0.74774027, + "learning_rate": 7.423482511467733e-07, + "loss": 0.76987636, + "num_input_tokens_seen": 129686125, + "step": 6029, + "time_per_iteration": 2.82848858833313 + }, + { + "auxiliary_loss_clip": 0.0120095, + "auxiliary_loss_mlp": 0.01027506, + "balance_loss_clip": 1.04040003, + "balance_loss_mlp": 1.02031422, + "epoch": 0.7250646305537185, + "flos": 26359294268160.0, + "grad_norm": 2.5335144349212984, + "language_loss": 0.65042645, + "learning_rate": 7.417426601470099e-07, + "loss": 0.67271101, + "num_input_tokens_seen": 129706485, + "step": 6030, + "time_per_iteration": 2.7200732231140137 + }, + { + "auxiliary_loss_clip": 0.01180391, + "auxiliary_loss_mlp": 0.01026336, + "balance_loss_clip": 1.046965, + "balance_loss_mlp": 1.01854575, + "epoch": 0.7251848734443576, + "flos": 30081614728320.0, + "grad_norm": 2.131304735456805, + "language_loss": 0.78294492, + "learning_rate": 7.411372600238841e-07, + "loss": 0.80501223, + "num_input_tokens_seen": 129727100, + "step": 6031, + "time_per_iteration": 2.6721954345703125 + }, + { + "auxiliary_loss_clip": 0.01170063, + "auxiliary_loss_mlp": 0.01025182, + "balance_loss_clip": 1.04767704, + "balance_loss_mlp": 1.01789236, + "epoch": 0.7253051163349967, + "flos": 17785262943360.0, + "grad_norm": 2.2833054838430034, + "language_loss": 0.73802894, + "learning_rate": 7.405320508692346e-07, + "loss": 0.75998139, + "num_input_tokens_seen": 129745840, + "step": 6032, + "time_per_iteration": 2.509486675262451 + }, + { + "auxiliary_loss_clip": 0.01166257, + "auxiliary_loss_mlp": 0.01030029, + "balance_loss_clip": 1.04761338, + "balance_loss_mlp": 1.02308488, + "epoch": 0.7254253592256358, + "flos": 12641346938880.0, + "grad_norm": 2.0503261496637943, + "language_loss": 0.75589406, + "learning_rate": 7.399270327748727e-07, + "loss": 0.77785695, + "num_input_tokens_seen": 129763500, + "step": 6033, + "time_per_iteration": 2.5706191062927246 + }, + { + "auxiliary_loss_clip": 0.01196785, + "auxiliary_loss_mlp": 0.01378058, + "balance_loss_clip": 1.04294944, + "balance_loss_mlp": 1.00017595, + "epoch": 0.7255456021162748, + "flos": 27199208966400.0, + "grad_norm": 1.7785387731475246, + "language_loss": 0.74190509, + "learning_rate": 7.39322205832577e-07, + "loss": 0.76765358, + "num_input_tokens_seen": 129784390, + "step": 6034, + "time_per_iteration": 2.7318832874298096 + }, + { + "auxiliary_loss_clip": 0.01182964, + "auxiliary_loss_mlp": 0.01028921, + "balance_loss_clip": 1.04572868, + "balance_loss_mlp": 1.02107739, + "epoch": 0.725665845006914, + "flos": 21288205088640.0, + "grad_norm": 2.2075859393479167, + "language_loss": 0.80842203, + "learning_rate": 7.387175701341009e-07, + "loss": 0.8305409, + "num_input_tokens_seen": 129803060, + "step": 6035, + "time_per_iteration": 2.6563949584960938 + }, + { + "auxiliary_loss_clip": 0.01178203, + "auxiliary_loss_mlp": 0.01027596, + "balance_loss_clip": 1.04455328, + "balance_loss_mlp": 1.01988888, + "epoch": 0.7257860878975531, + "flos": 16033684129920.0, + "grad_norm": 2.528688211039615, + "language_loss": 0.71745813, + "learning_rate": 7.381131257711659e-07, + "loss": 0.73951608, + "num_input_tokens_seen": 129820165, + "step": 6036, + "time_per_iteration": 2.5442371368408203 + }, + { + "auxiliary_loss_clip": 0.01184162, + "auxiliary_loss_mlp": 0.01027096, + "balance_loss_clip": 1.04807639, + "balance_loss_mlp": 1.02017641, + "epoch": 0.7259063307881921, + "flos": 12129943052160.0, + "grad_norm": 1.7989256441552914, + "language_loss": 0.83298898, + "learning_rate": 7.375088728354677e-07, + "loss": 0.85510153, + "num_input_tokens_seen": 129835195, + "step": 6037, + "time_per_iteration": 3.4761762619018555 + }, + { + "auxiliary_loss_clip": 0.01199158, + "auxiliary_loss_mlp": 0.01027371, + "balance_loss_clip": 1.04356432, + "balance_loss_mlp": 1.02075803, + "epoch": 0.7260265736788313, + "flos": 30443845432320.0, + "grad_norm": 3.6902561395019764, + "language_loss": 0.6747067, + "learning_rate": 7.369048114186691e-07, + "loss": 0.69697195, + "num_input_tokens_seen": 129856240, + "step": 6038, + "time_per_iteration": 2.7585601806640625 + }, + { + "auxiliary_loss_clip": 0.01201084, + "auxiliary_loss_mlp": 0.01378551, + "balance_loss_clip": 1.04303741, + "balance_loss_mlp": 1.00020504, + "epoch": 0.7261468165694703, + "flos": 21142264129920.0, + "grad_norm": 1.836540556609348, + "language_loss": 0.83121419, + "learning_rate": 7.363009416124055e-07, + "loss": 0.85701054, + "num_input_tokens_seen": 129875565, + "step": 6039, + "time_per_iteration": 3.573007583618164 + }, + { + "auxiliary_loss_clip": 0.01201025, + "auxiliary_loss_mlp": 0.0102385, + "balance_loss_clip": 1.04633784, + "balance_loss_mlp": 1.01633382, + "epoch": 0.7262670594601094, + "flos": 22306308180480.0, + "grad_norm": 6.205297741608696, + "language_loss": 0.6283465, + "learning_rate": 7.356972635082852e-07, + "loss": 0.65059525, + "num_input_tokens_seen": 129894420, + "step": 6040, + "time_per_iteration": 2.7008700370788574 + }, + { + "auxiliary_loss_clip": 0.01206501, + "auxiliary_loss_mlp": 0.01026371, + "balance_loss_clip": 1.04768264, + "balance_loss_mlp": 1.01803815, + "epoch": 0.7263873023507486, + "flos": 25335049950720.0, + "grad_norm": 1.9406897655742137, + "language_loss": 0.75475132, + "learning_rate": 7.35093777197884e-07, + "loss": 0.77708, + "num_input_tokens_seen": 129914490, + "step": 6041, + "time_per_iteration": 2.8030083179473877 + }, + { + "auxiliary_loss_clip": 0.0118593, + "auxiliary_loss_mlp": 0.01024339, + "balance_loss_clip": 1.04612207, + "balance_loss_mlp": 1.01716256, + "epoch": 0.7265075452413876, + "flos": 23878621192320.0, + "grad_norm": 2.915397215192861, + "language_loss": 0.85814601, + "learning_rate": 7.344904827727525e-07, + "loss": 0.88024867, + "num_input_tokens_seen": 129931670, + "step": 6042, + "time_per_iteration": 3.5559122562408447 + }, + { + "auxiliary_loss_clip": 0.01195239, + "auxiliary_loss_mlp": 0.0102614, + "balance_loss_clip": 1.0417192, + "balance_loss_mlp": 1.01892447, + "epoch": 0.7266277881320267, + "flos": 28724549967360.0, + "grad_norm": 3.5641362521491953, + "language_loss": 0.73911893, + "learning_rate": 7.338873803244076e-07, + "loss": 0.76133275, + "num_input_tokens_seen": 129946905, + "step": 6043, + "time_per_iteration": 2.754167079925537 + }, + { + "auxiliary_loss_clip": 0.01184744, + "auxiliary_loss_mlp": 0.01027663, + "balance_loss_clip": 1.04661679, + "balance_loss_mlp": 1.02077603, + "epoch": 0.7267480310226658, + "flos": 24863507182080.0, + "grad_norm": 2.006736100618886, + "language_loss": 0.80845463, + "learning_rate": 7.332844699443401e-07, + "loss": 0.83057868, + "num_input_tokens_seen": 129965505, + "step": 6044, + "time_per_iteration": 3.5816147327423096 + }, + { + "auxiliary_loss_clip": 0.01195182, + "auxiliary_loss_mlp": 0.01027966, + "balance_loss_clip": 1.04160345, + "balance_loss_mlp": 1.02064693, + "epoch": 0.7268682739133049, + "flos": 27198490694400.0, + "grad_norm": 2.027395796460133, + "language_loss": 0.75629508, + "learning_rate": 7.326817517240121e-07, + "loss": 0.77852654, + "num_input_tokens_seen": 129987210, + "step": 6045, + "time_per_iteration": 2.7765259742736816 + }, + { + "auxiliary_loss_clip": 0.01177996, + "auxiliary_loss_mlp": 0.0137827, + "balance_loss_clip": 1.04550576, + "balance_loss_mlp": 1.00017369, + "epoch": 0.7269885168039439, + "flos": 33508138688640.0, + "grad_norm": 3.0477389671211124, + "language_loss": 0.83498675, + "learning_rate": 7.320792257548545e-07, + "loss": 0.86054945, + "num_input_tokens_seen": 130008385, + "step": 6046, + "time_per_iteration": 2.751546621322632 + }, + { + "auxiliary_loss_clip": 0.01192709, + "auxiliary_loss_mlp": 0.01025426, + "balance_loss_clip": 1.04718685, + "balance_loss_mlp": 1.01815379, + "epoch": 0.7271087596945831, + "flos": 24313750548480.0, + "grad_norm": 2.0534527459289817, + "language_loss": 0.76294822, + "learning_rate": 7.314768921282704e-07, + "loss": 0.78512955, + "num_input_tokens_seen": 130029040, + "step": 6047, + "time_per_iteration": 2.64748215675354 + }, + { + "auxiliary_loss_clip": 0.01180802, + "auxiliary_loss_mlp": 0.01022722, + "balance_loss_clip": 1.04649973, + "balance_loss_mlp": 1.01558781, + "epoch": 0.7272290025852222, + "flos": 23805147922560.0, + "grad_norm": 3.2051106725683156, + "language_loss": 0.71583748, + "learning_rate": 7.30874750935633e-07, + "loss": 0.73787272, + "num_input_tokens_seen": 130048725, + "step": 6048, + "time_per_iteration": 2.7473857402801514 + }, + { + "auxiliary_loss_clip": 0.01196926, + "auxiliary_loss_mlp": 0.01023462, + "balance_loss_clip": 1.04597306, + "balance_loss_mlp": 1.01563585, + "epoch": 0.7273492454758612, + "flos": 16720367408640.0, + "grad_norm": 2.4381028293737526, + "language_loss": 0.79374325, + "learning_rate": 7.30272802268286e-07, + "loss": 0.81594718, + "num_input_tokens_seen": 130065720, + "step": 6049, + "time_per_iteration": 2.6383049488067627 + }, + { + "auxiliary_loss_clip": 0.01186062, + "auxiliary_loss_mlp": 0.01021996, + "balance_loss_clip": 1.03796315, + "balance_loss_mlp": 1.01568115, + "epoch": 0.7274694883665004, + "flos": 28031330413440.0, + "grad_norm": 1.7381472097121484, + "language_loss": 0.76301634, + "learning_rate": 7.29671046217547e-07, + "loss": 0.785097, + "num_input_tokens_seen": 130084830, + "step": 6050, + "time_per_iteration": 2.795030117034912 + }, + { + "auxiliary_loss_clip": 0.01198659, + "auxiliary_loss_mlp": 0.01026259, + "balance_loss_clip": 1.04493868, + "balance_loss_mlp": 1.0192256, + "epoch": 0.7275897312571394, + "flos": 30372706546560.0, + "grad_norm": 2.1704444121714506, + "language_loss": 0.8186928, + "learning_rate": 7.290694828746988e-07, + "loss": 0.84094197, + "num_input_tokens_seen": 130104495, + "step": 6051, + "time_per_iteration": 2.7544920444488525 + }, + { + "auxiliary_loss_clip": 0.01201128, + "auxiliary_loss_mlp": 0.01023469, + "balance_loss_clip": 1.04415345, + "balance_loss_mlp": 1.01631939, + "epoch": 0.7277099741477785, + "flos": 19204775498880.0, + "grad_norm": 1.824842360981672, + "language_loss": 0.8583374, + "learning_rate": 7.284681123310004e-07, + "loss": 0.88058341, + "num_input_tokens_seen": 130123210, + "step": 6052, + "time_per_iteration": 2.745758056640625 + }, + { + "auxiliary_loss_clip": 0.01176238, + "auxiliary_loss_mlp": 0.0102293, + "balance_loss_clip": 1.04612815, + "balance_loss_mlp": 1.01547348, + "epoch": 0.7278302170384175, + "flos": 20667884186880.0, + "grad_norm": 5.063608089053939, + "language_loss": 0.79702652, + "learning_rate": 7.27866934677678e-07, + "loss": 0.81901813, + "num_input_tokens_seen": 130142880, + "step": 6053, + "time_per_iteration": 2.574784755706787 + }, + { + "auxiliary_loss_clip": 0.01205174, + "auxiliary_loss_mlp": 0.01027467, + "balance_loss_clip": 1.04546309, + "balance_loss_mlp": 1.01983154, + "epoch": 0.7279504599290567, + "flos": 19093200877440.0, + "grad_norm": 2.0547814788472007, + "language_loss": 0.78082669, + "learning_rate": 7.272659500059297e-07, + "loss": 0.8031531, + "num_input_tokens_seen": 130160220, + "step": 6054, + "time_per_iteration": 2.744903087615967 + }, + { + "auxiliary_loss_clip": 0.0117043, + "auxiliary_loss_mlp": 0.01025998, + "balance_loss_clip": 1.04549599, + "balance_loss_mlp": 1.01801074, + "epoch": 0.7280707028196958, + "flos": 19062174504960.0, + "grad_norm": 6.7623281184743025, + "language_loss": 0.80330729, + "learning_rate": 7.266651584069264e-07, + "loss": 0.82527155, + "num_input_tokens_seen": 130177885, + "step": 6055, + "time_per_iteration": 2.6722373962402344 + }, + { + "auxiliary_loss_clip": 0.01185156, + "auxiliary_loss_mlp": 0.01030594, + "balance_loss_clip": 1.05119872, + "balance_loss_mlp": 1.02345371, + "epoch": 0.7281909457103348, + "flos": 37196308293120.0, + "grad_norm": 1.7971927792746718, + "language_loss": 0.56936973, + "learning_rate": 7.260645599718045e-07, + "loss": 0.59152722, + "num_input_tokens_seen": 130204240, + "step": 6056, + "time_per_iteration": 2.755206823348999 + }, + { + "auxiliary_loss_clip": 0.01188219, + "auxiliary_loss_mlp": 0.01024655, + "balance_loss_clip": 1.0459094, + "balance_loss_mlp": 1.01693618, + "epoch": 0.728311188600974, + "flos": 20667094087680.0, + "grad_norm": 2.196513876142502, + "language_loss": 0.67058754, + "learning_rate": 7.254641547916767e-07, + "loss": 0.6927163, + "num_input_tokens_seen": 130221735, + "step": 6057, + "time_per_iteration": 2.6978182792663574 + }, + { + "auxiliary_loss_clip": 0.01170917, + "auxiliary_loss_mlp": 0.01027579, + "balance_loss_clip": 1.05075407, + "balance_loss_mlp": 1.02008688, + "epoch": 0.728431431491613, + "flos": 28840685616000.0, + "grad_norm": 1.888765548044618, + "language_loss": 0.69200224, + "learning_rate": 7.248639429576226e-07, + "loss": 0.71398723, + "num_input_tokens_seen": 130241190, + "step": 6058, + "time_per_iteration": 2.630272150039673 + }, + { + "auxiliary_loss_clip": 0.01181372, + "auxiliary_loss_mlp": 0.01025829, + "balance_loss_clip": 1.04758406, + "balance_loss_mlp": 1.01840782, + "epoch": 0.7285516743822521, + "flos": 25991856092160.0, + "grad_norm": 2.055413048205087, + "language_loss": 0.71882492, + "learning_rate": 7.242639245606959e-07, + "loss": 0.74089694, + "num_input_tokens_seen": 130260980, + "step": 6059, + "time_per_iteration": 2.667597770690918 + }, + { + "auxiliary_loss_clip": 0.01190528, + "auxiliary_loss_mlp": 0.01024536, + "balance_loss_clip": 1.04303741, + "balance_loss_mlp": 1.01717532, + "epoch": 0.7286719172728913, + "flos": 16399721675520.0, + "grad_norm": 1.589839794554265, + "language_loss": 0.82265919, + "learning_rate": 7.236640996919168e-07, + "loss": 0.84480989, + "num_input_tokens_seen": 130280025, + "step": 6060, + "time_per_iteration": 2.768470525741577 + }, + { + "auxiliary_loss_clip": 0.01181528, + "auxiliary_loss_mlp": 0.01025386, + "balance_loss_clip": 1.04687738, + "balance_loss_mlp": 1.01825464, + "epoch": 0.7287921601635303, + "flos": 22018161277440.0, + "grad_norm": 1.7030264172915883, + "language_loss": 0.70640844, + "learning_rate": 7.230644684422782e-07, + "loss": 0.7284776, + "num_input_tokens_seen": 130300255, + "step": 6061, + "time_per_iteration": 2.638244152069092 + }, + { + "auxiliary_loss_clip": 0.01191639, + "auxiliary_loss_mlp": 0.01024467, + "balance_loss_clip": 1.042804, + "balance_loss_mlp": 1.01689124, + "epoch": 0.7289124030541694, + "flos": 24600927784320.0, + "grad_norm": 1.7337853267661987, + "language_loss": 0.81551641, + "learning_rate": 7.224650309027451e-07, + "loss": 0.83767748, + "num_input_tokens_seen": 130320005, + "step": 6062, + "time_per_iteration": 2.7018988132476807 + }, + { + "auxiliary_loss_clip": 0.01183937, + "auxiliary_loss_mlp": 0.01025625, + "balance_loss_clip": 1.04929936, + "balance_loss_mlp": 1.01877308, + "epoch": 0.7290326459448085, + "flos": 21393638484480.0, + "grad_norm": 2.7068119732793665, + "language_loss": 0.68438137, + "learning_rate": 7.218657871642506e-07, + "loss": 0.70647699, + "num_input_tokens_seen": 130338810, + "step": 6063, + "time_per_iteration": 3.4926767349243164 + }, + { + "auxiliary_loss_clip": 0.01172921, + "auxiliary_loss_mlp": 0.0102669, + "balance_loss_clip": 1.04960883, + "balance_loss_mlp": 1.01857758, + "epoch": 0.7291528888354476, + "flos": 18587686821120.0, + "grad_norm": 2.3374379585163823, + "language_loss": 0.62403363, + "learning_rate": 7.212667373177012e-07, + "loss": 0.64602971, + "num_input_tokens_seen": 130353805, + "step": 6064, + "time_per_iteration": 3.4809300899505615 + }, + { + "auxiliary_loss_clip": 0.01196482, + "auxiliary_loss_mlp": 0.01028678, + "balance_loss_clip": 1.04349566, + "balance_loss_mlp": 1.02161169, + "epoch": 0.7292731317260867, + "flos": 18951066760320.0, + "grad_norm": 1.861160195799337, + "language_loss": 0.75183237, + "learning_rate": 7.206678814539704e-07, + "loss": 0.77408397, + "num_input_tokens_seen": 130372105, + "step": 6065, + "time_per_iteration": 2.6677329540252686 + }, + { + "auxiliary_loss_clip": 0.0121623, + "auxiliary_loss_mlp": 0.01023178, + "balance_loss_clip": 1.04385948, + "balance_loss_mlp": 1.01636231, + "epoch": 0.7293933746167258, + "flos": 21067569797760.0, + "grad_norm": 1.5032020559890729, + "language_loss": 0.72695565, + "learning_rate": 7.20069219663904e-07, + "loss": 0.74934971, + "num_input_tokens_seen": 130391990, + "step": 6066, + "time_per_iteration": 2.7818243503570557 + }, + { + "auxiliary_loss_clip": 0.011798, + "auxiliary_loss_mlp": 0.01022812, + "balance_loss_clip": 1.04412043, + "balance_loss_mlp": 1.01489639, + "epoch": 0.7295136175073649, + "flos": 22453326547200.0, + "grad_norm": 1.6672513697665357, + "language_loss": 0.79251158, + "learning_rate": 7.1947075203832e-07, + "loss": 0.81453764, + "num_input_tokens_seen": 130411970, + "step": 6067, + "time_per_iteration": 2.6209566593170166 + }, + { + "auxiliary_loss_clip": 0.0106401, + "auxiliary_loss_mlp": 0.01000064, + "balance_loss_clip": 1.01101995, + "balance_loss_mlp": 0.99897903, + "epoch": 0.7296338603980039, + "flos": 56125506648960.0, + "grad_norm": 0.8648804799720224, + "language_loss": 0.60121965, + "learning_rate": 7.188724786680049e-07, + "loss": 0.62186038, + "num_input_tokens_seen": 130472440, + "step": 6068, + "time_per_iteration": 4.014440059661865 + }, + { + "auxiliary_loss_clip": 0.01188935, + "auxiliary_loss_mlp": 0.01024187, + "balance_loss_clip": 1.04491007, + "balance_loss_mlp": 1.01738, + "epoch": 0.7297541032886431, + "flos": 25228287751680.0, + "grad_norm": 1.5639542168003644, + "language_loss": 0.7561034, + "learning_rate": 7.182743996437162e-07, + "loss": 0.77823472, + "num_input_tokens_seen": 130491975, + "step": 6069, + "time_per_iteration": 3.6178057193756104 + }, + { + "auxiliary_loss_clip": 0.0120118, + "auxiliary_loss_mlp": 0.0103078, + "balance_loss_clip": 1.04309237, + "balance_loss_mlp": 1.02309716, + "epoch": 0.7298743461792822, + "flos": 26467600752000.0, + "grad_norm": 2.051943293240661, + "language_loss": 0.68636823, + "learning_rate": 7.176765150561819e-07, + "loss": 0.70868784, + "num_input_tokens_seen": 130510580, + "step": 6070, + "time_per_iteration": 2.7776057720184326 + }, + { + "auxiliary_loss_clip": 0.01170155, + "auxiliary_loss_mlp": 0.01028295, + "balance_loss_clip": 1.04513586, + "balance_loss_mlp": 1.02079356, + "epoch": 0.7299945890699212, + "flos": 19569053278080.0, + "grad_norm": 2.296087713104795, + "language_loss": 0.79930359, + "learning_rate": 7.170788249961002e-07, + "loss": 0.82128811, + "num_input_tokens_seen": 130529090, + "step": 6071, + "time_per_iteration": 2.628854513168335 + }, + { + "auxiliary_loss_clip": 0.01168567, + "auxiliary_loss_mlp": 0.01030756, + "balance_loss_clip": 1.04705346, + "balance_loss_mlp": 1.02344871, + "epoch": 0.7301148319605604, + "flos": 22928963466240.0, + "grad_norm": 1.9218451681829987, + "language_loss": 0.88114464, + "learning_rate": 7.164813295541418e-07, + "loss": 0.90313786, + "num_input_tokens_seen": 130548655, + "step": 6072, + "time_per_iteration": 2.573256254196167 + }, + { + "auxiliary_loss_clip": 0.01188307, + "auxiliary_loss_mlp": 0.01021794, + "balance_loss_clip": 1.0429734, + "balance_loss_mlp": 1.01457334, + "epoch": 0.7302350748511994, + "flos": 25369703596800.0, + "grad_norm": 1.6127446086862804, + "language_loss": 0.70415288, + "learning_rate": 7.15884028820944e-07, + "loss": 0.72625387, + "num_input_tokens_seen": 130567710, + "step": 6073, + "time_per_iteration": 2.711238145828247 + }, + { + "auxiliary_loss_clip": 0.01189245, + "auxiliary_loss_mlp": 0.01023056, + "balance_loss_clip": 1.0407474, + "balance_loss_mlp": 1.01581156, + "epoch": 0.7303553177418385, + "flos": 27819170732160.0, + "grad_norm": 2.0085240489354366, + "language_loss": 0.60349864, + "learning_rate": 7.152869228871185e-07, + "loss": 0.62562168, + "num_input_tokens_seen": 130590195, + "step": 6074, + "time_per_iteration": 2.7425546646118164 + }, + { + "auxiliary_loss_clip": 0.01180478, + "auxiliary_loss_mlp": 0.01026336, + "balance_loss_clip": 1.04381347, + "balance_loss_mlp": 1.01907659, + "epoch": 0.7304755606324776, + "flos": 24426510318720.0, + "grad_norm": 1.7996436249853922, + "language_loss": 0.72029638, + "learning_rate": 7.146900118432457e-07, + "loss": 0.74236453, + "num_input_tokens_seen": 130609940, + "step": 6075, + "time_per_iteration": 2.7288436889648438 + }, + { + "auxiliary_loss_clip": 0.01215403, + "auxiliary_loss_mlp": 0.01021198, + "balance_loss_clip": 1.03409803, + "balance_loss_mlp": 1.01393247, + "epoch": 0.7305958035231167, + "flos": 23840483927040.0, + "grad_norm": 1.5812634936764853, + "language_loss": 0.86049306, + "learning_rate": 7.140932957798753e-07, + "loss": 0.88285911, + "num_input_tokens_seen": 130628380, + "step": 6076, + "time_per_iteration": 2.8614115715026855 + }, + { + "auxiliary_loss_clip": 0.01190218, + "auxiliary_loss_mlp": 0.01025794, + "balance_loss_clip": 1.04406655, + "balance_loss_mlp": 1.01835775, + "epoch": 0.7307160464137558, + "flos": 16726939597440.0, + "grad_norm": 2.156049211047905, + "language_loss": 0.71731395, + "learning_rate": 7.134967747875309e-07, + "loss": 0.73947406, + "num_input_tokens_seen": 130646590, + "step": 6077, + "time_per_iteration": 3.0485148429870605 + }, + { + "auxiliary_loss_clip": 0.01173466, + "auxiliary_loss_mlp": 0.0102741, + "balance_loss_clip": 1.04532385, + "balance_loss_mlp": 1.01967382, + "epoch": 0.7308362893043949, + "flos": 21798280172160.0, + "grad_norm": 1.8405196496961436, + "language_loss": 0.81755722, + "learning_rate": 7.129004489567014e-07, + "loss": 0.83956593, + "num_input_tokens_seen": 130664070, + "step": 6078, + "time_per_iteration": 2.6702964305877686 + }, + { + "auxiliary_loss_clip": 0.01199949, + "auxiliary_loss_mlp": 0.01025554, + "balance_loss_clip": 1.04400802, + "balance_loss_mlp": 1.01828575, + "epoch": 0.730956532195034, + "flos": 10707377840640.0, + "grad_norm": 2.3871242478863124, + "language_loss": 0.77941382, + "learning_rate": 7.123043183778512e-07, + "loss": 0.80166894, + "num_input_tokens_seen": 130681400, + "step": 6079, + "time_per_iteration": 2.6576285362243652 + }, + { + "auxiliary_loss_clip": 0.01200578, + "auxiliary_loss_mlp": 0.01022225, + "balance_loss_clip": 1.0448072, + "balance_loss_mlp": 1.01484609, + "epoch": 0.731076775085673, + "flos": 19791987039360.0, + "grad_norm": 1.5277777787617683, + "language_loss": 0.65324593, + "learning_rate": 7.117083831414114e-07, + "loss": 0.67547393, + "num_input_tokens_seen": 130700675, + "step": 6080, + "time_per_iteration": 2.708158016204834 + }, + { + "auxiliary_loss_clip": 0.01167625, + "auxiliary_loss_mlp": 0.01022368, + "balance_loss_clip": 1.04830146, + "balance_loss_mlp": 1.01480389, + "epoch": 0.7311970179763122, + "flos": 20447033414400.0, + "grad_norm": 1.9324290443151089, + "language_loss": 0.69625211, + "learning_rate": 7.11112643337787e-07, + "loss": 0.71815205, + "num_input_tokens_seen": 130719720, + "step": 6081, + "time_per_iteration": 2.525129795074463 + }, + { + "auxiliary_loss_clip": 0.01190135, + "auxiliary_loss_mlp": 0.01021788, + "balance_loss_clip": 1.04708076, + "balance_loss_mlp": 1.01416469, + "epoch": 0.7313172608669513, + "flos": 18513818501760.0, + "grad_norm": 2.4577456202253254, + "language_loss": 0.7636953, + "learning_rate": 7.10517099057349e-07, + "loss": 0.78581452, + "num_input_tokens_seen": 130736670, + "step": 6082, + "time_per_iteration": 2.7003471851348877 + }, + { + "auxiliary_loss_clip": 0.01186088, + "auxiliary_loss_mlp": 0.01028351, + "balance_loss_clip": 1.04417968, + "balance_loss_mlp": 1.01985145, + "epoch": 0.7314375037575903, + "flos": 16180738410240.0, + "grad_norm": 4.227569147057043, + "language_loss": 0.61408579, + "learning_rate": 7.099217503904411e-07, + "loss": 0.63623023, + "num_input_tokens_seen": 130754525, + "step": 6083, + "time_per_iteration": 2.660508155822754 + }, + { + "auxiliary_loss_clip": 0.01191381, + "auxiliary_loss_mlp": 0.01027604, + "balance_loss_clip": 1.04501843, + "balance_loss_mlp": 1.02020681, + "epoch": 0.7315577466482295, + "flos": 17967940536960.0, + "grad_norm": 2.8762190811251775, + "language_loss": 0.89965397, + "learning_rate": 7.093265974273788e-07, + "loss": 0.92184389, + "num_input_tokens_seen": 130772420, + "step": 6084, + "time_per_iteration": 2.647650718688965 + }, + { + "auxiliary_loss_clip": 0.01178977, + "auxiliary_loss_mlp": 0.01023679, + "balance_loss_clip": 1.04539371, + "balance_loss_mlp": 1.016819, + "epoch": 0.7316779895388685, + "flos": 18405440190720.0, + "grad_norm": 1.8511632274531677, + "language_loss": 0.71864903, + "learning_rate": 7.087316402584447e-07, + "loss": 0.74067557, + "num_input_tokens_seen": 130791245, + "step": 6085, + "time_per_iteration": 2.5818440914154053 + }, + { + "auxiliary_loss_clip": 0.01170522, + "auxiliary_loss_mlp": 0.01029632, + "balance_loss_clip": 1.04906154, + "balance_loss_mlp": 1.0220505, + "epoch": 0.7317982324295076, + "flos": 17928294900480.0, + "grad_norm": 2.4956223611838655, + "language_loss": 0.86670518, + "learning_rate": 7.081368789738953e-07, + "loss": 0.8887068, + "num_input_tokens_seen": 130808445, + "step": 6086, + "time_per_iteration": 2.5931246280670166 + }, + { + "auxiliary_loss_clip": 0.01180357, + "auxiliary_loss_mlp": 0.0102765, + "balance_loss_clip": 1.04202676, + "balance_loss_mlp": 1.01982939, + "epoch": 0.7319184753201466, + "flos": 27229840289280.0, + "grad_norm": 1.9789066006958729, + "language_loss": 0.77718216, + "learning_rate": 7.075423136639537e-07, + "loss": 0.79926223, + "num_input_tokens_seen": 130827700, + "step": 6087, + "time_per_iteration": 2.666550874710083 + }, + { + "auxiliary_loss_clip": 0.01189039, + "auxiliary_loss_mlp": 0.01025022, + "balance_loss_clip": 1.04215217, + "balance_loss_mlp": 1.01755929, + "epoch": 0.7320387182107858, + "flos": 37448544574080.0, + "grad_norm": 1.689474199829158, + "language_loss": 0.74596494, + "learning_rate": 7.069479444188149e-07, + "loss": 0.76810557, + "num_input_tokens_seen": 130848290, + "step": 6088, + "time_per_iteration": 2.852756977081299 + }, + { + "auxiliary_loss_clip": 0.01177115, + "auxiliary_loss_mlp": 0.0102565, + "balance_loss_clip": 1.04412234, + "balance_loss_mlp": 1.01737058, + "epoch": 0.7321589611014249, + "flos": 17859023521920.0, + "grad_norm": 1.7231034086182144, + "language_loss": 0.82297993, + "learning_rate": 7.063537713286453e-07, + "loss": 0.8450076, + "num_input_tokens_seen": 130865970, + "step": 6089, + "time_per_iteration": 3.5424041748046875 + }, + { + "auxiliary_loss_clip": 0.01192398, + "auxiliary_loss_mlp": 0.01023442, + "balance_loss_clip": 1.04463923, + "balance_loss_mlp": 1.01608443, + "epoch": 0.7322792039920639, + "flos": 26100593539200.0, + "grad_norm": 3.307800292984502, + "language_loss": 0.80989897, + "learning_rate": 7.057597944835803e-07, + "loss": 0.83205736, + "num_input_tokens_seen": 130885245, + "step": 6090, + "time_per_iteration": 3.4782044887542725 + }, + { + "auxiliary_loss_clip": 0.0120048, + "auxiliary_loss_mlp": 0.01028589, + "balance_loss_clip": 1.0426327, + "balance_loss_mlp": 1.02120447, + "epoch": 0.7323994468827031, + "flos": 25369093065600.0, + "grad_norm": 1.924366011861328, + "language_loss": 0.74741125, + "learning_rate": 7.051660139737253e-07, + "loss": 0.76970184, + "num_input_tokens_seen": 130903465, + "step": 6091, + "time_per_iteration": 2.688746213912964 + }, + { + "auxiliary_loss_clip": 0.01175503, + "auxiliary_loss_mlp": 0.01378747, + "balance_loss_clip": 1.04688644, + "balance_loss_mlp": 1.00022411, + "epoch": 0.7325196897733421, + "flos": 26907075653760.0, + "grad_norm": 2.2724708887975593, + "language_loss": 0.76454926, + "learning_rate": 7.045724298891565e-07, + "loss": 0.79009175, + "num_input_tokens_seen": 130922935, + "step": 6092, + "time_per_iteration": 2.6312081813812256 + }, + { + "auxiliary_loss_clip": 0.01178267, + "auxiliary_loss_mlp": 0.01024752, + "balance_loss_clip": 1.04663444, + "balance_loss_mlp": 1.01767075, + "epoch": 0.7326399326639812, + "flos": 25775781828480.0, + "grad_norm": 1.948586851555813, + "language_loss": 0.69294655, + "learning_rate": 7.039790423199192e-07, + "loss": 0.71497679, + "num_input_tokens_seen": 130942575, + "step": 6093, + "time_per_iteration": 2.6971118450164795 + }, + { + "auxiliary_loss_clip": 0.01189344, + "auxiliary_loss_mlp": 0.01024712, + "balance_loss_clip": 1.04453897, + "balance_loss_mlp": 1.01764917, + "epoch": 0.7327601755546204, + "flos": 21032269706880.0, + "grad_norm": 2.587289352474979, + "language_loss": 0.78003788, + "learning_rate": 7.033858513560322e-07, + "loss": 0.80217838, + "num_input_tokens_seen": 130958870, + "step": 6094, + "time_per_iteration": 3.609234094619751 + }, + { + "auxiliary_loss_clip": 0.01180559, + "auxiliary_loss_mlp": 0.01025292, + "balance_loss_clip": 1.0476079, + "balance_loss_mlp": 1.01821053, + "epoch": 0.7328804184452594, + "flos": 16289224462080.0, + "grad_norm": 2.4707709147557053, + "language_loss": 0.7634908, + "learning_rate": 7.027928570874794e-07, + "loss": 0.78554928, + "num_input_tokens_seen": 130977060, + "step": 6095, + "time_per_iteration": 2.579631805419922 + }, + { + "auxiliary_loss_clip": 0.01168628, + "auxiliary_loss_mlp": 0.01028469, + "balance_loss_clip": 1.0469507, + "balance_loss_mlp": 1.02136135, + "epoch": 0.7330006613358985, + "flos": 17858233422720.0, + "grad_norm": 1.9434185742390278, + "language_loss": 0.85688603, + "learning_rate": 7.022000596042194e-07, + "loss": 0.87885702, + "num_input_tokens_seen": 130994160, + "step": 6096, + "time_per_iteration": 3.4990663528442383 + }, + { + "auxiliary_loss_clip": 0.01197036, + "auxiliary_loss_mlp": 0.01023535, + "balance_loss_clip": 1.04065609, + "balance_loss_mlp": 1.01618552, + "epoch": 0.7331209042265376, + "flos": 22492074343680.0, + "grad_norm": 2.13268038240026, + "language_loss": 0.81863928, + "learning_rate": 7.016074589961784e-07, + "loss": 0.84084499, + "num_input_tokens_seen": 131012725, + "step": 6097, + "time_per_iteration": 2.7133679389953613 + }, + { + "auxiliary_loss_clip": 0.01182107, + "auxiliary_loss_mlp": 0.01028419, + "balance_loss_clip": 1.044204, + "balance_loss_mlp": 1.02111709, + "epoch": 0.7332411471171767, + "flos": 33072757937280.0, + "grad_norm": 1.7187253210942752, + "language_loss": 0.66960561, + "learning_rate": 7.01015055353253e-07, + "loss": 0.69171083, + "num_input_tokens_seen": 131035150, + "step": 6098, + "time_per_iteration": 2.7663004398345947 + }, + { + "auxiliary_loss_clip": 0.01193778, + "auxiliary_loss_mlp": 0.01031669, + "balance_loss_clip": 1.04161763, + "balance_loss_mlp": 1.0239383, + "epoch": 0.7333613900078157, + "flos": 22743017735040.0, + "grad_norm": 1.710120058354996, + "language_loss": 0.77775049, + "learning_rate": 7.004228487653123e-07, + "loss": 0.80000496, + "num_input_tokens_seen": 131055955, + "step": 6099, + "time_per_iteration": 2.7863609790802 + }, + { + "auxiliary_loss_clip": 0.0118925, + "auxiliary_loss_mlp": 0.01029655, + "balance_loss_clip": 1.03970313, + "balance_loss_mlp": 1.02235377, + "epoch": 0.7334816328984549, + "flos": 22346133384960.0, + "grad_norm": 1.8865169753340905, + "language_loss": 0.78291655, + "learning_rate": 6.998308393221906e-07, + "loss": 0.80510557, + "num_input_tokens_seen": 131074360, + "step": 6100, + "time_per_iteration": 2.6910626888275146 + }, + { + "auxiliary_loss_clip": 0.01199626, + "auxiliary_loss_mlp": 0.01029059, + "balance_loss_clip": 1.04607797, + "balance_loss_mlp": 1.02209735, + "epoch": 0.733601875789094, + "flos": 20736149984640.0, + "grad_norm": 2.101593586581799, + "language_loss": 0.71002603, + "learning_rate": 6.992390271136977e-07, + "loss": 0.73231292, + "num_input_tokens_seen": 131090070, + "step": 6101, + "time_per_iteration": 2.6742427349090576 + }, + { + "auxiliary_loss_clip": 0.01170685, + "auxiliary_loss_mlp": 0.01027222, + "balance_loss_clip": 1.04438806, + "balance_loss_mlp": 1.01956868, + "epoch": 0.733722118679733, + "flos": 22564362464640.0, + "grad_norm": 1.8618780843678522, + "language_loss": 0.85445005, + "learning_rate": 6.986474122296094e-07, + "loss": 0.87642908, + "num_input_tokens_seen": 131109185, + "step": 6102, + "time_per_iteration": 2.612504243850708 + }, + { + "auxiliary_loss_clip": 0.01173031, + "auxiliary_loss_mlp": 0.01026272, + "balance_loss_clip": 1.04922462, + "balance_loss_mlp": 1.01817822, + "epoch": 0.7338423615703722, + "flos": 20084192179200.0, + "grad_norm": 1.8924541459721953, + "language_loss": 0.72256911, + "learning_rate": 6.980559947596751e-07, + "loss": 0.74456215, + "num_input_tokens_seen": 131127725, + "step": 6103, + "time_per_iteration": 2.552619218826294 + }, + { + "auxiliary_loss_clip": 0.01203472, + "auxiliary_loss_mlp": 0.01023715, + "balance_loss_clip": 1.04179049, + "balance_loss_mlp": 1.01628196, + "epoch": 0.7339626044610112, + "flos": 21687675217920.0, + "grad_norm": 1.9151152012544503, + "language_loss": 0.75978482, + "learning_rate": 6.974647747936109e-07, + "loss": 0.78205669, + "num_input_tokens_seen": 131146110, + "step": 6104, + "time_per_iteration": 2.743434429168701 + }, + { + "auxiliary_loss_clip": 0.01171518, + "auxiliary_loss_mlp": 0.01379022, + "balance_loss_clip": 1.0487299, + "balance_loss_mlp": 1.00015557, + "epoch": 0.7340828473516503, + "flos": 15268248282240.0, + "grad_norm": 2.0788996162751303, + "language_loss": 0.82504153, + "learning_rate": 6.968737524211039e-07, + "loss": 0.85054702, + "num_input_tokens_seen": 131162920, + "step": 6105, + "time_per_iteration": 2.581481695175171 + }, + { + "auxiliary_loss_clip": 0.0118092, + "auxiliary_loss_mlp": 0.01025655, + "balance_loss_clip": 1.04969299, + "balance_loss_mlp": 1.01792407, + "epoch": 0.7342030902422895, + "flos": 22930112701440.0, + "grad_norm": 5.310867227557299, + "language_loss": 0.80077994, + "learning_rate": 6.962829277318132e-07, + "loss": 0.82284576, + "num_input_tokens_seen": 131182515, + "step": 6106, + "time_per_iteration": 2.5882279872894287 + }, + { + "auxiliary_loss_clip": 0.01181153, + "auxiliary_loss_mlp": 0.01023725, + "balance_loss_clip": 1.0487479, + "balance_loss_mlp": 1.0165782, + "epoch": 0.7343233331329285, + "flos": 25847890381440.0, + "grad_norm": 1.7736854091350782, + "language_loss": 0.83636844, + "learning_rate": 6.956923008153652e-07, + "loss": 0.85841721, + "num_input_tokens_seen": 131202280, + "step": 6107, + "time_per_iteration": 2.6737921237945557 + }, + { + "auxiliary_loss_clip": 0.0117918, + "auxiliary_loss_mlp": 0.01027186, + "balance_loss_clip": 1.04479468, + "balance_loss_mlp": 1.01989603, + "epoch": 0.7344435760235676, + "flos": 18478985287680.0, + "grad_norm": 1.9917501563887448, + "language_loss": 0.84426433, + "learning_rate": 6.951018717613593e-07, + "loss": 0.866328, + "num_input_tokens_seen": 131221295, + "step": 6108, + "time_per_iteration": 2.606050968170166 + }, + { + "auxiliary_loss_clip": 0.01176683, + "auxiliary_loss_mlp": 0.01029462, + "balance_loss_clip": 1.04664326, + "balance_loss_mlp": 1.02176142, + "epoch": 0.7345638189142067, + "flos": 17640040256640.0, + "grad_norm": 1.987261410983196, + "language_loss": 0.77980667, + "learning_rate": 6.945116406593614e-07, + "loss": 0.80186808, + "num_input_tokens_seen": 131240150, + "step": 6109, + "time_per_iteration": 2.615793228149414 + }, + { + "auxiliary_loss_clip": 0.0120908, + "auxiliary_loss_mlp": 0.01026035, + "balance_loss_clip": 1.04456532, + "balance_loss_mlp": 1.01882267, + "epoch": 0.7346840618048458, + "flos": 20260225756800.0, + "grad_norm": 2.432100175884562, + "language_loss": 0.74331546, + "learning_rate": 6.939216075989089e-07, + "loss": 0.7656666, + "num_input_tokens_seen": 131258080, + "step": 6110, + "time_per_iteration": 2.6850364208221436 + }, + { + "auxiliary_loss_clip": 0.01183897, + "auxiliary_loss_mlp": 0.01024016, + "balance_loss_clip": 1.0438931, + "balance_loss_mlp": 1.01669681, + "epoch": 0.7348043046954849, + "flos": 29023183641600.0, + "grad_norm": 2.628910261287824, + "language_loss": 0.65708685, + "learning_rate": 6.933317726695109e-07, + "loss": 0.67916596, + "num_input_tokens_seen": 131279310, + "step": 6111, + "time_per_iteration": 2.733431577682495 + }, + { + "auxiliary_loss_clip": 0.01195023, + "auxiliary_loss_mlp": 0.01026134, + "balance_loss_clip": 1.04624963, + "balance_loss_mlp": 1.01913655, + "epoch": 0.734924547586124, + "flos": 17931203902080.0, + "grad_norm": 2.8800822542652593, + "language_loss": 0.79872346, + "learning_rate": 6.92742135960644e-07, + "loss": 0.82093501, + "num_input_tokens_seen": 131297010, + "step": 6112, + "time_per_iteration": 2.695558547973633 + }, + { + "auxiliary_loss_clip": 0.01078072, + "auxiliary_loss_mlp": 0.00999653, + "balance_loss_clip": 1.01154435, + "balance_loss_mlp": 0.99857962, + "epoch": 0.7350447904767631, + "flos": 63588319850880.0, + "grad_norm": 0.8256245968338667, + "language_loss": 0.55724812, + "learning_rate": 6.921526975617556e-07, + "loss": 0.57802534, + "num_input_tokens_seen": 131356470, + "step": 6113, + "time_per_iteration": 3.2586100101470947 + }, + { + "auxiliary_loss_clip": 0.01188214, + "auxiliary_loss_mlp": 0.01025368, + "balance_loss_clip": 1.04413962, + "balance_loss_mlp": 1.01786423, + "epoch": 0.7351650333674021, + "flos": 21580015178880.0, + "grad_norm": 1.8594715950715932, + "language_loss": 0.75475806, + "learning_rate": 6.915634575622631e-07, + "loss": 0.77689397, + "num_input_tokens_seen": 131374985, + "step": 6114, + "time_per_iteration": 2.7182188034057617 + }, + { + "auxiliary_loss_clip": 0.01167562, + "auxiliary_loss_mlp": 0.01024272, + "balance_loss_clip": 1.04717326, + "balance_loss_mlp": 1.01711977, + "epoch": 0.7352852762580413, + "flos": 18186349184640.0, + "grad_norm": 1.8218099064950792, + "language_loss": 0.70590562, + "learning_rate": 6.909744160515532e-07, + "loss": 0.72782397, + "num_input_tokens_seen": 131393125, + "step": 6115, + "time_per_iteration": 3.4187064170837402 + }, + { + "auxiliary_loss_clip": 0.01185302, + "auxiliary_loss_mlp": 0.01028757, + "balance_loss_clip": 1.04642463, + "balance_loss_mlp": 1.02147961, + "epoch": 0.7354055191486804, + "flos": 38910073063680.0, + "grad_norm": 1.9020318262735258, + "language_loss": 0.69204116, + "learning_rate": 6.903855731189849e-07, + "loss": 0.71418172, + "num_input_tokens_seen": 131415760, + "step": 6116, + "time_per_iteration": 3.696760416030884 + }, + { + "auxiliary_loss_clip": 0.01194503, + "auxiliary_loss_mlp": 0.01023349, + "balance_loss_clip": 1.04505134, + "balance_loss_mlp": 1.0155766, + "epoch": 0.7355257620393194, + "flos": 16289978647680.0, + "grad_norm": 2.7182622772274505, + "language_loss": 0.81738555, + "learning_rate": 6.897969288538825e-07, + "loss": 0.83956409, + "num_input_tokens_seen": 131433705, + "step": 6117, + "time_per_iteration": 2.6538214683532715 + }, + { + "auxiliary_loss_clip": 0.01182573, + "auxiliary_loss_mlp": 0.01024545, + "balance_loss_clip": 1.04439688, + "balance_loss_mlp": 1.01691008, + "epoch": 0.7356460049299585, + "flos": 18114240631680.0, + "grad_norm": 2.050623683417816, + "language_loss": 0.80883414, + "learning_rate": 6.892084833455452e-07, + "loss": 0.83090532, + "num_input_tokens_seen": 131453275, + "step": 6118, + "time_per_iteration": 2.616422176361084 + }, + { + "auxiliary_loss_clip": 0.01175899, + "auxiliary_loss_mlp": 0.01024698, + "balance_loss_clip": 1.04664922, + "balance_loss_mlp": 1.0169524, + "epoch": 0.7357662478205976, + "flos": 21325193118720.0, + "grad_norm": 1.4283116396424362, + "language_loss": 0.83719414, + "learning_rate": 6.886202366832384e-07, + "loss": 0.85920012, + "num_input_tokens_seen": 131474960, + "step": 6119, + "time_per_iteration": 2.637176752090454 + }, + { + "auxiliary_loss_clip": 0.01201679, + "auxiliary_loss_mlp": 0.01023827, + "balance_loss_clip": 1.04223013, + "balance_loss_mlp": 1.01639438, + "epoch": 0.7358864907112367, + "flos": 14246841139200.0, + "grad_norm": 1.8240992960575757, + "language_loss": 0.73586786, + "learning_rate": 6.880321889561987e-07, + "loss": 0.75812292, + "num_input_tokens_seen": 131492935, + "step": 6120, + "time_per_iteration": 3.5742530822753906 + }, + { + "auxiliary_loss_clip": 0.0118775, + "auxiliary_loss_mlp": 0.01027621, + "balance_loss_clip": 1.04217863, + "balance_loss_mlp": 1.01933551, + "epoch": 0.7360067336018757, + "flos": 22309684058880.0, + "grad_norm": 3.892944045466075, + "language_loss": 0.65314102, + "learning_rate": 6.874443402536338e-07, + "loss": 0.67529476, + "num_input_tokens_seen": 131512025, + "step": 6121, + "time_per_iteration": 3.562696933746338 + }, + { + "auxiliary_loss_clip": 0.01192908, + "auxiliary_loss_mlp": 0.01024564, + "balance_loss_clip": 1.0456661, + "balance_loss_mlp": 1.01688719, + "epoch": 0.7361269764925149, + "flos": 25554607833600.0, + "grad_norm": 1.6580072515049353, + "language_loss": 0.80550313, + "learning_rate": 6.868566906647177e-07, + "loss": 0.82767785, + "num_input_tokens_seen": 131532975, + "step": 6122, + "time_per_iteration": 2.732435464859009 + }, + { + "auxiliary_loss_clip": 0.01177414, + "auxiliary_loss_mlp": 0.01026815, + "balance_loss_clip": 1.04558873, + "balance_loss_mlp": 1.01926351, + "epoch": 0.736247219383154, + "flos": 20376505059840.0, + "grad_norm": 1.8734304899695977, + "language_loss": 0.835585, + "learning_rate": 6.862692402785984e-07, + "loss": 0.85762727, + "num_input_tokens_seen": 131553225, + "step": 6123, + "time_per_iteration": 2.6375269889831543 + }, + { + "auxiliary_loss_clip": 0.01104854, + "auxiliary_loss_mlp": 0.01000557, + "balance_loss_clip": 1.01776195, + "balance_loss_mlp": 0.9995138, + "epoch": 0.736367462273793, + "flos": 70339525735680.0, + "grad_norm": 0.6857274791874975, + "language_loss": 0.49633551, + "learning_rate": 6.856819891843899e-07, + "loss": 0.51738966, + "num_input_tokens_seen": 131617930, + "step": 6124, + "time_per_iteration": 3.4358608722686768 + }, + { + "auxiliary_loss_clip": 0.01215665, + "auxiliary_loss_mlp": 0.01025922, + "balance_loss_clip": 1.04369259, + "balance_loss_mlp": 1.01869154, + "epoch": 0.7364877051644322, + "flos": 22412711243520.0, + "grad_norm": 1.9262928043299417, + "language_loss": 0.72356665, + "learning_rate": 6.8509493747118e-07, + "loss": 0.74598253, + "num_input_tokens_seen": 131636740, + "step": 6125, + "time_per_iteration": 2.842404365539551 + }, + { + "auxiliary_loss_clip": 0.01171959, + "auxiliary_loss_mlp": 0.01021867, + "balance_loss_clip": 1.05002737, + "balance_loss_mlp": 1.01461649, + "epoch": 0.7366079480550712, + "flos": 12130266274560.0, + "grad_norm": 2.203016742862321, + "language_loss": 0.88399684, + "learning_rate": 6.845080852280221e-07, + "loss": 0.90593511, + "num_input_tokens_seen": 131653810, + "step": 6126, + "time_per_iteration": 2.7948224544525146 + }, + { + "auxiliary_loss_clip": 0.01198153, + "auxiliary_loss_mlp": 0.01023324, + "balance_loss_clip": 1.04401433, + "balance_loss_mlp": 1.01620162, + "epoch": 0.7367281909457103, + "flos": 15049336844160.0, + "grad_norm": 1.7103036709410797, + "language_loss": 0.74070209, + "learning_rate": 6.839214325439409e-07, + "loss": 0.76291692, + "num_input_tokens_seen": 131671505, + "step": 6127, + "time_per_iteration": 2.751359224319458 + }, + { + "auxiliary_loss_clip": 0.01180794, + "auxiliary_loss_mlp": 0.01024709, + "balance_loss_clip": 1.04651022, + "balance_loss_mlp": 1.01804554, + "epoch": 0.7368484338363495, + "flos": 23510752053120.0, + "grad_norm": 1.6333779330384006, + "language_loss": 0.71729165, + "learning_rate": 6.833349795079327e-07, + "loss": 0.73934668, + "num_input_tokens_seen": 131690615, + "step": 6128, + "time_per_iteration": 2.635765314102173 + }, + { + "auxiliary_loss_clip": 0.01196412, + "auxiliary_loss_mlp": 0.01026779, + "balance_loss_clip": 1.04605818, + "balance_loss_mlp": 1.01952195, + "epoch": 0.7369686767269885, + "flos": 27417833095680.0, + "grad_norm": 1.5801470890957883, + "language_loss": 0.68780017, + "learning_rate": 6.827487262089613e-07, + "loss": 0.71003205, + "num_input_tokens_seen": 131711120, + "step": 6129, + "time_per_iteration": 2.7348692417144775 + }, + { + "auxiliary_loss_clip": 0.01086009, + "auxiliary_loss_mlp": 0.00999962, + "balance_loss_clip": 1.01331329, + "balance_loss_mlp": 0.99879944, + "epoch": 0.7370889196176276, + "flos": 70293343824000.0, + "grad_norm": 0.9592922839064266, + "language_loss": 0.56863147, + "learning_rate": 6.821626727359606e-07, + "loss": 0.58949119, + "num_input_tokens_seen": 131776680, + "step": 6130, + "time_per_iteration": 3.3855865001678467 + }, + { + "auxiliary_loss_clip": 0.01188593, + "auxiliary_loss_mlp": 0.01029333, + "balance_loss_clip": 1.04809451, + "balance_loss_mlp": 1.02097082, + "epoch": 0.7372091625082667, + "flos": 18040839189120.0, + "grad_norm": 2.125959925839556, + "language_loss": 0.76761591, + "learning_rate": 6.815768191778348e-07, + "loss": 0.78979522, + "num_input_tokens_seen": 131794760, + "step": 6131, + "time_per_iteration": 2.7181568145751953 + }, + { + "auxiliary_loss_clip": 0.01172094, + "auxiliary_loss_mlp": 0.01026163, + "balance_loss_clip": 1.04439163, + "balance_loss_mlp": 1.01814055, + "epoch": 0.7373294053989058, + "flos": 33726331854720.0, + "grad_norm": 2.1198037842604722, + "language_loss": 0.73287576, + "learning_rate": 6.809911656234569e-07, + "loss": 0.75485831, + "num_input_tokens_seen": 131816735, + "step": 6132, + "time_per_iteration": 2.7158570289611816 + }, + { + "auxiliary_loss_clip": 0.01197064, + "auxiliary_loss_mlp": 0.0102394, + "balance_loss_clip": 1.04245043, + "balance_loss_mlp": 1.01640582, + "epoch": 0.7374496482895448, + "flos": 21506326427520.0, + "grad_norm": 2.428602642110819, + "language_loss": 0.7824353, + "learning_rate": 6.804057121616707e-07, + "loss": 0.80464542, + "num_input_tokens_seen": 131834940, + "step": 6133, + "time_per_iteration": 2.7394402027130127 + }, + { + "auxiliary_loss_clip": 0.01178607, + "auxiliary_loss_mlp": 0.01023159, + "balance_loss_clip": 1.04487002, + "balance_loss_mlp": 1.01621795, + "epoch": 0.737569891180184, + "flos": 24936908624640.0, + "grad_norm": 2.001886420827213, + "language_loss": 0.72026873, + "learning_rate": 6.798204588812888e-07, + "loss": 0.74228638, + "num_input_tokens_seen": 131854355, + "step": 6134, + "time_per_iteration": 2.6419131755828857 + }, + { + "auxiliary_loss_clip": 0.01201672, + "auxiliary_loss_mlp": 0.01378905, + "balance_loss_clip": 1.03916979, + "balance_loss_mlp": 1.00018299, + "epoch": 0.7376901340708231, + "flos": 20664544222080.0, + "grad_norm": 1.6237481247345276, + "language_loss": 0.75456381, + "learning_rate": 6.792354058710937e-07, + "loss": 0.78036958, + "num_input_tokens_seen": 131871825, + "step": 6135, + "time_per_iteration": 2.744086742401123 + }, + { + "auxiliary_loss_clip": 0.01165244, + "auxiliary_loss_mlp": 0.01027069, + "balance_loss_clip": 1.04726648, + "balance_loss_mlp": 1.02011299, + "epoch": 0.7378103769614621, + "flos": 23805794367360.0, + "grad_norm": 1.962872091608032, + "language_loss": 0.65098351, + "learning_rate": 6.786505532198374e-07, + "loss": 0.67290664, + "num_input_tokens_seen": 131890770, + "step": 6136, + "time_per_iteration": 2.606752872467041 + }, + { + "auxiliary_loss_clip": 0.01170322, + "auxiliary_loss_mlp": 0.01023044, + "balance_loss_clip": 1.04809022, + "balance_loss_mlp": 1.0154053, + "epoch": 0.7379306198521013, + "flos": 22237216369920.0, + "grad_norm": 1.651187159798726, + "language_loss": 0.85410726, + "learning_rate": 6.780659010162411e-07, + "loss": 0.87604094, + "num_input_tokens_seen": 131909720, + "step": 6137, + "time_per_iteration": 2.665950298309326 + }, + { + "auxiliary_loss_clip": 0.01199768, + "auxiliary_loss_mlp": 0.01032129, + "balance_loss_clip": 1.04527879, + "balance_loss_mlp": 1.02496433, + "epoch": 0.7380508627427403, + "flos": 14903108576640.0, + "grad_norm": 1.6395403843647411, + "language_loss": 0.83049303, + "learning_rate": 6.774814493489975e-07, + "loss": 0.85281199, + "num_input_tokens_seen": 131927395, + "step": 6138, + "time_per_iteration": 2.7712533473968506 + }, + { + "auxiliary_loss_clip": 0.01174608, + "auxiliary_loss_mlp": 0.01023211, + "balance_loss_clip": 1.04569542, + "balance_loss_mlp": 1.01593935, + "epoch": 0.7381711056333794, + "flos": 21685843624320.0, + "grad_norm": 1.6499957548157946, + "language_loss": 0.6593861, + "learning_rate": 6.768971983067655e-07, + "loss": 0.6813643, + "num_input_tokens_seen": 131947725, + "step": 6139, + "time_per_iteration": 2.6053121089935303 + }, + { + "auxiliary_loss_clip": 0.01064396, + "auxiliary_loss_mlp": 0.00999157, + "balance_loss_clip": 1.0114857, + "balance_loss_mlp": 0.99804252, + "epoch": 0.7382913485240186, + "flos": 52404263596800.0, + "grad_norm": 1.01444410376606, + "language_loss": 0.67855954, + "learning_rate": 6.763131479781772e-07, + "loss": 0.69919515, + "num_input_tokens_seen": 131997485, + "step": 6140, + "time_per_iteration": 2.9848392009735107 + }, + { + "auxiliary_loss_clip": 0.01181999, + "auxiliary_loss_mlp": 0.01026909, + "balance_loss_clip": 1.04768336, + "balance_loss_mlp": 1.01970911, + "epoch": 0.7384115914146576, + "flos": 21798818876160.0, + "grad_norm": 1.8052762464520709, + "language_loss": 0.7633574, + "learning_rate": 6.757292984518316e-07, + "loss": 0.78544641, + "num_input_tokens_seen": 132016885, + "step": 6141, + "time_per_iteration": 4.700171232223511 + }, + { + "auxiliary_loss_clip": 0.01078126, + "auxiliary_loss_mlp": 0.01002737, + "balance_loss_clip": 1.0116744, + "balance_loss_mlp": 1.00154483, + "epoch": 0.7385318343052967, + "flos": 61494331662720.0, + "grad_norm": 0.7387857954343258, + "language_loss": 0.56455868, + "learning_rate": 6.751456498162981e-07, + "loss": 0.58536732, + "num_input_tokens_seen": 132075920, + "step": 6142, + "time_per_iteration": 3.0820260047912598 + }, + { + "auxiliary_loss_clip": 0.01176186, + "auxiliary_loss_mlp": 0.01023223, + "balance_loss_clip": 1.04384351, + "balance_loss_mlp": 1.01623106, + "epoch": 0.7386520771959358, + "flos": 17013757697280.0, + "grad_norm": 1.87867181071198, + "language_loss": 0.85355937, + "learning_rate": 6.745622021601174e-07, + "loss": 0.87555349, + "num_input_tokens_seen": 132092945, + "step": 6143, + "time_per_iteration": 2.6514956951141357 + }, + { + "auxiliary_loss_clip": 0.01197177, + "auxiliary_loss_mlp": 0.01023116, + "balance_loss_clip": 1.04370904, + "balance_loss_mlp": 1.01528358, + "epoch": 0.7387723200865749, + "flos": 18770759464320.0, + "grad_norm": 1.8401170733430272, + "language_loss": 0.69674492, + "learning_rate": 6.739789555717954e-07, + "loss": 0.71894783, + "num_input_tokens_seen": 132109920, + "step": 6144, + "time_per_iteration": 2.6768674850463867 + }, + { + "auxiliary_loss_clip": 0.01166883, + "auxiliary_loss_mlp": 0.01027316, + "balance_loss_clip": 1.04636812, + "balance_loss_mlp": 1.01961482, + "epoch": 0.738892562977214, + "flos": 22525542840960.0, + "grad_norm": 2.064914421683879, + "language_loss": 0.77322292, + "learning_rate": 6.733959101398124e-07, + "loss": 0.79516494, + "num_input_tokens_seen": 132128050, + "step": 6145, + "time_per_iteration": 3.4612956047058105 + }, + { + "auxiliary_loss_clip": 0.01183895, + "auxiliary_loss_mlp": 0.01029931, + "balance_loss_clip": 1.04363036, + "balance_loss_mlp": 1.02275455, + "epoch": 0.7390128058678531, + "flos": 21501478091520.0, + "grad_norm": 2.949898182054298, + "language_loss": 0.81570411, + "learning_rate": 6.728130659526143e-07, + "loss": 0.8378424, + "num_input_tokens_seen": 132145860, + "step": 6146, + "time_per_iteration": 2.6530067920684814 + }, + { + "auxiliary_loss_clip": 0.01190569, + "auxiliary_loss_mlp": 0.01028799, + "balance_loss_clip": 1.04711962, + "balance_loss_mlp": 1.02116919, + "epoch": 0.7391330487584922, + "flos": 25776176878080.0, + "grad_norm": 2.4961514101234274, + "language_loss": 0.70915294, + "learning_rate": 6.7223042309862e-07, + "loss": 0.73134655, + "num_input_tokens_seen": 132166060, + "step": 6147, + "time_per_iteration": 3.5781755447387695 + }, + { + "auxiliary_loss_clip": 0.01174987, + "auxiliary_loss_mlp": 0.01021783, + "balance_loss_clip": 1.04528952, + "balance_loss_mlp": 1.01440716, + "epoch": 0.7392532916491312, + "flos": 28366736636160.0, + "grad_norm": 1.9211100520051518, + "language_loss": 0.73386604, + "learning_rate": 6.716479816662144e-07, + "loss": 0.75583375, + "num_input_tokens_seen": 132187790, + "step": 6148, + "time_per_iteration": 2.722088098526001 + }, + { + "auxiliary_loss_clip": 0.01189067, + "auxiliary_loss_mlp": 0.01025359, + "balance_loss_clip": 1.04318988, + "balance_loss_mlp": 1.01810467, + "epoch": 0.7393735345397703, + "flos": 23585877348480.0, + "grad_norm": 1.874088961667303, + "language_loss": 0.72990251, + "learning_rate": 6.710657417437531e-07, + "loss": 0.75204682, + "num_input_tokens_seen": 132207495, + "step": 6149, + "time_per_iteration": 2.6610074043273926 + }, + { + "auxiliary_loss_clip": 0.01186671, + "auxiliary_loss_mlp": 0.01027067, + "balance_loss_clip": 1.04545724, + "balance_loss_mlp": 1.01941395, + "epoch": 0.7394937774304094, + "flos": 19974772373760.0, + "grad_norm": 2.062689634023383, + "language_loss": 0.80269349, + "learning_rate": 6.704837034195628e-07, + "loss": 0.82483089, + "num_input_tokens_seen": 132225960, + "step": 6150, + "time_per_iteration": 2.5997025966644287 + }, + { + "auxiliary_loss_clip": 0.01171072, + "auxiliary_loss_mlp": 0.01030539, + "balance_loss_clip": 1.04718149, + "balance_loss_mlp": 1.0226922, + "epoch": 0.7396140203210485, + "flos": 23478037741440.0, + "grad_norm": 1.7151236327256325, + "language_loss": 0.84821737, + "learning_rate": 6.699018667819376e-07, + "loss": 0.87023348, + "num_input_tokens_seen": 132245360, + "step": 6151, + "time_per_iteration": 2.62148118019104 + }, + { + "auxiliary_loss_clip": 0.01173683, + "auxiliary_loss_mlp": 0.0102667, + "balance_loss_clip": 1.04454637, + "balance_loss_mlp": 1.01916564, + "epoch": 0.7397342632116876, + "flos": 25555433846400.0, + "grad_norm": 1.602538365240831, + "language_loss": 0.72608155, + "learning_rate": 6.693202319191415e-07, + "loss": 0.74808514, + "num_input_tokens_seen": 132267095, + "step": 6152, + "time_per_iteration": 2.6851608753204346 + }, + { + "auxiliary_loss_clip": 0.01170434, + "auxiliary_loss_mlp": 0.01025765, + "balance_loss_clip": 1.05103087, + "balance_loss_mlp": 1.0187794, + "epoch": 0.7398545061023267, + "flos": 24755021130240.0, + "grad_norm": 1.8877583717851492, + "language_loss": 0.74796736, + "learning_rate": 6.687387989194084e-07, + "loss": 0.76992935, + "num_input_tokens_seen": 132286610, + "step": 6153, + "time_per_iteration": 2.595264434814453 + }, + { + "auxiliary_loss_clip": 0.0118222, + "auxiliary_loss_mlp": 0.01023956, + "balance_loss_clip": 1.0449481, + "balance_loss_mlp": 1.0162257, + "epoch": 0.7399747489929658, + "flos": 16508602776960.0, + "grad_norm": 1.829665900774435, + "language_loss": 0.79143, + "learning_rate": 6.681575678709404e-07, + "loss": 0.8134917, + "num_input_tokens_seen": 132305300, + "step": 6154, + "time_per_iteration": 2.6055219173431396 + }, + { + "auxiliary_loss_clip": 0.01175026, + "auxiliary_loss_mlp": 0.01023265, + "balance_loss_clip": 1.04518437, + "balance_loss_mlp": 1.01568937, + "epoch": 0.7400949918836048, + "flos": 24097065753600.0, + "grad_norm": 1.970866712913524, + "language_loss": 0.70910883, + "learning_rate": 6.67576538861911e-07, + "loss": 0.7310918, + "num_input_tokens_seen": 132323875, + "step": 6155, + "time_per_iteration": 2.652096748352051 + }, + { + "auxiliary_loss_clip": 0.01183759, + "auxiliary_loss_mlp": 0.01024465, + "balance_loss_clip": 1.04570901, + "balance_loss_mlp": 1.01730013, + "epoch": 0.740215234774244, + "flos": 21802517976960.0, + "grad_norm": 1.499581191511844, + "language_loss": 0.82196927, + "learning_rate": 6.669957119804612e-07, + "loss": 0.84405148, + "num_input_tokens_seen": 132345510, + "step": 6156, + "time_per_iteration": 2.661189556121826 + }, + { + "auxiliary_loss_clip": 0.01193348, + "auxiliary_loss_mlp": 0.01023351, + "balance_loss_clip": 1.04396701, + "balance_loss_mlp": 1.01585913, + "epoch": 0.7403354776648831, + "flos": 18733196816640.0, + "grad_norm": 4.033228400306593, + "language_loss": 0.72091651, + "learning_rate": 6.66415087314702e-07, + "loss": 0.74308348, + "num_input_tokens_seen": 132360465, + "step": 6157, + "time_per_iteration": 2.683716058731079 + }, + { + "auxiliary_loss_clip": 0.01187661, + "auxiliary_loss_mlp": 0.01025901, + "balance_loss_clip": 1.04496729, + "balance_loss_mlp": 1.01826525, + "epoch": 0.7404557205555221, + "flos": 16909581277440.0, + "grad_norm": 2.5588395679971874, + "language_loss": 0.72980034, + "learning_rate": 6.65834664952714e-07, + "loss": 0.75193596, + "num_input_tokens_seen": 132377915, + "step": 6158, + "time_per_iteration": 2.7035129070281982 + }, + { + "auxiliary_loss_clip": 0.01195987, + "auxiliary_loss_mlp": 0.01023289, + "balance_loss_clip": 1.04232383, + "balance_loss_mlp": 1.01585591, + "epoch": 0.7405759634461613, + "flos": 21214408596480.0, + "grad_norm": 1.706639631799951, + "language_loss": 0.75983518, + "learning_rate": 6.652544449825457e-07, + "loss": 0.78202796, + "num_input_tokens_seen": 132398170, + "step": 6159, + "time_per_iteration": 2.7123212814331055 + }, + { + "auxiliary_loss_clip": 0.01194457, + "auxiliary_loss_mlp": 0.01029248, + "balance_loss_clip": 1.04532743, + "balance_loss_mlp": 1.0212667, + "epoch": 0.7406962063368003, + "flos": 20480106862080.0, + "grad_norm": 1.8260225785803408, + "language_loss": 0.764902, + "learning_rate": 6.646744274922182e-07, + "loss": 0.78713906, + "num_input_tokens_seen": 132416615, + "step": 6160, + "time_per_iteration": 2.6648082733154297 + }, + { + "auxiliary_loss_clip": 0.01187741, + "auxiliary_loss_mlp": 0.01025725, + "balance_loss_clip": 1.04411697, + "balance_loss_mlp": 1.01839352, + "epoch": 0.7408164492274394, + "flos": 19791915212160.0, + "grad_norm": 4.178106789435028, + "language_loss": 0.75330406, + "learning_rate": 6.640946125697171e-07, + "loss": 0.77543867, + "num_input_tokens_seen": 132434145, + "step": 6161, + "time_per_iteration": 2.6535356044769287 + }, + { + "auxiliary_loss_clip": 0.01177168, + "auxiliary_loss_mlp": 0.01029326, + "balance_loss_clip": 1.0445199, + "balance_loss_mlp": 1.02065396, + "epoch": 0.7409366921180786, + "flos": 29204855654400.0, + "grad_norm": 5.773454731563942, + "language_loss": 0.7540592, + "learning_rate": 6.635150003030017e-07, + "loss": 0.77612418, + "num_input_tokens_seen": 132452670, + "step": 6162, + "time_per_iteration": 2.640589952468872 + }, + { + "auxiliary_loss_clip": 0.01201298, + "auxiliary_loss_mlp": 0.01022266, + "balance_loss_clip": 1.03904045, + "balance_loss_mlp": 1.01473832, + "epoch": 0.7410569350087176, + "flos": 22930004960640.0, + "grad_norm": 2.7992801385932884, + "language_loss": 0.85834885, + "learning_rate": 6.629355907799981e-07, + "loss": 0.88058454, + "num_input_tokens_seen": 132472475, + "step": 6163, + "time_per_iteration": 2.726942777633667 + }, + { + "auxiliary_loss_clip": 0.01182461, + "auxiliary_loss_mlp": 0.0103334, + "balance_loss_clip": 1.04698563, + "balance_loss_mlp": 1.02573466, + "epoch": 0.7411771778993567, + "flos": 30440397726720.0, + "grad_norm": 1.8410140280130625, + "language_loss": 0.69107115, + "learning_rate": 6.623563840886015e-07, + "loss": 0.71322918, + "num_input_tokens_seen": 132493400, + "step": 6164, + "time_per_iteration": 2.689176559448242 + }, + { + "auxiliary_loss_clip": 0.01170709, + "auxiliary_loss_mlp": 0.01025603, + "balance_loss_clip": 1.04325223, + "balance_loss_mlp": 1.01828957, + "epoch": 0.7412974207899958, + "flos": 20522050968960.0, + "grad_norm": 1.650911449967234, + "language_loss": 0.69865739, + "learning_rate": 6.617773803166795e-07, + "loss": 0.72062051, + "num_input_tokens_seen": 132511725, + "step": 6165, + "time_per_iteration": 2.6204442977905273 + }, + { + "auxiliary_loss_clip": 0.01190533, + "auxiliary_loss_mlp": 0.0137907, + "balance_loss_clip": 1.04417384, + "balance_loss_mlp": 1.00024056, + "epoch": 0.7414176636806349, + "flos": 22090700793600.0, + "grad_norm": 2.233380767857893, + "language_loss": 0.81712413, + "learning_rate": 6.611985795520634e-07, + "loss": 0.84282017, + "num_input_tokens_seen": 132530270, + "step": 6166, + "time_per_iteration": 2.6187026500701904 + }, + { + "auxiliary_loss_clip": 0.01203212, + "auxiliary_loss_mlp": 0.01030145, + "balance_loss_clip": 1.04462719, + "balance_loss_mlp": 1.02223802, + "epoch": 0.7415379065712739, + "flos": 25155245445120.0, + "grad_norm": 2.5879560627261604, + "language_loss": 0.77941287, + "learning_rate": 6.606199818825588e-07, + "loss": 0.80174637, + "num_input_tokens_seen": 132550725, + "step": 6167, + "time_per_iteration": 3.6541895866394043 + }, + { + "auxiliary_loss_clip": 0.01189632, + "auxiliary_loss_mlp": 0.01024137, + "balance_loss_clip": 1.04377913, + "balance_loss_mlp": 1.01660252, + "epoch": 0.7416581494619131, + "flos": 16871731320960.0, + "grad_norm": 2.2933354312801493, + "language_loss": 0.81869078, + "learning_rate": 6.600415873959377e-07, + "loss": 0.84082842, + "num_input_tokens_seen": 132568600, + "step": 6168, + "time_per_iteration": 3.5328972339630127 + }, + { + "auxiliary_loss_clip": 0.01207932, + "auxiliary_loss_mlp": 0.01378289, + "balance_loss_clip": 1.03706908, + "balance_loss_mlp": 1.00018656, + "epoch": 0.7417783923525522, + "flos": 28438881102720.0, + "grad_norm": 2.0914928951436287, + "language_loss": 0.64745277, + "learning_rate": 6.594633961799437e-07, + "loss": 0.67331505, + "num_input_tokens_seen": 132587640, + "step": 6169, + "time_per_iteration": 2.902355432510376 + }, + { + "auxiliary_loss_clip": 0.01205016, + "auxiliary_loss_mlp": 0.01024708, + "balance_loss_clip": 1.04432631, + "balance_loss_mlp": 1.01746869, + "epoch": 0.7418986352431912, + "flos": 20084299920000.0, + "grad_norm": 1.5814476201633227, + "language_loss": 0.81607366, + "learning_rate": 6.588854083222857e-07, + "loss": 0.83837092, + "num_input_tokens_seen": 132607075, + "step": 6170, + "time_per_iteration": 2.7229225635528564 + }, + { + "auxiliary_loss_clip": 0.01188284, + "auxiliary_loss_mlp": 0.01028599, + "balance_loss_clip": 1.04468513, + "balance_loss_mlp": 1.02089179, + "epoch": 0.7420188781338304, + "flos": 18259571059200.0, + "grad_norm": 2.126174572429566, + "language_loss": 0.80744946, + "learning_rate": 6.583076239106444e-07, + "loss": 0.82961822, + "num_input_tokens_seen": 132625580, + "step": 6171, + "time_per_iteration": 2.691995859146118 + }, + { + "auxiliary_loss_clip": 0.01191968, + "auxiliary_loss_mlp": 0.01023506, + "balance_loss_clip": 1.04429626, + "balance_loss_mlp": 1.0162282, + "epoch": 0.7421391210244694, + "flos": 13771994319360.0, + "grad_norm": 2.160266283587928, + "language_loss": 0.75244826, + "learning_rate": 6.577300430326707e-07, + "loss": 0.77460301, + "num_input_tokens_seen": 132640525, + "step": 6172, + "time_per_iteration": 3.5149307250976562 + }, + { + "auxiliary_loss_clip": 0.01194588, + "auxiliary_loss_mlp": 0.01025195, + "balance_loss_clip": 1.04444385, + "balance_loss_mlp": 1.01802444, + "epoch": 0.7422593639151085, + "flos": 15961683317760.0, + "grad_norm": 2.7111593204966473, + "language_loss": 0.72034943, + "learning_rate": 6.571526657759821e-07, + "loss": 0.74254727, + "num_input_tokens_seen": 132656265, + "step": 6173, + "time_per_iteration": 3.5444655418395996 + }, + { + "auxiliary_loss_clip": 0.01170791, + "auxiliary_loss_mlp": 0.0102592, + "balance_loss_clip": 1.0430913, + "balance_loss_mlp": 1.01869619, + "epoch": 0.7423796068057477, + "flos": 30114400867200.0, + "grad_norm": 1.570947779499198, + "language_loss": 0.70553386, + "learning_rate": 6.565754922281663e-07, + "loss": 0.72750098, + "num_input_tokens_seen": 132678510, + "step": 6174, + "time_per_iteration": 2.6801960468292236 + }, + { + "auxiliary_loss_clip": 0.01181977, + "auxiliary_loss_mlp": 0.01024377, + "balance_loss_clip": 1.04172301, + "balance_loss_mlp": 1.01693511, + "epoch": 0.7424998496963867, + "flos": 20521907314560.0, + "grad_norm": 1.647773944265166, + "language_loss": 0.78337193, + "learning_rate": 6.559985224767801e-07, + "loss": 0.80543554, + "num_input_tokens_seen": 132696385, + "step": 6175, + "time_per_iteration": 2.629326105117798 + }, + { + "auxiliary_loss_clip": 0.01200842, + "auxiliary_loss_mlp": 0.01025283, + "balance_loss_clip": 1.04383051, + "balance_loss_mlp": 1.01762938, + "epoch": 0.7426200925870258, + "flos": 21871573873920.0, + "grad_norm": 2.484453847790887, + "language_loss": 0.75406235, + "learning_rate": 6.55421756609349e-07, + "loss": 0.77632362, + "num_input_tokens_seen": 132714640, + "step": 6176, + "time_per_iteration": 2.64723539352417 + }, + { + "auxiliary_loss_clip": 0.01177204, + "auxiliary_loss_mlp": 0.01029287, + "balance_loss_clip": 1.04914045, + "balance_loss_mlp": 1.02170539, + "epoch": 0.7427403354776649, + "flos": 26432049265920.0, + "grad_norm": 1.969611265315953, + "language_loss": 0.78689241, + "learning_rate": 6.54845194713369e-07, + "loss": 0.80895734, + "num_input_tokens_seen": 132735590, + "step": 6177, + "time_per_iteration": 2.655153512954712 + }, + { + "auxiliary_loss_clip": 0.01173015, + "auxiliary_loss_mlp": 0.01024243, + "balance_loss_clip": 1.04694343, + "balance_loss_mlp": 1.01679265, + "epoch": 0.742860578368304, + "flos": 19898390102400.0, + "grad_norm": 1.9498707498291643, + "language_loss": 0.79742545, + "learning_rate": 6.542688368763034e-07, + "loss": 0.81939805, + "num_input_tokens_seen": 132753995, + "step": 6178, + "time_per_iteration": 2.6299891471862793 + }, + { + "auxiliary_loss_clip": 0.01176281, + "auxiliary_loss_mlp": 0.01024409, + "balance_loss_clip": 1.04684961, + "balance_loss_mlp": 1.01691031, + "epoch": 0.742980821258943, + "flos": 24827201510400.0, + "grad_norm": 2.2748485010094184, + "language_loss": 0.77007699, + "learning_rate": 6.536926831855854e-07, + "loss": 0.79208386, + "num_input_tokens_seen": 132773160, + "step": 6179, + "time_per_iteration": 2.6133766174316406 + }, + { + "auxiliary_loss_clip": 0.01181965, + "auxiliary_loss_mlp": 0.01025367, + "balance_loss_clip": 1.04437232, + "balance_loss_mlp": 1.01767778, + "epoch": 0.7431010641495821, + "flos": 25228646887680.0, + "grad_norm": 2.5582426573159087, + "language_loss": 0.73119491, + "learning_rate": 6.531167337286165e-07, + "loss": 0.75326824, + "num_input_tokens_seen": 132793180, + "step": 6180, + "time_per_iteration": 2.649750232696533 + }, + { + "auxiliary_loss_clip": 0.01187095, + "auxiliary_loss_mlp": 0.0102535, + "balance_loss_clip": 1.04750538, + "balance_loss_mlp": 1.01832581, + "epoch": 0.7432213070402213, + "flos": 21762369550080.0, + "grad_norm": 1.6944339333057314, + "language_loss": 0.79501683, + "learning_rate": 6.52540988592768e-07, + "loss": 0.81714129, + "num_input_tokens_seen": 132814200, + "step": 6181, + "time_per_iteration": 2.647848129272461 + }, + { + "auxiliary_loss_clip": 0.01188757, + "auxiliary_loss_mlp": 0.01030388, + "balance_loss_clip": 1.04455447, + "balance_loss_mlp": 1.02302659, + "epoch": 0.7433415499308603, + "flos": 14793832425600.0, + "grad_norm": 2.356901687241828, + "language_loss": 0.83567041, + "learning_rate": 6.519654478653814e-07, + "loss": 0.85786188, + "num_input_tokens_seen": 132832565, + "step": 6182, + "time_per_iteration": 2.628309488296509 + }, + { + "auxiliary_loss_clip": 0.01091646, + "auxiliary_loss_mlp": 0.01000623, + "balance_loss_clip": 1.0115459, + "balance_loss_mlp": 0.9995147, + "epoch": 0.7434617928214994, + "flos": 67155577297920.0, + "grad_norm": 0.7608410363291875, + "language_loss": 0.56071526, + "learning_rate": 6.51390111633763e-07, + "loss": 0.58163798, + "num_input_tokens_seen": 132897840, + "step": 6183, + "time_per_iteration": 3.327423572540283 + }, + { + "auxiliary_loss_clip": 0.01208625, + "auxiliary_loss_mlp": 0.01025296, + "balance_loss_clip": 1.03835261, + "balance_loss_mlp": 1.01840234, + "epoch": 0.7435820357121385, + "flos": 27377576928000.0, + "grad_norm": 1.65096211283241, + "language_loss": 0.76136029, + "learning_rate": 6.508149799851932e-07, + "loss": 0.78369951, + "num_input_tokens_seen": 132919505, + "step": 6184, + "time_per_iteration": 2.832552909851074 + }, + { + "auxiliary_loss_clip": 0.01184342, + "auxiliary_loss_mlp": 0.01021618, + "balance_loss_clip": 1.04616261, + "balance_loss_mlp": 1.01452541, + "epoch": 0.7437022786027776, + "flos": 23987645948160.0, + "grad_norm": 3.040917662984888, + "language_loss": 0.61569273, + "learning_rate": 6.502400530069183e-07, + "loss": 0.63775235, + "num_input_tokens_seen": 132939390, + "step": 6185, + "time_per_iteration": 2.66658353805542 + }, + { + "auxiliary_loss_clip": 0.01193503, + "auxiliary_loss_mlp": 0.01024714, + "balance_loss_clip": 1.04319525, + "balance_loss_mlp": 1.01652992, + "epoch": 0.7438225214934167, + "flos": 21866761451520.0, + "grad_norm": 2.1047530887902517, + "language_loss": 0.68631452, + "learning_rate": 6.496653307861535e-07, + "loss": 0.70849669, + "num_input_tokens_seen": 132960060, + "step": 6186, + "time_per_iteration": 2.71455717086792 + }, + { + "auxiliary_loss_clip": 0.0118132, + "auxiliary_loss_mlp": 0.01024023, + "balance_loss_clip": 1.04512942, + "balance_loss_mlp": 1.01733208, + "epoch": 0.7439427643840558, + "flos": 20230097224320.0, + "grad_norm": 1.7940850845042287, + "language_loss": 0.65757817, + "learning_rate": 6.490908134100857e-07, + "loss": 0.67963159, + "num_input_tokens_seen": 132978525, + "step": 6187, + "time_per_iteration": 2.5878937244415283 + }, + { + "auxiliary_loss_clip": 0.01180755, + "auxiliary_loss_mlp": 0.01024676, + "balance_loss_clip": 1.04540181, + "balance_loss_mlp": 1.01720214, + "epoch": 0.7440630072746949, + "flos": 20849915335680.0, + "grad_norm": 2.1351244606837048, + "language_loss": 0.69437337, + "learning_rate": 6.48516500965866e-07, + "loss": 0.71642774, + "num_input_tokens_seen": 132998460, + "step": 6188, + "time_per_iteration": 2.6305034160614014 + }, + { + "auxiliary_loss_clip": 0.01180016, + "auxiliary_loss_mlp": 0.01019783, + "balance_loss_clip": 1.04450727, + "balance_loss_mlp": 1.01292276, + "epoch": 0.7441832501653339, + "flos": 26503762769280.0, + "grad_norm": 1.8088832273520759, + "language_loss": 0.81850958, + "learning_rate": 6.479423935406192e-07, + "loss": 0.84050751, + "num_input_tokens_seen": 133018445, + "step": 6189, + "time_per_iteration": 2.6737215518951416 + }, + { + "auxiliary_loss_clip": 0.01086181, + "auxiliary_loss_mlp": 0.01002598, + "balance_loss_clip": 1.0144136, + "balance_loss_mlp": 1.00142431, + "epoch": 0.7443034930559731, + "flos": 68602848088320.0, + "grad_norm": 0.8041047449685225, + "language_loss": 0.61994624, + "learning_rate": 6.473684912214357e-07, + "loss": 0.64083397, + "num_input_tokens_seen": 133082005, + "step": 6190, + "time_per_iteration": 3.3278884887695312 + }, + { + "auxiliary_loss_clip": 0.0117697, + "auxiliary_loss_mlp": 0.01026228, + "balance_loss_clip": 1.0459373, + "balance_loss_mlp": 1.01841938, + "epoch": 0.7444237359466122, + "flos": 18654982951680.0, + "grad_norm": 2.4523603442550668, + "language_loss": 0.7000224, + "learning_rate": 6.467947940953778e-07, + "loss": 0.72205436, + "num_input_tokens_seen": 133100530, + "step": 6191, + "time_per_iteration": 2.6663036346435547 + }, + { + "auxiliary_loss_clip": 0.01185218, + "auxiliary_loss_mlp": 0.01019187, + "balance_loss_clip": 1.0447073, + "balance_loss_mlp": 1.01187015, + "epoch": 0.7445439788372512, + "flos": 22817604326400.0, + "grad_norm": 1.786774384089426, + "language_loss": 0.72649431, + "learning_rate": 6.462213022494732e-07, + "loss": 0.74853837, + "num_input_tokens_seen": 133119775, + "step": 6192, + "time_per_iteration": 2.705139636993408 + }, + { + "auxiliary_loss_clip": 0.01077928, + "auxiliary_loss_mlp": 0.01002344, + "balance_loss_clip": 1.01151276, + "balance_loss_mlp": 1.00124764, + "epoch": 0.7446642217278904, + "flos": 67045690615680.0, + "grad_norm": 0.7712649500801977, + "language_loss": 0.61001194, + "learning_rate": 6.456480157707201e-07, + "loss": 0.63081461, + "num_input_tokens_seen": 133184550, + "step": 6193, + "time_per_iteration": 4.0833775997161865 + }, + { + "auxiliary_loss_clip": 0.01187811, + "auxiliary_loss_mlp": 0.01024524, + "balance_loss_clip": 1.04272413, + "balance_loss_mlp": 1.01704109, + "epoch": 0.7447844646185294, + "flos": 17417465631360.0, + "grad_norm": 2.0898281807185857, + "language_loss": 0.85146976, + "learning_rate": 6.450749347460866e-07, + "loss": 0.87359309, + "num_input_tokens_seen": 133201525, + "step": 6194, + "time_per_iteration": 2.755890369415283 + }, + { + "auxiliary_loss_clip": 0.01170159, + "auxiliary_loss_mlp": 0.01028104, + "balance_loss_clip": 1.04799068, + "balance_loss_mlp": 1.01990247, + "epoch": 0.7449047075091685, + "flos": 26615876094720.0, + "grad_norm": 1.8877395706464308, + "language_loss": 0.79121125, + "learning_rate": 6.445020592625083e-07, + "loss": 0.81319392, + "num_input_tokens_seen": 133222175, + "step": 6195, + "time_per_iteration": 2.5985701084136963 + }, + { + "auxiliary_loss_clip": 0.0116835, + "auxiliary_loss_mlp": 0.0102626, + "balance_loss_clip": 1.04651189, + "balance_loss_mlp": 1.01925898, + "epoch": 0.7450249503998077, + "flos": 14170458867840.0, + "grad_norm": 2.0474155677975845, + "language_loss": 0.80282664, + "learning_rate": 6.4392938940689e-07, + "loss": 0.82477272, + "num_input_tokens_seen": 133237590, + "step": 6196, + "time_per_iteration": 2.588310956954956 + }, + { + "auxiliary_loss_clip": 0.0119905, + "auxiliary_loss_mlp": 0.01379081, + "balance_loss_clip": 1.04099107, + "balance_loss_mlp": 1.0001694, + "epoch": 0.7451451932904467, + "flos": 19606687752960.0, + "grad_norm": 2.4834753612592158, + "language_loss": 0.71276939, + "learning_rate": 6.433569252661049e-07, + "loss": 0.73855066, + "num_input_tokens_seen": 133255590, + "step": 6197, + "time_per_iteration": 2.709059238433838 + }, + { + "auxiliary_loss_clip": 0.01188985, + "auxiliary_loss_mlp": 0.01022474, + "balance_loss_clip": 1.04308534, + "balance_loss_mlp": 1.01565814, + "epoch": 0.7452654361810858, + "flos": 12495405980160.0, + "grad_norm": 1.9036964829317669, + "language_loss": 0.71071726, + "learning_rate": 6.427846669269952e-07, + "loss": 0.73283184, + "num_input_tokens_seen": 133273210, + "step": 6198, + "time_per_iteration": 3.5734400749206543 + }, + { + "auxiliary_loss_clip": 0.01169132, + "auxiliary_loss_mlp": 0.01024866, + "balance_loss_clip": 1.04836106, + "balance_loss_mlp": 1.01793957, + "epoch": 0.7453856790717249, + "flos": 22127329687680.0, + "grad_norm": 2.0615265499830233, + "language_loss": 0.82403773, + "learning_rate": 6.422126144763729e-07, + "loss": 0.84597772, + "num_input_tokens_seen": 133292600, + "step": 6199, + "time_per_iteration": 3.473098039627075 + }, + { + "auxiliary_loss_clip": 0.01188751, + "auxiliary_loss_mlp": 0.01379147, + "balance_loss_clip": 1.03948092, + "balance_loss_mlp": 1.00010204, + "epoch": 0.745505921962364, + "flos": 20010682995840.0, + "grad_norm": 1.9974194948661208, + "language_loss": 0.7663638, + "learning_rate": 6.416407680010174e-07, + "loss": 0.79204273, + "num_input_tokens_seen": 133306960, + "step": 6200, + "time_per_iteration": 2.656806707382202 + }, + { + "auxiliary_loss_clip": 0.01216905, + "auxiliary_loss_mlp": 0.01028382, + "balance_loss_clip": 1.04289699, + "balance_loss_mlp": 1.0208714, + "epoch": 0.745626164853003, + "flos": 24677884673280.0, + "grad_norm": 2.3883749717624094, + "language_loss": 0.81127369, + "learning_rate": 6.410691275876774e-07, + "loss": 0.83372653, + "num_input_tokens_seen": 133326380, + "step": 6201, + "time_per_iteration": 2.77439284324646 + }, + { + "auxiliary_loss_clip": 0.011957, + "auxiliary_loss_mlp": 0.01026806, + "balance_loss_clip": 1.04682851, + "balance_loss_mlp": 1.01915836, + "epoch": 0.7457464077436422, + "flos": 14538830797440.0, + "grad_norm": 5.824991475396436, + "language_loss": 0.76620722, + "learning_rate": 6.404976933230704e-07, + "loss": 0.78843224, + "num_input_tokens_seen": 133342900, + "step": 6202, + "time_per_iteration": 2.5898025035858154 + }, + { + "auxiliary_loss_clip": 0.01189988, + "auxiliary_loss_mlp": 0.01030588, + "balance_loss_clip": 1.0453527, + "balance_loss_mlp": 1.02246344, + "epoch": 0.7458666506342813, + "flos": 34021194600960.0, + "grad_norm": 1.8690151841216516, + "language_loss": 0.72605467, + "learning_rate": 6.399264652938813e-07, + "loss": 0.74826038, + "num_input_tokens_seen": 133363805, + "step": 6203, + "time_per_iteration": 2.753545045852661 + }, + { + "auxiliary_loss_clip": 0.01183095, + "auxiliary_loss_mlp": 0.01025231, + "balance_loss_clip": 1.04299736, + "balance_loss_mlp": 1.01886535, + "epoch": 0.7459868935249203, + "flos": 24279025075200.0, + "grad_norm": 2.23785200829016, + "language_loss": 0.74455416, + "learning_rate": 6.393554435867679e-07, + "loss": 0.76663744, + "num_input_tokens_seen": 133384655, + "step": 6204, + "time_per_iteration": 2.6878371238708496 + }, + { + "auxiliary_loss_clip": 0.01191554, + "auxiliary_loss_mlp": 0.01031484, + "balance_loss_clip": 1.04373193, + "balance_loss_mlp": 1.02385414, + "epoch": 0.7461071364155595, + "flos": 21908777385600.0, + "grad_norm": 2.1788100717236025, + "language_loss": 0.83685839, + "learning_rate": 6.387846282883502e-07, + "loss": 0.85908878, + "num_input_tokens_seen": 133401185, + "step": 6205, + "time_per_iteration": 2.7037577629089355 + }, + { + "auxiliary_loss_clip": 0.0116676, + "auxiliary_loss_mlp": 0.01028531, + "balance_loss_clip": 1.04764163, + "balance_loss_mlp": 1.0209552, + "epoch": 0.7462273793061985, + "flos": 22889712879360.0, + "grad_norm": 2.0151928245513466, + "language_loss": 0.76597315, + "learning_rate": 6.38214019485223e-07, + "loss": 0.78792608, + "num_input_tokens_seen": 133420010, + "step": 6206, + "time_per_iteration": 2.5499253273010254 + }, + { + "auxiliary_loss_clip": 0.01205795, + "auxiliary_loss_mlp": 0.01024159, + "balance_loss_clip": 1.03823042, + "balance_loss_mlp": 1.01656568, + "epoch": 0.7463476221968376, + "flos": 19968451580160.0, + "grad_norm": 1.8251701516803898, + "language_loss": 0.71664751, + "learning_rate": 6.376436172639461e-07, + "loss": 0.73894703, + "num_input_tokens_seen": 133437855, + "step": 6207, + "time_per_iteration": 2.7554831504821777 + }, + { + "auxiliary_loss_clip": 0.01222106, + "auxiliary_loss_mlp": 0.01028474, + "balance_loss_clip": 1.03887987, + "balance_loss_mlp": 1.02040339, + "epoch": 0.7464678650874768, + "flos": 16836610798080.0, + "grad_norm": 2.5740279163196664, + "language_loss": 0.6510489, + "learning_rate": 6.370734217110487e-07, + "loss": 0.67355472, + "num_input_tokens_seen": 133456600, + "step": 6208, + "time_per_iteration": 2.725243091583252 + }, + { + "auxiliary_loss_clip": 0.01191785, + "auxiliary_loss_mlp": 0.01028834, + "balance_loss_clip": 1.04808307, + "balance_loss_mlp": 1.02091289, + "epoch": 0.7465881079781158, + "flos": 48100869843840.0, + "grad_norm": 1.4809068778938526, + "language_loss": 0.64260817, + "learning_rate": 6.36503432913031e-07, + "loss": 0.66481429, + "num_input_tokens_seen": 133479745, + "step": 6209, + "time_per_iteration": 2.828695297241211 + }, + { + "auxiliary_loss_clip": 0.01176787, + "auxiliary_loss_mlp": 0.01025799, + "balance_loss_clip": 1.04688859, + "balance_loss_mlp": 1.01791883, + "epoch": 0.7467083508687549, + "flos": 19677359761920.0, + "grad_norm": 1.9960357412888328, + "language_loss": 0.6911546, + "learning_rate": 6.359336509563569e-07, + "loss": 0.71318048, + "num_input_tokens_seen": 133495765, + "step": 6210, + "time_per_iteration": 2.620892286300659 + }, + { + "auxiliary_loss_clip": 0.01179737, + "auxiliary_loss_mlp": 0.01025953, + "balance_loss_clip": 1.04149985, + "balance_loss_mlp": 1.01843655, + "epoch": 0.7468285937593939, + "flos": 17895436934400.0, + "grad_norm": 1.7562920064722731, + "language_loss": 0.80715781, + "learning_rate": 6.353640759274641e-07, + "loss": 0.82921469, + "num_input_tokens_seen": 133514655, + "step": 6211, + "time_per_iteration": 2.6197941303253174 + }, + { + "auxiliary_loss_clip": 0.01174251, + "auxiliary_loss_mlp": 0.01024189, + "balance_loss_clip": 1.04370177, + "balance_loss_mlp": 1.01659262, + "epoch": 0.7469488366500331, + "flos": 23141446369920.0, + "grad_norm": 10.853423829373218, + "language_loss": 0.74976218, + "learning_rate": 6.347947079127556e-07, + "loss": 0.77174658, + "num_input_tokens_seen": 133532555, + "step": 6212, + "time_per_iteration": 2.680996894836426 + }, + { + "auxiliary_loss_clip": 0.01180525, + "auxiliary_loss_mlp": 0.01028528, + "balance_loss_clip": 1.04358697, + "balance_loss_mlp": 1.02141738, + "epoch": 0.7470690795406721, + "flos": 16690849407360.0, + "grad_norm": 2.0643062052622, + "language_loss": 0.77139062, + "learning_rate": 6.342255469986053e-07, + "loss": 0.79348111, + "num_input_tokens_seen": 133551300, + "step": 6213, + "time_per_iteration": 2.617138624191284 + }, + { + "auxiliary_loss_clip": 0.01167382, + "auxiliary_loss_mlp": 0.01022975, + "balance_loss_clip": 1.04658532, + "balance_loss_mlp": 1.01541102, + "epoch": 0.7471893224313112, + "flos": 25192700352000.0, + "grad_norm": 1.710052268138982, + "language_loss": 0.75969589, + "learning_rate": 6.336565932713533e-07, + "loss": 0.78159952, + "num_input_tokens_seen": 133570725, + "step": 6214, + "time_per_iteration": 2.604619264602661 + }, + { + "auxiliary_loss_clip": 0.01183352, + "auxiliary_loss_mlp": 0.01029371, + "balance_loss_clip": 1.04673862, + "balance_loss_mlp": 1.02174735, + "epoch": 0.7473095653219504, + "flos": 22526225199360.0, + "grad_norm": 2.0142939132440447, + "language_loss": 0.77841234, + "learning_rate": 6.330878468173088e-07, + "loss": 0.80053961, + "num_input_tokens_seen": 133590790, + "step": 6215, + "time_per_iteration": 2.6579537391662598 + }, + { + "auxiliary_loss_clip": 0.01166371, + "auxiliary_loss_mlp": 0.01025682, + "balance_loss_clip": 1.04348135, + "balance_loss_mlp": 1.01867211, + "epoch": 0.7474298082125894, + "flos": 18113989236480.0, + "grad_norm": 2.1086716953794427, + "language_loss": 0.72806633, + "learning_rate": 6.32519307722752e-07, + "loss": 0.74998683, + "num_input_tokens_seen": 133608685, + "step": 6216, + "time_per_iteration": 2.655616044998169 + }, + { + "auxiliary_loss_clip": 0.01103278, + "auxiliary_loss_mlp": 0.01000493, + "balance_loss_clip": 1.01696777, + "balance_loss_mlp": 0.99948609, + "epoch": 0.7475500511032285, + "flos": 62086535193600.0, + "grad_norm": 0.9162628749307754, + "language_loss": 0.54960811, + "learning_rate": 6.31950976073929e-07, + "loss": 0.57064575, + "num_input_tokens_seen": 133662775, + "step": 6217, + "time_per_iteration": 3.2585561275482178 + }, + { + "auxiliary_loss_clip": 0.01202062, + "auxiliary_loss_mlp": 0.01025562, + "balance_loss_clip": 1.04361498, + "balance_loss_mlp": 1.01828158, + "epoch": 0.7476702939938676, + "flos": 17785586165760.0, + "grad_norm": 2.2400988572125193, + "language_loss": 0.80605352, + "learning_rate": 6.31382851957055e-07, + "loss": 0.82832968, + "num_input_tokens_seen": 133679595, + "step": 6218, + "time_per_iteration": 2.6983981132507324 + }, + { + "auxiliary_loss_clip": 0.01193243, + "auxiliary_loss_mlp": 0.01378646, + "balance_loss_clip": 1.04556525, + "balance_loss_mlp": 1.0001905, + "epoch": 0.7477905368845067, + "flos": 27927944092800.0, + "grad_norm": 2.0157459742293007, + "language_loss": 0.71530282, + "learning_rate": 6.308149354583143e-07, + "loss": 0.74102175, + "num_input_tokens_seen": 133699000, + "step": 6219, + "time_per_iteration": 3.6076974868774414 + }, + { + "auxiliary_loss_clip": 0.01184617, + "auxiliary_loss_mlp": 0.01028308, + "balance_loss_clip": 1.04906881, + "balance_loss_mlp": 1.02087545, + "epoch": 0.7479107797751458, + "flos": 26870374932480.0, + "grad_norm": 2.0494489657577977, + "language_loss": 0.8189624, + "learning_rate": 6.302472266638586e-07, + "loss": 0.84109169, + "num_input_tokens_seen": 133719540, + "step": 6220, + "time_per_iteration": 3.540938377380371 + }, + { + "auxiliary_loss_clip": 0.01175846, + "auxiliary_loss_mlp": 0.01027881, + "balance_loss_clip": 1.05002928, + "balance_loss_mlp": 1.01981068, + "epoch": 0.7480310226657849, + "flos": 33943375785600.0, + "grad_norm": 1.9506329404467364, + "language_loss": 0.70031559, + "learning_rate": 6.296797256598101e-07, + "loss": 0.7223528, + "num_input_tokens_seen": 133741020, + "step": 6221, + "time_per_iteration": 2.6496152877807617 + }, + { + "auxiliary_loss_clip": 0.01184679, + "auxiliary_loss_mlp": 0.01024936, + "balance_loss_clip": 1.04240954, + "balance_loss_mlp": 1.0180279, + "epoch": 0.748151265556424, + "flos": 24826555065600.0, + "grad_norm": 1.88394500053882, + "language_loss": 0.81499642, + "learning_rate": 6.291124325322576e-07, + "loss": 0.83709252, + "num_input_tokens_seen": 133761145, + "step": 6222, + "time_per_iteration": 2.6706202030181885 + }, + { + "auxiliary_loss_clip": 0.01191305, + "auxiliary_loss_mlp": 0.01025317, + "balance_loss_clip": 1.04491711, + "balance_loss_mlp": 1.01797938, + "epoch": 0.748271508447063, + "flos": 38399351535360.0, + "grad_norm": 1.583066685728043, + "language_loss": 0.62514961, + "learning_rate": 6.285453473672595e-07, + "loss": 0.64731574, + "num_input_tokens_seen": 133783715, + "step": 6223, + "time_per_iteration": 2.7801225185394287 + }, + { + "auxiliary_loss_clip": 0.01169638, + "auxiliary_loss_mlp": 0.01025962, + "balance_loss_clip": 1.04849458, + "balance_loss_mlp": 1.01900315, + "epoch": 0.7483917513377022, + "flos": 21541842000000.0, + "grad_norm": 1.8750600988890518, + "language_loss": 0.75522053, + "learning_rate": 6.279784702508415e-07, + "loss": 0.7771765, + "num_input_tokens_seen": 133804465, + "step": 6224, + "time_per_iteration": 3.5286543369293213 + }, + { + "auxiliary_loss_clip": 0.01104433, + "auxiliary_loss_mlp": 0.01000771, + "balance_loss_clip": 1.01072526, + "balance_loss_mlp": 0.99966842, + "epoch": 0.7485119942283412, + "flos": 62314532772480.0, + "grad_norm": 0.7779945094712117, + "language_loss": 0.58554149, + "learning_rate": 6.274118012689979e-07, + "loss": 0.60659349, + "num_input_tokens_seen": 133866365, + "step": 6225, + "time_per_iteration": 4.267701148986816 + }, + { + "auxiliary_loss_clip": 0.01179076, + "auxiliary_loss_mlp": 0.01028162, + "balance_loss_clip": 1.04432964, + "balance_loss_mlp": 1.02029991, + "epoch": 0.7486322371189803, + "flos": 29937613104000.0, + "grad_norm": 1.7466060363131453, + "language_loss": 0.68078208, + "learning_rate": 6.268453405076943e-07, + "loss": 0.70285439, + "num_input_tokens_seen": 133888760, + "step": 6226, + "time_per_iteration": 2.699772357940674 + }, + { + "auxiliary_loss_clip": 0.01184751, + "auxiliary_loss_mlp": 0.01022249, + "balance_loss_clip": 1.04370618, + "balance_loss_mlp": 1.01531148, + "epoch": 0.7487524800096195, + "flos": 18949414734720.0, + "grad_norm": 2.0775409915629544, + "language_loss": 0.82248819, + "learning_rate": 6.262790880528592e-07, + "loss": 0.84455812, + "num_input_tokens_seen": 133906380, + "step": 6227, + "time_per_iteration": 2.7503726482391357 + }, + { + "auxiliary_loss_clip": 0.01201325, + "auxiliary_loss_mlp": 0.01026561, + "balance_loss_clip": 1.04007328, + "balance_loss_mlp": 1.01904535, + "epoch": 0.7488727229002585, + "flos": 18697393935360.0, + "grad_norm": 3.2879115249463866, + "language_loss": 0.79708946, + "learning_rate": 6.257130439903951e-07, + "loss": 0.8193683, + "num_input_tokens_seen": 133922875, + "step": 6228, + "time_per_iteration": 2.6344265937805176 + }, + { + "auxiliary_loss_clip": 0.011722, + "auxiliary_loss_mlp": 0.01027048, + "balance_loss_clip": 1.0496459, + "balance_loss_mlp": 1.02025342, + "epoch": 0.7489929657908976, + "flos": 23623368168960.0, + "grad_norm": 3.715557640163127, + "language_loss": 0.80854928, + "learning_rate": 6.251472084061695e-07, + "loss": 0.83054173, + "num_input_tokens_seen": 133941795, + "step": 6229, + "time_per_iteration": 2.5960640907287598 + }, + { + "auxiliary_loss_clip": 0.0117687, + "auxiliary_loss_mlp": 0.01030202, + "balance_loss_clip": 1.04792047, + "balance_loss_mlp": 1.02287102, + "epoch": 0.7491132086815367, + "flos": 20551533056640.0, + "grad_norm": 1.9021854847039759, + "language_loss": 0.89257109, + "learning_rate": 6.245815813860191e-07, + "loss": 0.9146418, + "num_input_tokens_seen": 133957305, + "step": 6230, + "time_per_iteration": 2.5719356536865234 + }, + { + "auxiliary_loss_clip": 0.01171642, + "auxiliary_loss_mlp": 0.01026164, + "balance_loss_clip": 1.04743731, + "balance_loss_mlp": 1.01804042, + "epoch": 0.7492334515721758, + "flos": 23003011353600.0, + "grad_norm": 2.0471793338938657, + "language_loss": 0.70632982, + "learning_rate": 6.240161630157495e-07, + "loss": 0.72830784, + "num_input_tokens_seen": 133976660, + "step": 6231, + "time_per_iteration": 2.588921070098877 + }, + { + "auxiliary_loss_clip": 0.01171222, + "auxiliary_loss_mlp": 0.01028973, + "balance_loss_clip": 1.04824948, + "balance_loss_mlp": 1.02134407, + "epoch": 0.7493536944628149, + "flos": 16398823835520.0, + "grad_norm": 2.4022136711445436, + "language_loss": 0.70385933, + "learning_rate": 6.23450953381133e-07, + "loss": 0.72586131, + "num_input_tokens_seen": 133994750, + "step": 6232, + "time_per_iteration": 2.533184766769409 + }, + { + "auxiliary_loss_clip": 0.01176277, + "auxiliary_loss_mlp": 0.01023497, + "balance_loss_clip": 1.04145765, + "balance_loss_mlp": 1.01642776, + "epoch": 0.749473937353454, + "flos": 15338561155200.0, + "grad_norm": 1.9803231686204132, + "language_loss": 0.67772824, + "learning_rate": 6.228859525679131e-07, + "loss": 0.69972599, + "num_input_tokens_seen": 134009165, + "step": 6233, + "time_per_iteration": 2.6095800399780273 + }, + { + "auxiliary_loss_clip": 0.01180837, + "auxiliary_loss_mlp": 0.01028265, + "balance_loss_clip": 1.04876423, + "balance_loss_mlp": 1.02131844, + "epoch": 0.7495941802440931, + "flos": 18951138587520.0, + "grad_norm": 2.399106748593753, + "language_loss": 0.7958107, + "learning_rate": 6.223211606617986e-07, + "loss": 0.81790173, + "num_input_tokens_seen": 134027585, + "step": 6234, + "time_per_iteration": 2.556164264678955 + }, + { + "auxiliary_loss_clip": 0.01178534, + "auxiliary_loss_mlp": 0.01023447, + "balance_loss_clip": 1.0500592, + "balance_loss_mlp": 1.01731706, + "epoch": 0.7497144231347321, + "flos": 22492469393280.0, + "grad_norm": 1.7444537341602488, + "language_loss": 0.83798653, + "learning_rate": 6.217565777484701e-07, + "loss": 0.86000633, + "num_input_tokens_seen": 134046680, + "step": 6235, + "time_per_iteration": 2.6212711334228516 + }, + { + "auxiliary_loss_clip": 0.01185694, + "auxiliary_loss_mlp": 0.0137885, + "balance_loss_clip": 1.04549742, + "balance_loss_mlp": 1.00018156, + "epoch": 0.7498346660253713, + "flos": 24243509502720.0, + "grad_norm": 1.7870332281652812, + "language_loss": 0.79917252, + "learning_rate": 6.211922039135722e-07, + "loss": 0.8248179, + "num_input_tokens_seen": 134066825, + "step": 6236, + "time_per_iteration": 2.614243268966675 + }, + { + "auxiliary_loss_clip": 0.01171198, + "auxiliary_loss_mlp": 0.01024896, + "balance_loss_clip": 1.04946983, + "balance_loss_mlp": 1.01776683, + "epoch": 0.7499549089160104, + "flos": 24387080163840.0, + "grad_norm": 2.467525211430689, + "language_loss": 0.81072438, + "learning_rate": 6.206280392427201e-07, + "loss": 0.83268523, + "num_input_tokens_seen": 134086410, + "step": 6237, + "time_per_iteration": 2.5795392990112305 + }, + { + "auxiliary_loss_clip": 0.0116905, + "auxiliary_loss_mlp": 0.01024848, + "balance_loss_clip": 1.04355359, + "balance_loss_mlp": 1.01757038, + "epoch": 0.7500751518066494, + "flos": 34057320704640.0, + "grad_norm": 2.016188919686416, + "language_loss": 0.73808002, + "learning_rate": 6.200640838214983e-07, + "loss": 0.76001906, + "num_input_tokens_seen": 134109185, + "step": 6238, + "time_per_iteration": 2.7403724193573 + }, + { + "auxiliary_loss_clip": 0.01167031, + "auxiliary_loss_mlp": 0.01022077, + "balance_loss_clip": 1.04675889, + "balance_loss_mlp": 1.01521027, + "epoch": 0.7501953946972886, + "flos": 18843586289280.0, + "grad_norm": 1.814884402479432, + "language_loss": 0.66955411, + "learning_rate": 6.195003377354578e-07, + "loss": 0.69144523, + "num_input_tokens_seen": 134128455, + "step": 6239, + "time_per_iteration": 2.6322784423828125 + }, + { + "auxiliary_loss_clip": 0.01172732, + "auxiliary_loss_mlp": 0.01021629, + "balance_loss_clip": 1.04419708, + "balance_loss_mlp": 1.01397276, + "epoch": 0.7503156375879276, + "flos": 20257675891200.0, + "grad_norm": 2.603470670149287, + "language_loss": 0.73989332, + "learning_rate": 6.189368010701183e-07, + "loss": 0.76183695, + "num_input_tokens_seen": 134145515, + "step": 6240, + "time_per_iteration": 2.5788381099700928 + }, + { + "auxiliary_loss_clip": 0.01182383, + "auxiliary_loss_mlp": 0.01027835, + "balance_loss_clip": 1.04532051, + "balance_loss_mlp": 1.02042913, + "epoch": 0.7504358804785667, + "flos": 13480040574720.0, + "grad_norm": 1.9584760299276378, + "language_loss": 0.76508093, + "learning_rate": 6.183734739109683e-07, + "loss": 0.78718311, + "num_input_tokens_seen": 134163335, + "step": 6241, + "time_per_iteration": 2.6248021125793457 + }, + { + "auxiliary_loss_clip": 0.01186552, + "auxiliary_loss_mlp": 0.01028145, + "balance_loss_clip": 1.04911494, + "balance_loss_mlp": 1.02012837, + "epoch": 0.7505561233692057, + "flos": 29461042431360.0, + "grad_norm": 1.9954252521229099, + "language_loss": 0.69066703, + "learning_rate": 6.178103563434629e-07, + "loss": 0.71281397, + "num_input_tokens_seen": 134182335, + "step": 6242, + "time_per_iteration": 2.666409969329834 + }, + { + "auxiliary_loss_clip": 0.01169334, + "auxiliary_loss_mlp": 0.0102168, + "balance_loss_clip": 1.04720974, + "balance_loss_mlp": 1.01425004, + "epoch": 0.7506763662598449, + "flos": 20302457172480.0, + "grad_norm": 1.635126615061084, + "language_loss": 0.83693027, + "learning_rate": 6.172474484530283e-07, + "loss": 0.85884035, + "num_input_tokens_seen": 134201070, + "step": 6243, + "time_per_iteration": 2.5338973999023438 + }, + { + "auxiliary_loss_clip": 0.0117328, + "auxiliary_loss_mlp": 0.01021047, + "balance_loss_clip": 1.04050398, + "balance_loss_mlp": 1.01345968, + "epoch": 0.750796609150484, + "flos": 37230961939200.0, + "grad_norm": 3.267525545314218, + "language_loss": 0.75746715, + "learning_rate": 6.166847503250563e-07, + "loss": 0.77941042, + "num_input_tokens_seen": 134223310, + "step": 6244, + "time_per_iteration": 2.723073720932007 + }, + { + "auxiliary_loss_clip": 0.01185331, + "auxiliary_loss_mlp": 0.01032649, + "balance_loss_clip": 1.04416156, + "balance_loss_mlp": 1.02479935, + "epoch": 0.750916852041123, + "flos": 19609417186560.0, + "grad_norm": 2.991196859586884, + "language_loss": 0.78432965, + "learning_rate": 6.161222620449078e-07, + "loss": 0.80650949, + "num_input_tokens_seen": 134242085, + "step": 6245, + "time_per_iteration": 3.4827334880828857 + }, + { + "auxiliary_loss_clip": 0.01200365, + "auxiliary_loss_mlp": 0.01026605, + "balance_loss_clip": 1.0455904, + "balance_loss_mlp": 1.01887441, + "epoch": 0.7510370949317622, + "flos": 25112690807040.0, + "grad_norm": 2.3131320595962603, + "language_loss": 0.80496335, + "learning_rate": 6.155599836979117e-07, + "loss": 0.82723296, + "num_input_tokens_seen": 134260770, + "step": 6246, + "time_per_iteration": 3.6407759189605713 + }, + { + "auxiliary_loss_clip": 0.01203816, + "auxiliary_loss_mlp": 0.01029625, + "balance_loss_clip": 1.04219019, + "balance_loss_mlp": 1.02127445, + "epoch": 0.7511573378224012, + "flos": 19062282245760.0, + "grad_norm": 2.0068834682315715, + "language_loss": 0.81745678, + "learning_rate": 6.149979153693649e-07, + "loss": 0.83979118, + "num_input_tokens_seen": 134278025, + "step": 6247, + "time_per_iteration": 2.6721348762512207 + }, + { + "auxiliary_loss_clip": 0.01175047, + "auxiliary_loss_mlp": 0.01025753, + "balance_loss_clip": 1.04615593, + "balance_loss_mlp": 1.01815069, + "epoch": 0.7512775807130403, + "flos": 19937676602880.0, + "grad_norm": 2.000421439277236, + "language_loss": 0.77072173, + "learning_rate": 6.144360571445343e-07, + "loss": 0.79272974, + "num_input_tokens_seen": 134297170, + "step": 6248, + "time_per_iteration": 2.5934133529663086 + }, + { + "auxiliary_loss_clip": 0.01177329, + "auxiliary_loss_mlp": 0.01028258, + "balance_loss_clip": 1.0484302, + "balance_loss_mlp": 1.02024698, + "epoch": 0.7513978236036795, + "flos": 20739920912640.0, + "grad_norm": 1.719597503761655, + "language_loss": 0.80223858, + "learning_rate": 6.138744091086509e-07, + "loss": 0.82429445, + "num_input_tokens_seen": 134316755, + "step": 6249, + "time_per_iteration": 2.650000810623169 + }, + { + "auxiliary_loss_clip": 0.01201669, + "auxiliary_loss_mlp": 0.01026479, + "balance_loss_clip": 1.04597425, + "balance_loss_mlp": 1.01966596, + "epoch": 0.7515180664943185, + "flos": 27563163523200.0, + "grad_norm": 2.555591611465025, + "language_loss": 0.72370386, + "learning_rate": 6.133129713469183e-07, + "loss": 0.74598539, + "num_input_tokens_seen": 134335960, + "step": 6250, + "time_per_iteration": 3.628310441970825 + }, + { + "auxiliary_loss_clip": 0.01201726, + "auxiliary_loss_mlp": 0.01022504, + "balance_loss_clip": 1.04056823, + "balance_loss_mlp": 1.01510406, + "epoch": 0.7516383093849576, + "flos": 33803181002880.0, + "grad_norm": 2.417091928812237, + "language_loss": 0.63890243, + "learning_rate": 6.127517439445053e-07, + "loss": 0.66114473, + "num_input_tokens_seen": 134356805, + "step": 6251, + "time_per_iteration": 3.7421441078186035 + }, + { + "auxiliary_loss_clip": 0.01197363, + "auxiliary_loss_mlp": 0.01024362, + "balance_loss_clip": 1.04306388, + "balance_loss_mlp": 1.01744509, + "epoch": 0.7517585522755967, + "flos": 29746172592000.0, + "grad_norm": 1.9826394723639962, + "language_loss": 0.82203484, + "learning_rate": 6.121907269865498e-07, + "loss": 0.84425211, + "num_input_tokens_seen": 134376295, + "step": 6252, + "time_per_iteration": 2.7978739738464355 + }, + { + "auxiliary_loss_clip": 0.01101345, + "auxiliary_loss_mlp": 0.01005082, + "balance_loss_clip": 1.0157423, + "balance_loss_mlp": 1.00395, + "epoch": 0.7518787951662358, + "flos": 69807974319360.0, + "grad_norm": 0.9254433885297444, + "language_loss": 0.67222434, + "learning_rate": 6.116299205581577e-07, + "loss": 0.69328856, + "num_input_tokens_seen": 134431125, + "step": 6253, + "time_per_iteration": 3.2214932441711426 + }, + { + "auxiliary_loss_clip": 0.01174292, + "auxiliary_loss_mlp": 0.01028067, + "balance_loss_clip": 1.0488925, + "balance_loss_mlp": 1.01979327, + "epoch": 0.7519990380568748, + "flos": 34203225749760.0, + "grad_norm": 1.6522302979411017, + "language_loss": 0.68296564, + "learning_rate": 6.110693247444018e-07, + "loss": 0.70498919, + "num_input_tokens_seen": 134452960, + "step": 6254, + "time_per_iteration": 2.7304837703704834 + }, + { + "auxiliary_loss_clip": 0.01183005, + "auxiliary_loss_mlp": 0.01024968, + "balance_loss_clip": 1.0416851, + "balance_loss_mlp": 1.01762164, + "epoch": 0.752119280947514, + "flos": 21725704742400.0, + "grad_norm": 1.7090875304411597, + "language_loss": 0.8255465, + "learning_rate": 6.105089396303258e-07, + "loss": 0.84762621, + "num_input_tokens_seen": 134471350, + "step": 6255, + "time_per_iteration": 2.6461479663848877 + }, + { + "auxiliary_loss_clip": 0.01188096, + "auxiliary_loss_mlp": 0.01024264, + "balance_loss_clip": 1.04552448, + "balance_loss_mlp": 1.01662898, + "epoch": 0.7522395238381531, + "flos": 32742774668160.0, + "grad_norm": 1.997389570620817, + "language_loss": 0.7552911, + "learning_rate": 6.099487653009383e-07, + "loss": 0.77741474, + "num_input_tokens_seen": 134490695, + "step": 6256, + "time_per_iteration": 2.761528491973877 + }, + { + "auxiliary_loss_clip": 0.01176503, + "auxiliary_loss_mlp": 0.0102461, + "balance_loss_clip": 1.04540634, + "balance_loss_mlp": 1.01779127, + "epoch": 0.7523597667287921, + "flos": 23476026579840.0, + "grad_norm": 2.075424021423268, + "language_loss": 0.83171213, + "learning_rate": 6.093888018412192e-07, + "loss": 0.85372329, + "num_input_tokens_seen": 134506885, + "step": 6257, + "time_per_iteration": 2.6006078720092773 + }, + { + "auxiliary_loss_clip": 0.0107874, + "auxiliary_loss_mlp": 0.010014, + "balance_loss_clip": 1.01249671, + "balance_loss_mlp": 1.00028515, + "epoch": 0.7524800096194313, + "flos": 67346730501120.0, + "grad_norm": 0.7119258412818666, + "language_loss": 0.54688853, + "learning_rate": 6.088290493361125e-07, + "loss": 0.56768996, + "num_input_tokens_seen": 134571770, + "step": 6258, + "time_per_iteration": 3.331667423248291 + }, + { + "auxiliary_loss_clip": 0.01193021, + "auxiliary_loss_mlp": 0.01026038, + "balance_loss_clip": 1.03992581, + "balance_loss_mlp": 1.01889491, + "epoch": 0.7526002525100703, + "flos": 13006055681280.0, + "grad_norm": 2.0469567286500125, + "language_loss": 0.71291548, + "learning_rate": 6.082695078705322e-07, + "loss": 0.73510605, + "num_input_tokens_seen": 134589250, + "step": 6259, + "time_per_iteration": 2.661074638366699 + }, + { + "auxiliary_loss_clip": 0.01168481, + "auxiliary_loss_mlp": 0.01025844, + "balance_loss_clip": 1.04464459, + "balance_loss_mlp": 1.01758909, + "epoch": 0.7527204954007094, + "flos": 21397229844480.0, + "grad_norm": 2.2012986980698805, + "language_loss": 0.68919134, + "learning_rate": 6.077101775293618e-07, + "loss": 0.71113455, + "num_input_tokens_seen": 134608075, + "step": 6260, + "time_per_iteration": 2.588512420654297 + }, + { + "auxiliary_loss_clip": 0.01179667, + "auxiliary_loss_mlp": 0.01026312, + "balance_loss_clip": 1.04515433, + "balance_loss_mlp": 1.01777089, + "epoch": 0.7528407382913486, + "flos": 18947188091520.0, + "grad_norm": 3.435086780409005, + "language_loss": 0.82710016, + "learning_rate": 6.071510583974504e-07, + "loss": 0.84915996, + "num_input_tokens_seen": 134623260, + "step": 6261, + "time_per_iteration": 2.7025363445281982 + }, + { + "auxiliary_loss_clip": 0.01171837, + "auxiliary_loss_mlp": 0.0102509, + "balance_loss_clip": 1.05019712, + "balance_loss_mlp": 1.01775885, + "epoch": 0.7529609811819876, + "flos": 15231798956160.0, + "grad_norm": 1.7950653439942466, + "language_loss": 0.71898878, + "learning_rate": 6.065921505596161e-07, + "loss": 0.74095803, + "num_input_tokens_seen": 134641540, + "step": 6262, + "time_per_iteration": 2.5484800338745117 + }, + { + "auxiliary_loss_clip": 0.01194727, + "auxiliary_loss_mlp": 0.01027171, + "balance_loss_clip": 1.04532361, + "balance_loss_mlp": 1.01996183, + "epoch": 0.7530812240726267, + "flos": 19354487385600.0, + "grad_norm": 1.5868544123127668, + "language_loss": 0.76916683, + "learning_rate": 6.060334541006445e-07, + "loss": 0.79138583, + "num_input_tokens_seen": 134660035, + "step": 6263, + "time_per_iteration": 2.7315988540649414 + }, + { + "auxiliary_loss_clip": 0.01193998, + "auxiliary_loss_mlp": 0.01022611, + "balance_loss_clip": 1.04044461, + "balance_loss_mlp": 1.0156076, + "epoch": 0.7532014669632658, + "flos": 27748247328000.0, + "grad_norm": 1.7255798204827715, + "language_loss": 0.69035578, + "learning_rate": 6.05474969105289e-07, + "loss": 0.71252185, + "num_input_tokens_seen": 134683025, + "step": 6264, + "time_per_iteration": 2.763059139251709 + }, + { + "auxiliary_loss_clip": 0.01178463, + "auxiliary_loss_mlp": 0.01027703, + "balance_loss_clip": 1.0460881, + "balance_loss_mlp": 1.02011561, + "epoch": 0.7533217098539049, + "flos": 14137421333760.0, + "grad_norm": 3.434597362326491, + "language_loss": 0.7376458, + "learning_rate": 6.049166956582725e-07, + "loss": 0.75970745, + "num_input_tokens_seen": 134701290, + "step": 6265, + "time_per_iteration": 2.6370012760162354 + }, + { + "auxiliary_loss_clip": 0.01171815, + "auxiliary_loss_mlp": 0.01025811, + "balance_loss_clip": 1.04534674, + "balance_loss_mlp": 1.01898038, + "epoch": 0.753441952744544, + "flos": 26429068437120.0, + "grad_norm": 1.918485586107922, + "language_loss": 0.87182796, + "learning_rate": 6.043586338442841e-07, + "loss": 0.89380431, + "num_input_tokens_seen": 134720345, + "step": 6266, + "time_per_iteration": 2.639569044113159 + }, + { + "auxiliary_loss_clip": 0.01164106, + "auxiliary_loss_mlp": 0.01024296, + "balance_loss_clip": 1.0466342, + "balance_loss_mlp": 1.01770103, + "epoch": 0.7535621956351831, + "flos": 23878621192320.0, + "grad_norm": 1.5679843059167669, + "language_loss": 0.73166662, + "learning_rate": 6.038007837479815e-07, + "loss": 0.75355065, + "num_input_tokens_seen": 134741450, + "step": 6267, + "time_per_iteration": 2.579724073410034 + }, + { + "auxiliary_loss_clip": 0.01177796, + "auxiliary_loss_mlp": 0.01026648, + "balance_loss_clip": 1.04844832, + "balance_loss_mlp": 1.0194658, + "epoch": 0.7536824385258222, + "flos": 21795873960960.0, + "grad_norm": 1.845104605643479, + "language_loss": 0.64307857, + "learning_rate": 6.032431454539897e-07, + "loss": 0.66512305, + "num_input_tokens_seen": 134760295, + "step": 6268, + "time_per_iteration": 2.591648578643799 + }, + { + "auxiliary_loss_clip": 0.01196954, + "auxiliary_loss_mlp": 0.01026292, + "balance_loss_clip": 1.04430485, + "balance_loss_mlp": 1.01871037, + "epoch": 0.7538026814164612, + "flos": 28911644933760.0, + "grad_norm": 1.6827096686334235, + "language_loss": 0.81307518, + "learning_rate": 6.026857190469014e-07, + "loss": 0.83530772, + "num_input_tokens_seen": 134782050, + "step": 6269, + "time_per_iteration": 2.799913167953491 + }, + { + "auxiliary_loss_clip": 0.0118935, + "auxiliary_loss_mlp": 0.01023898, + "balance_loss_clip": 1.04662371, + "balance_loss_mlp": 1.01691794, + "epoch": 0.7539229243071004, + "flos": 21104701482240.0, + "grad_norm": 1.9011536856819664, + "language_loss": 0.7406252, + "learning_rate": 6.0212850461128e-07, + "loss": 0.76275772, + "num_input_tokens_seen": 134801170, + "step": 6270, + "time_per_iteration": 2.6102824211120605 + }, + { + "auxiliary_loss_clip": 0.01187211, + "auxiliary_loss_mlp": 0.01021618, + "balance_loss_clip": 1.0436697, + "balance_loss_mlp": 1.01389313, + "epoch": 0.7540431671977395, + "flos": 15158469340800.0, + "grad_norm": 2.2004870234367533, + "language_loss": 0.74593914, + "learning_rate": 6.015715022316516e-07, + "loss": 0.76802742, + "num_input_tokens_seen": 134819150, + "step": 6271, + "time_per_iteration": 3.525312900543213 + }, + { + "auxiliary_loss_clip": 0.01203623, + "auxiliary_loss_mlp": 0.01022731, + "balance_loss_clip": 1.04075313, + "balance_loss_mlp": 1.0150001, + "epoch": 0.7541634100883785, + "flos": 18770579896320.0, + "grad_norm": 2.853993420580086, + "language_loss": 0.78116894, + "learning_rate": 6.010147119925154e-07, + "loss": 0.80343246, + "num_input_tokens_seen": 134836905, + "step": 6272, + "time_per_iteration": 3.7501237392425537 + }, + { + "auxiliary_loss_clip": 0.01184569, + "auxiliary_loss_mlp": 0.01022358, + "balance_loss_clip": 1.04159343, + "balance_loss_mlp": 1.01539636, + "epoch": 0.7542836529790176, + "flos": 20594770053120.0, + "grad_norm": 2.1527828978149692, + "language_loss": 0.66696495, + "learning_rate": 6.004581339783348e-07, + "loss": 0.68903422, + "num_input_tokens_seen": 134855225, + "step": 6273, + "time_per_iteration": 2.6449825763702393 + }, + { + "auxiliary_loss_clip": 0.01181674, + "auxiliary_loss_mlp": 0.01025842, + "balance_loss_clip": 1.04606414, + "balance_loss_mlp": 1.01820636, + "epoch": 0.7544038958696567, + "flos": 19095104298240.0, + "grad_norm": 2.686549498491915, + "language_loss": 0.68486309, + "learning_rate": 5.999017682735425e-07, + "loss": 0.70693827, + "num_input_tokens_seen": 134871615, + "step": 6274, + "time_per_iteration": 2.6162269115448 + }, + { + "auxiliary_loss_clip": 0.01220959, + "auxiliary_loss_mlp": 0.01027874, + "balance_loss_clip": 1.04139256, + "balance_loss_mlp": 1.02014291, + "epoch": 0.7545241387602958, + "flos": 31723306859520.0, + "grad_norm": 1.8550262386852208, + "language_loss": 0.66529775, + "learning_rate": 5.993456149625387e-07, + "loss": 0.68778598, + "num_input_tokens_seen": 134892765, + "step": 6275, + "time_per_iteration": 2.863647699356079 + }, + { + "auxiliary_loss_clip": 0.01185023, + "auxiliary_loss_mlp": 0.01024949, + "balance_loss_clip": 1.04297483, + "balance_loss_mlp": 1.01736116, + "epoch": 0.7546443816509348, + "flos": 20296495514880.0, + "grad_norm": 1.9007438283042404, + "language_loss": 0.82483423, + "learning_rate": 5.987896741296909e-07, + "loss": 0.84693396, + "num_input_tokens_seen": 134910505, + "step": 6276, + "time_per_iteration": 3.5660226345062256 + }, + { + "auxiliary_loss_clip": 0.01190055, + "auxiliary_loss_mlp": 0.0102594, + "balance_loss_clip": 1.04832649, + "balance_loss_mlp": 1.01826644, + "epoch": 0.754764624541574, + "flos": 23696159080320.0, + "grad_norm": 2.1690898168100032, + "language_loss": 0.78363991, + "learning_rate": 5.982339458593361e-07, + "loss": 0.80579984, + "num_input_tokens_seen": 134930445, + "step": 6277, + "time_per_iteration": 3.5792148113250732 + }, + { + "auxiliary_loss_clip": 0.01176436, + "auxiliary_loss_mlp": 0.01378729, + "balance_loss_clip": 1.04801464, + "balance_loss_mlp": 1.00017881, + "epoch": 0.7548848674322131, + "flos": 25337204766720.0, + "grad_norm": 1.580230813738504, + "language_loss": 0.84086573, + "learning_rate": 5.976784302357767e-07, + "loss": 0.86641735, + "num_input_tokens_seen": 134951010, + "step": 6278, + "time_per_iteration": 2.6221868991851807 + }, + { + "auxiliary_loss_clip": 0.01180864, + "auxiliary_loss_mlp": 0.0102387, + "balance_loss_clip": 1.04751945, + "balance_loss_mlp": 1.0164845, + "epoch": 0.7550051103228521, + "flos": 19573147428480.0, + "grad_norm": 2.1113427221808405, + "language_loss": 0.73531264, + "learning_rate": 5.971231273432855e-07, + "loss": 0.75735998, + "num_input_tokens_seen": 134970495, + "step": 6279, + "time_per_iteration": 2.6298818588256836 + }, + { + "auxiliary_loss_clip": 0.01078552, + "auxiliary_loss_mlp": 0.01002775, + "balance_loss_clip": 1.01289749, + "balance_loss_mlp": 1.00173223, + "epoch": 0.7551253532134913, + "flos": 64150068648960.0, + "grad_norm": 0.8144928799836356, + "language_loss": 0.54539788, + "learning_rate": 5.965680372661e-07, + "loss": 0.5662111, + "num_input_tokens_seen": 135028060, + "step": 6280, + "time_per_iteration": 3.115767002105713 + }, + { + "auxiliary_loss_clip": 0.01188678, + "auxiliary_loss_mlp": 0.01022625, + "balance_loss_clip": 1.04577208, + "balance_loss_mlp": 1.01576126, + "epoch": 0.7552455961041303, + "flos": 26067986968320.0, + "grad_norm": 1.7518592540656153, + "language_loss": 0.56348252, + "learning_rate": 5.960131600884266e-07, + "loss": 0.58559549, + "num_input_tokens_seen": 135047330, + "step": 6281, + "time_per_iteration": 2.627013921737671 + }, + { + "auxiliary_loss_clip": 0.01195032, + "auxiliary_loss_mlp": 0.01024162, + "balance_loss_clip": 1.04187822, + "balance_loss_mlp": 1.01751006, + "epoch": 0.7553658389947694, + "flos": 24498223822080.0, + "grad_norm": 1.6030138702179109, + "language_loss": 0.76033676, + "learning_rate": 5.954584958944413e-07, + "loss": 0.78252876, + "num_input_tokens_seen": 135065995, + "step": 6282, + "time_per_iteration": 2.7379095554351807 + }, + { + "auxiliary_loss_clip": 0.01195793, + "auxiliary_loss_mlp": 0.01378708, + "balance_loss_clip": 1.04147243, + "balance_loss_mlp": 1.00021172, + "epoch": 0.7554860818854086, + "flos": 21799465320960.0, + "grad_norm": 2.5724529963632286, + "language_loss": 0.81314605, + "learning_rate": 5.949040447682854e-07, + "loss": 0.83889103, + "num_input_tokens_seen": 135085820, + "step": 6283, + "time_per_iteration": 2.7415552139282227 + }, + { + "auxiliary_loss_clip": 0.01191685, + "auxiliary_loss_mlp": 0.01027551, + "balance_loss_clip": 1.04491568, + "balance_loss_mlp": 1.02009439, + "epoch": 0.7556063247760476, + "flos": 16362123114240.0, + "grad_norm": 1.9694914594103796, + "language_loss": 0.68237805, + "learning_rate": 5.943498067940686e-07, + "loss": 0.70457041, + "num_input_tokens_seen": 135102845, + "step": 6284, + "time_per_iteration": 2.6378798484802246 + }, + { + "auxiliary_loss_clip": 0.01182681, + "auxiliary_loss_mlp": 0.01028829, + "balance_loss_clip": 1.0498184, + "balance_loss_mlp": 1.02132177, + "epoch": 0.7557265676666867, + "flos": 27235155502080.0, + "grad_norm": 1.754082363678534, + "language_loss": 0.81675941, + "learning_rate": 5.937957820558686e-07, + "loss": 0.83887452, + "num_input_tokens_seen": 135122190, + "step": 6285, + "time_per_iteration": 2.6638753414154053 + }, + { + "auxiliary_loss_clip": 0.01091543, + "auxiliary_loss_mlp": 0.00998265, + "balance_loss_clip": 1.0115819, + "balance_loss_mlp": 0.9971506, + "epoch": 0.7558468105573258, + "flos": 62189131415040.0, + "grad_norm": 0.851414729409587, + "language_loss": 0.65499282, + "learning_rate": 5.932419706377296e-07, + "loss": 0.67589092, + "num_input_tokens_seen": 135180495, + "step": 6286, + "time_per_iteration": 3.201930046081543 + }, + { + "auxiliary_loss_clip": 0.01194991, + "auxiliary_loss_mlp": 0.01029819, + "balance_loss_clip": 1.04635882, + "balance_loss_mlp": 1.02261269, + "epoch": 0.7559670534479649, + "flos": 33249078823680.0, + "grad_norm": 2.0124058089944747, + "language_loss": 0.73780632, + "learning_rate": 5.92688372623666e-07, + "loss": 0.76005435, + "num_input_tokens_seen": 135199200, + "step": 6287, + "time_per_iteration": 2.7479023933410645 + }, + { + "auxiliary_loss_clip": 0.01177583, + "auxiliary_loss_mlp": 0.01024939, + "balance_loss_clip": 1.04476714, + "balance_loss_mlp": 1.01765501, + "epoch": 0.7560872963386039, + "flos": 14064379027200.0, + "grad_norm": 1.9883906221546097, + "language_loss": 0.74091858, + "learning_rate": 5.921349880976574e-07, + "loss": 0.76294374, + "num_input_tokens_seen": 135217035, + "step": 6288, + "time_per_iteration": 2.6485278606414795 + }, + { + "auxiliary_loss_clip": 0.01189483, + "auxiliary_loss_mlp": 0.0137909, + "balance_loss_clip": 1.04343486, + "balance_loss_mlp": 1.0001626, + "epoch": 0.7562075392292431, + "flos": 20412307941120.0, + "grad_norm": 1.895687740643294, + "language_loss": 0.81487101, + "learning_rate": 5.915818171436515e-07, + "loss": 0.84055674, + "num_input_tokens_seen": 135236370, + "step": 6289, + "time_per_iteration": 2.6263246536254883 + }, + { + "auxiliary_loss_clip": 0.01181442, + "auxiliary_loss_mlp": 0.01025927, + "balance_loss_clip": 1.04055512, + "balance_loss_mlp": 1.01838446, + "epoch": 0.7563277821198822, + "flos": 20376792368640.0, + "grad_norm": 1.546127600835559, + "language_loss": 0.74791932, + "learning_rate": 5.910288598455642e-07, + "loss": 0.76999301, + "num_input_tokens_seen": 135255720, + "step": 6290, + "time_per_iteration": 2.6441078186035156 + }, + { + "auxiliary_loss_clip": 0.01183287, + "auxiliary_loss_mlp": 0.01024999, + "balance_loss_clip": 1.04650497, + "balance_loss_mlp": 1.01749444, + "epoch": 0.7564480250105212, + "flos": 18588261438720.0, + "grad_norm": 2.901364690111718, + "language_loss": 0.74219763, + "learning_rate": 5.90476116287278e-07, + "loss": 0.7642805, + "num_input_tokens_seen": 135273320, + "step": 6291, + "time_per_iteration": 2.616825580596924 + }, + { + "auxiliary_loss_clip": 0.01188338, + "auxiliary_loss_mlp": 0.0102509, + "balance_loss_clip": 1.04718995, + "balance_loss_mlp": 1.01761258, + "epoch": 0.7565682679011604, + "flos": 21215521918080.0, + "grad_norm": 1.7885438122224881, + "language_loss": 0.67895091, + "learning_rate": 5.899235865526456e-07, + "loss": 0.70108521, + "num_input_tokens_seen": 135292615, + "step": 6292, + "time_per_iteration": 2.6207311153411865 + }, + { + "auxiliary_loss_clip": 0.01185616, + "auxiliary_loss_mlp": 0.01026125, + "balance_loss_clip": 1.04192424, + "balance_loss_mlp": 1.01891315, + "epoch": 0.7566885107917994, + "flos": 20449008662400.0, + "grad_norm": 1.6931351465901465, + "language_loss": 0.82153445, + "learning_rate": 5.893712707254825e-07, + "loss": 0.84365189, + "num_input_tokens_seen": 135310075, + "step": 6293, + "time_per_iteration": 2.679926872253418 + }, + { + "auxiliary_loss_clip": 0.01198182, + "auxiliary_loss_mlp": 0.0103071, + "balance_loss_clip": 1.03994107, + "balance_loss_mlp": 1.02300882, + "epoch": 0.7568087536824385, + "flos": 19025832919680.0, + "grad_norm": 2.9005681271537656, + "language_loss": 0.65946019, + "learning_rate": 5.888191688895769e-07, + "loss": 0.68174905, + "num_input_tokens_seen": 135327335, + "step": 6294, + "time_per_iteration": 2.6752703189849854 + }, + { + "auxiliary_loss_clip": 0.01169375, + "auxiliary_loss_mlp": 0.01022343, + "balance_loss_clip": 1.0464015, + "balance_loss_mlp": 1.01419842, + "epoch": 0.7569289965730777, + "flos": 15225442248960.0, + "grad_norm": 2.3952602732021893, + "language_loss": 0.62024397, + "learning_rate": 5.882672811286813e-07, + "loss": 0.64216113, + "num_input_tokens_seen": 135343615, + "step": 6295, + "time_per_iteration": 2.6715047359466553 + }, + { + "auxiliary_loss_clip": 0.01172304, + "auxiliary_loss_mlp": 0.01025659, + "balance_loss_clip": 1.04802704, + "balance_loss_mlp": 1.01769018, + "epoch": 0.7570492394637167, + "flos": 20769367086720.0, + "grad_norm": 2.0187214575385286, + "language_loss": 0.69377005, + "learning_rate": 5.877156075265166e-07, + "loss": 0.71574974, + "num_input_tokens_seen": 135359880, + "step": 6296, + "time_per_iteration": 2.5825154781341553 + }, + { + "auxiliary_loss_clip": 0.01181183, + "auxiliary_loss_mlp": 0.01028832, + "balance_loss_clip": 1.04117584, + "balance_loss_mlp": 1.02123237, + "epoch": 0.7571694823543558, + "flos": 15664091137920.0, + "grad_norm": 2.7983924447063746, + "language_loss": 0.70192432, + "learning_rate": 5.871641481667715e-07, + "loss": 0.72402447, + "num_input_tokens_seen": 135374325, + "step": 6297, + "time_per_iteration": 3.6396656036376953 + }, + { + "auxiliary_loss_clip": 0.01209112, + "auxiliary_loss_mlp": 0.01028885, + "balance_loss_clip": 1.04300332, + "balance_loss_mlp": 1.02165484, + "epoch": 0.7572897252449949, + "flos": 25409241492480.0, + "grad_norm": 1.893175902033906, + "language_loss": 0.84511399, + "learning_rate": 5.866129031331011e-07, + "loss": 0.86749399, + "num_input_tokens_seen": 135393980, + "step": 6298, + "time_per_iteration": 3.858666181564331 + }, + { + "auxiliary_loss_clip": 0.01187728, + "auxiliary_loss_mlp": 0.01028085, + "balance_loss_clip": 1.04412341, + "balance_loss_mlp": 1.02030659, + "epoch": 0.757409968135634, + "flos": 24279348297600.0, + "grad_norm": 2.360476105868263, + "language_loss": 0.83434153, + "learning_rate": 5.8606187250913e-07, + "loss": 0.85649967, + "num_input_tokens_seen": 135412030, + "step": 6299, + "time_per_iteration": 2.6613779067993164 + }, + { + "auxiliary_loss_clip": 0.01182487, + "auxiliary_loss_mlp": 0.01378856, + "balance_loss_clip": 1.0509901, + "balance_loss_mlp": 1.0002135, + "epoch": 0.757530211026273, + "flos": 24133766474880.0, + "grad_norm": 1.8947759755271167, + "language_loss": 0.84067613, + "learning_rate": 5.855110563784482e-07, + "loss": 0.86628956, + "num_input_tokens_seen": 135430565, + "step": 6300, + "time_per_iteration": 2.6768600940704346 + }, + { + "auxiliary_loss_clip": 0.01170385, + "auxiliary_loss_mlp": 0.01378489, + "balance_loss_clip": 1.04471564, + "balance_loss_mlp": 1.00017285, + "epoch": 0.7576504539169122, + "flos": 23951807153280.0, + "grad_norm": 1.7521602645338112, + "language_loss": 0.6403476, + "learning_rate": 5.849604548246156e-07, + "loss": 0.66583639, + "num_input_tokens_seen": 135451675, + "step": 6301, + "time_per_iteration": 2.638519763946533 + }, + { + "auxiliary_loss_clip": 0.01193424, + "auxiliary_loss_mlp": 0.01378845, + "balance_loss_clip": 1.04763317, + "balance_loss_mlp": 1.00021243, + "epoch": 0.7577706968075513, + "flos": 21251360712960.0, + "grad_norm": 1.8733368485080013, + "language_loss": 0.80332983, + "learning_rate": 5.844100679311565e-07, + "loss": 0.82905245, + "num_input_tokens_seen": 135470635, + "step": 6302, + "time_per_iteration": 3.5639822483062744 + }, + { + "auxiliary_loss_clip": 0.01186539, + "auxiliary_loss_mlp": 0.01022893, + "balance_loss_clip": 1.04610431, + "balance_loss_mlp": 1.01561785, + "epoch": 0.7578909396981903, + "flos": 18296595002880.0, + "grad_norm": 3.488511825048588, + "language_loss": 0.76312137, + "learning_rate": 5.838598957815637e-07, + "loss": 0.78521574, + "num_input_tokens_seen": 135487865, + "step": 6303, + "time_per_iteration": 3.530137062072754 + }, + { + "auxiliary_loss_clip": 0.01180177, + "auxiliary_loss_mlp": 0.01024278, + "balance_loss_clip": 1.043841, + "balance_loss_mlp": 1.01731944, + "epoch": 0.7580111825888295, + "flos": 25373869574400.0, + "grad_norm": 1.5275759947812322, + "language_loss": 0.85273898, + "learning_rate": 5.833099384592996e-07, + "loss": 0.87478352, + "num_input_tokens_seen": 135508440, + "step": 6304, + "time_per_iteration": 2.7296457290649414 + }, + { + "auxiliary_loss_clip": 0.01184415, + "auxiliary_loss_mlp": 0.01027003, + "balance_loss_clip": 1.04561257, + "balance_loss_mlp": 1.01914978, + "epoch": 0.7581314254794685, + "flos": 23768662682880.0, + "grad_norm": 2.7486706762969635, + "language_loss": 0.71488321, + "learning_rate": 5.827601960477913e-07, + "loss": 0.73699737, + "num_input_tokens_seen": 135526365, + "step": 6305, + "time_per_iteration": 2.7355079650878906 + }, + { + "auxiliary_loss_clip": 0.0117627, + "auxiliary_loss_mlp": 0.01026702, + "balance_loss_clip": 1.04583669, + "balance_loss_mlp": 1.01954973, + "epoch": 0.7582516683701076, + "flos": 22054610603520.0, + "grad_norm": 3.3723547009290193, + "language_loss": 0.70542246, + "learning_rate": 5.822106686304344e-07, + "loss": 0.72745216, + "num_input_tokens_seen": 135545655, + "step": 6306, + "time_per_iteration": 2.602128267288208 + }, + { + "auxiliary_loss_clip": 0.0119896, + "auxiliary_loss_mlp": 0.01026155, + "balance_loss_clip": 1.04326463, + "balance_loss_mlp": 1.018803, + "epoch": 0.7583719112607467, + "flos": 31649725848960.0, + "grad_norm": 1.938312965751115, + "language_loss": 0.57852137, + "learning_rate": 5.816613562905919e-07, + "loss": 0.60077262, + "num_input_tokens_seen": 135566840, + "step": 6307, + "time_per_iteration": 2.7448999881744385 + }, + { + "auxiliary_loss_clip": 0.01190649, + "auxiliary_loss_mlp": 0.01021764, + "balance_loss_clip": 1.04543328, + "balance_loss_mlp": 1.01430106, + "epoch": 0.7584921541513858, + "flos": 33068376478080.0, + "grad_norm": 1.5160473301762962, + "language_loss": 0.69767976, + "learning_rate": 5.811122591115933e-07, + "loss": 0.71980381, + "num_input_tokens_seen": 135587825, + "step": 6308, + "time_per_iteration": 2.8056063652038574 + }, + { + "auxiliary_loss_clip": 0.01196405, + "auxiliary_loss_mlp": 0.0103065, + "balance_loss_clip": 1.04780519, + "balance_loss_mlp": 1.02315724, + "epoch": 0.7586123970420249, + "flos": 23326350606720.0, + "grad_norm": 2.2308695383610275, + "language_loss": 0.71567893, + "learning_rate": 5.805633771767376e-07, + "loss": 0.73794949, + "num_input_tokens_seen": 135605220, + "step": 6309, + "time_per_iteration": 2.745985269546509 + }, + { + "auxiliary_loss_clip": 0.01183749, + "auxiliary_loss_mlp": 0.01025293, + "balance_loss_clip": 1.04658747, + "balance_loss_mlp": 1.01771998, + "epoch": 0.7587326399326639, + "flos": 18334229477760.0, + "grad_norm": 1.8672260759311148, + "language_loss": 0.77634108, + "learning_rate": 5.800147105692888e-07, + "loss": 0.79843146, + "num_input_tokens_seen": 135624795, + "step": 6310, + "time_per_iteration": 2.6691184043884277 + }, + { + "auxiliary_loss_clip": 0.01177455, + "auxiliary_loss_mlp": 0.01026265, + "balance_loss_clip": 1.04373276, + "balance_loss_mlp": 1.01908541, + "epoch": 0.7588528828233031, + "flos": 17275080119040.0, + "grad_norm": 1.6876914106474525, + "language_loss": 0.79177248, + "learning_rate": 5.794662593724795e-07, + "loss": 0.81380975, + "num_input_tokens_seen": 135643800, + "step": 6311, + "time_per_iteration": 2.624098300933838 + }, + { + "auxiliary_loss_clip": 0.01170549, + "auxiliary_loss_mlp": 0.01027632, + "balance_loss_clip": 1.04969215, + "balance_loss_mlp": 1.01992857, + "epoch": 0.7589731257139422, + "flos": 17713621267200.0, + "grad_norm": 6.911254871367875, + "language_loss": 0.75272912, + "learning_rate": 5.789180236695091e-07, + "loss": 0.77471095, + "num_input_tokens_seen": 135660655, + "step": 6312, + "time_per_iteration": 2.603900671005249 + }, + { + "auxiliary_loss_clip": 0.01171539, + "auxiliary_loss_mlp": 0.01022437, + "balance_loss_clip": 1.04523277, + "balance_loss_mlp": 1.01534152, + "epoch": 0.7590933686045812, + "flos": 15961072786560.0, + "grad_norm": 1.9633624488705739, + "language_loss": 0.85076404, + "learning_rate": 5.78370003543544e-07, + "loss": 0.87270379, + "num_input_tokens_seen": 135679410, + "step": 6313, + "time_per_iteration": 2.569019317626953 + }, + { + "auxiliary_loss_clip": 0.01180293, + "auxiliary_loss_mlp": 0.01378956, + "balance_loss_clip": 1.04707301, + "balance_loss_mlp": 1.00025487, + "epoch": 0.7592136114952204, + "flos": 21068072588160.0, + "grad_norm": 3.8353441790676603, + "language_loss": 0.83636403, + "learning_rate": 5.778221990777203e-07, + "loss": 0.86195648, + "num_input_tokens_seen": 135697150, + "step": 6314, + "time_per_iteration": 2.642076015472412 + }, + { + "auxiliary_loss_clip": 0.01187787, + "auxiliary_loss_mlp": 0.01030967, + "balance_loss_clip": 1.04668355, + "balance_loss_mlp": 1.02365899, + "epoch": 0.7593338543858594, + "flos": 25297666871040.0, + "grad_norm": 2.1526704460668253, + "language_loss": 0.82575154, + "learning_rate": 5.772746103551372e-07, + "loss": 0.84793913, + "num_input_tokens_seen": 135712545, + "step": 6315, + "time_per_iteration": 2.6405770778656006 + }, + { + "auxiliary_loss_clip": 0.01186045, + "auxiliary_loss_mlp": 0.01027033, + "balance_loss_clip": 1.04618168, + "balance_loss_mlp": 1.01980329, + "epoch": 0.7594540972764985, + "flos": 31832367528960.0, + "grad_norm": 1.8289677279700844, + "language_loss": 0.71926409, + "learning_rate": 5.767272374588648e-07, + "loss": 0.74139488, + "num_input_tokens_seen": 135733950, + "step": 6316, + "time_per_iteration": 2.737161636352539 + }, + { + "auxiliary_loss_clip": 0.01178503, + "auxiliary_loss_mlp": 0.01023815, + "balance_loss_clip": 1.0484004, + "balance_loss_mlp": 1.01632893, + "epoch": 0.7595743401671377, + "flos": 37597250880000.0, + "grad_norm": 2.316749829320917, + "language_loss": 0.78146482, + "learning_rate": 5.76180080471939e-07, + "loss": 0.80348796, + "num_input_tokens_seen": 135757120, + "step": 6317, + "time_per_iteration": 2.772904872894287 + }, + { + "auxiliary_loss_clip": 0.01173596, + "auxiliary_loss_mlp": 0.01026108, + "balance_loss_clip": 1.04824471, + "balance_loss_mlp": 1.01757312, + "epoch": 0.7596945830577767, + "flos": 18287724343680.0, + "grad_norm": 2.06745031461266, + "language_loss": 0.72195184, + "learning_rate": 5.756331394773631e-07, + "loss": 0.74394888, + "num_input_tokens_seen": 135773335, + "step": 6318, + "time_per_iteration": 2.570411443710327 + }, + { + "auxiliary_loss_clip": 0.01214242, + "auxiliary_loss_mlp": 0.01379305, + "balance_loss_clip": 1.04052067, + "balance_loss_mlp": 1.00016654, + "epoch": 0.7598148259484158, + "flos": 22233122219520.0, + "grad_norm": 1.7318318113733906, + "language_loss": 0.76167804, + "learning_rate": 5.750864145581071e-07, + "loss": 0.78761351, + "num_input_tokens_seen": 135792555, + "step": 6319, + "time_per_iteration": 2.766444683074951 + }, + { + "auxiliary_loss_clip": 0.01171743, + "auxiliary_loss_mlp": 0.01023666, + "balance_loss_clip": 1.05112505, + "balance_loss_mlp": 1.01616836, + "epoch": 0.7599350688390549, + "flos": 27161718145920.0, + "grad_norm": 1.9941843433670012, + "language_loss": 0.86210608, + "learning_rate": 5.745399057971085e-07, + "loss": 0.88406014, + "num_input_tokens_seen": 135813690, + "step": 6320, + "time_per_iteration": 2.669557571411133 + }, + { + "auxiliary_loss_clip": 0.01178621, + "auxiliary_loss_mlp": 0.01025521, + "balance_loss_clip": 1.04552436, + "balance_loss_mlp": 1.018062, + "epoch": 0.760055311729694, + "flos": 15560704817280.0, + "grad_norm": 2.1985750405911073, + "language_loss": 0.75569284, + "learning_rate": 5.739936132772738e-07, + "loss": 0.77773422, + "num_input_tokens_seen": 135832255, + "step": 6321, + "time_per_iteration": 2.590399980545044 + }, + { + "auxiliary_loss_clip": 0.011683, + "auxiliary_loss_mlp": 0.01029695, + "balance_loss_clip": 1.04656458, + "balance_loss_mlp": 1.02199411, + "epoch": 0.760175554620333, + "flos": 25155496840320.0, + "grad_norm": 2.1422189633799142, + "language_loss": 0.74402297, + "learning_rate": 5.734475370814733e-07, + "loss": 0.76600301, + "num_input_tokens_seen": 135851935, + "step": 6322, + "time_per_iteration": 2.625296115875244 + }, + { + "auxiliary_loss_clip": 0.01178989, + "auxiliary_loss_mlp": 0.01028058, + "balance_loss_clip": 1.04545307, + "balance_loss_mlp": 1.0207597, + "epoch": 0.7602957975109722, + "flos": 24353791234560.0, + "grad_norm": 1.669128044074278, + "language_loss": 0.78215921, + "learning_rate": 5.729016772925483e-07, + "loss": 0.80422974, + "num_input_tokens_seen": 135873510, + "step": 6323, + "time_per_iteration": 3.6072323322296143 + }, + { + "auxiliary_loss_clip": 0.01199349, + "auxiliary_loss_mlp": 0.01031116, + "balance_loss_clip": 1.0428369, + "balance_loss_mlp": 1.02318287, + "epoch": 0.7604160404016113, + "flos": 25192664438400.0, + "grad_norm": 1.7212566151465063, + "language_loss": 0.70827347, + "learning_rate": 5.723560339933038e-07, + "loss": 0.73057812, + "num_input_tokens_seen": 135893845, + "step": 6324, + "time_per_iteration": 3.6584219932556152 + }, + { + "auxiliary_loss_clip": 0.01174462, + "auxiliary_loss_mlp": 0.01379089, + "balance_loss_clip": 1.04532087, + "balance_loss_mlp": 1.00031364, + "epoch": 0.7605362832922503, + "flos": 29861841363840.0, + "grad_norm": 2.0424777076176754, + "language_loss": 0.6532144, + "learning_rate": 5.71810607266513e-07, + "loss": 0.67874992, + "num_input_tokens_seen": 135912430, + "step": 6325, + "time_per_iteration": 2.6996920108795166 + }, + { + "auxiliary_loss_clip": 0.01180807, + "auxiliary_loss_mlp": 0.01023742, + "balance_loss_clip": 1.04611492, + "balance_loss_mlp": 1.01641679, + "epoch": 0.7606565261828895, + "flos": 13917935278080.0, + "grad_norm": 1.8386125615451792, + "language_loss": 0.60797226, + "learning_rate": 5.712653971949184e-07, + "loss": 0.63001776, + "num_input_tokens_seen": 135930550, + "step": 6326, + "time_per_iteration": 2.6226909160614014 + }, + { + "auxiliary_loss_clip": 0.01171093, + "auxiliary_loss_mlp": 0.01026022, + "balance_loss_clip": 1.04435658, + "balance_loss_mlp": 1.01839244, + "epoch": 0.7607767690735285, + "flos": 18551273408640.0, + "grad_norm": 2.491882296356066, + "language_loss": 0.75146514, + "learning_rate": 5.707204038612268e-07, + "loss": 0.77343631, + "num_input_tokens_seen": 135947980, + "step": 6327, + "time_per_iteration": 2.599311113357544 + }, + { + "auxiliary_loss_clip": 0.0119346, + "auxiliary_loss_mlp": 0.01030091, + "balance_loss_clip": 1.04931378, + "balance_loss_mlp": 1.02200246, + "epoch": 0.7608970119641676, + "flos": 20922993555840.0, + "grad_norm": 2.1357847875443956, + "language_loss": 0.74072355, + "learning_rate": 5.701756273481138e-07, + "loss": 0.76295906, + "num_input_tokens_seen": 135965400, + "step": 6328, + "time_per_iteration": 3.657656192779541 + }, + { + "auxiliary_loss_clip": 0.01189057, + "auxiliary_loss_mlp": 0.01026708, + "balance_loss_clip": 1.04465842, + "balance_loss_mlp": 1.01981711, + "epoch": 0.7610172548548068, + "flos": 23807302738560.0, + "grad_norm": 1.5999339734209226, + "language_loss": 0.73822534, + "learning_rate": 5.696310677382212e-07, + "loss": 0.76038295, + "num_input_tokens_seen": 135986795, + "step": 6329, + "time_per_iteration": 3.5806305408477783 + }, + { + "auxiliary_loss_clip": 0.0110124, + "auxiliary_loss_mlp": 0.01000535, + "balance_loss_clip": 1.01409864, + "balance_loss_mlp": 0.999313, + "epoch": 0.7611374977454458, + "flos": 66496580426880.0, + "grad_norm": 0.8684489170511885, + "language_loss": 0.61763209, + "learning_rate": 5.690867251141576e-07, + "loss": 0.63864982, + "num_input_tokens_seen": 136053450, + "step": 6330, + "time_per_iteration": 3.375471830368042 + }, + { + "auxiliary_loss_clip": 0.01185234, + "auxiliary_loss_mlp": 0.01026238, + "balance_loss_clip": 1.04701209, + "balance_loss_mlp": 1.01808393, + "epoch": 0.7612577406360849, + "flos": 15633136592640.0, + "grad_norm": 3.0007764218946305, + "language_loss": 0.91784942, + "learning_rate": 5.685425995585013e-07, + "loss": 0.93996418, + "num_input_tokens_seen": 136071375, + "step": 6331, + "time_per_iteration": 2.636934518814087 + }, + { + "auxiliary_loss_clip": 0.01089683, + "auxiliary_loss_mlp": 0.01000216, + "balance_loss_clip": 1.01074159, + "balance_loss_mlp": 0.9991129, + "epoch": 0.761377983526724, + "flos": 60526253237760.0, + "grad_norm": 0.7549763092478111, + "language_loss": 0.59014541, + "learning_rate": 5.679986911537935e-07, + "loss": 0.61104435, + "num_input_tokens_seen": 136138905, + "step": 6332, + "time_per_iteration": 3.35371470451355 + }, + { + "auxiliary_loss_clip": 0.01189968, + "auxiliary_loss_mlp": 0.01025818, + "balance_loss_clip": 1.04016161, + "balance_loss_mlp": 1.01870167, + "epoch": 0.7614982264173631, + "flos": 35772522019200.0, + "grad_norm": 1.8213034062924733, + "language_loss": 0.6708374, + "learning_rate": 5.674549999825462e-07, + "loss": 0.69299531, + "num_input_tokens_seen": 136161720, + "step": 6333, + "time_per_iteration": 2.8359763622283936 + }, + { + "auxiliary_loss_clip": 0.01077718, + "auxiliary_loss_mlp": 0.01001933, + "balance_loss_clip": 1.01177657, + "balance_loss_mlp": 1.00081241, + "epoch": 0.7616184693080021, + "flos": 67925502345600.0, + "grad_norm": 0.9580555434001377, + "language_loss": 0.71360803, + "learning_rate": 5.669115261272363e-07, + "loss": 0.73440456, + "num_input_tokens_seen": 136222040, + "step": 6334, + "time_per_iteration": 3.1821322441101074 + }, + { + "auxiliary_loss_clip": 0.01176097, + "auxiliary_loss_mlp": 0.01024741, + "balance_loss_clip": 1.04533887, + "balance_loss_mlp": 1.01653945, + "epoch": 0.7617387121986413, + "flos": 20521979141760.0, + "grad_norm": 2.3785415266541845, + "language_loss": 0.72802222, + "learning_rate": 5.663682696703081e-07, + "loss": 0.75003052, + "num_input_tokens_seen": 136240305, + "step": 6335, + "time_per_iteration": 2.604335308074951 + }, + { + "auxiliary_loss_clip": 0.01169317, + "auxiliary_loss_mlp": 0.01027077, + "balance_loss_clip": 1.0486995, + "balance_loss_mlp": 1.0201807, + "epoch": 0.7618589550892804, + "flos": 18624495283200.0, + "grad_norm": 1.946077155268306, + "language_loss": 0.82099628, + "learning_rate": 5.658252306941746e-07, + "loss": 0.84296024, + "num_input_tokens_seen": 136259625, + "step": 6336, + "time_per_iteration": 2.5198445320129395 + }, + { + "auxiliary_loss_clip": 0.01208748, + "auxiliary_loss_mlp": 0.01025999, + "balance_loss_clip": 1.04458892, + "balance_loss_mlp": 1.01828599, + "epoch": 0.7619791979799194, + "flos": 17453735389440.0, + "grad_norm": 2.7988502057366182, + "language_loss": 0.7500056, + "learning_rate": 5.65282409281212e-07, + "loss": 0.77235305, + "num_input_tokens_seen": 136277090, + "step": 6337, + "time_per_iteration": 2.748253583908081 + }, + { + "auxiliary_loss_clip": 0.01180222, + "auxiliary_loss_mlp": 0.01027219, + "balance_loss_clip": 1.04260015, + "balance_loss_mlp": 1.01941037, + "epoch": 0.7620994408705585, + "flos": 14137421333760.0, + "grad_norm": 2.096833315449062, + "language_loss": 0.7004329, + "learning_rate": 5.64739805513768e-07, + "loss": 0.72250736, + "num_input_tokens_seen": 136294635, + "step": 6338, + "time_per_iteration": 2.640827178955078 + }, + { + "auxiliary_loss_clip": 0.01074598, + "auxiliary_loss_mlp": 0.01373523, + "balance_loss_clip": 1.01277828, + "balance_loss_mlp": 0.99959117, + "epoch": 0.7622196837611976, + "flos": 70708792527360.0, + "grad_norm": 0.7856149474236499, + "language_loss": 0.55724722, + "learning_rate": 5.641974194741541e-07, + "loss": 0.5817284, + "num_input_tokens_seen": 136350320, + "step": 6339, + "time_per_iteration": 3.1442346572875977 + }, + { + "auxiliary_loss_clip": 0.01094063, + "auxiliary_loss_mlp": 0.01001128, + "balance_loss_clip": 1.02550387, + "balance_loss_mlp": 1.00015688, + "epoch": 0.7623399266518367, + "flos": 60684150447360.0, + "grad_norm": 0.7828317524163455, + "language_loss": 0.63720286, + "learning_rate": 5.636552512446502e-07, + "loss": 0.65815473, + "num_input_tokens_seen": 136411375, + "step": 6340, + "time_per_iteration": 3.178037166595459 + }, + { + "auxiliary_loss_clip": 0.01173358, + "auxiliary_loss_mlp": 0.01024429, + "balance_loss_clip": 1.04576397, + "balance_loss_mlp": 1.01739264, + "epoch": 0.7624601695424758, + "flos": 26468893641600.0, + "grad_norm": 1.6411198111304868, + "language_loss": 0.77778077, + "learning_rate": 5.631133009075027e-07, + "loss": 0.79975855, + "num_input_tokens_seen": 136430560, + "step": 6341, + "time_per_iteration": 2.6070237159729004 + }, + { + "auxiliary_loss_clip": 0.01180547, + "auxiliary_loss_mlp": 0.01378471, + "balance_loss_clip": 1.0473758, + "balance_loss_mlp": 1.00017881, + "epoch": 0.7625804124331149, + "flos": 19135755515520.0, + "grad_norm": 4.752405062987361, + "language_loss": 0.68905604, + "learning_rate": 5.625715685449242e-07, + "loss": 0.71464622, + "num_input_tokens_seen": 136448665, + "step": 6342, + "time_per_iteration": 2.626126527786255 + }, + { + "auxiliary_loss_clip": 0.01195205, + "auxiliary_loss_mlp": 0.0102807, + "balance_loss_clip": 1.0468998, + "balance_loss_mlp": 1.02084553, + "epoch": 0.762700655323754, + "flos": 26213101914240.0, + "grad_norm": 1.5571487106684927, + "language_loss": 0.71759868, + "learning_rate": 5.620300542390966e-07, + "loss": 0.73983145, + "num_input_tokens_seen": 136469710, + "step": 6343, + "time_per_iteration": 2.710150718688965 + }, + { + "auxiliary_loss_clip": 0.01181324, + "auxiliary_loss_mlp": 0.01026887, + "balance_loss_clip": 1.0425303, + "balance_loss_mlp": 1.02013958, + "epoch": 0.762820898214393, + "flos": 22382582711040.0, + "grad_norm": 1.8829078444476735, + "language_loss": 0.85091758, + "learning_rate": 5.614887580721659e-07, + "loss": 0.87299967, + "num_input_tokens_seen": 136489855, + "step": 6344, + "time_per_iteration": 2.6722912788391113 + }, + { + "auxiliary_loss_clip": 0.01190683, + "auxiliary_loss_mlp": 0.01027199, + "balance_loss_clip": 1.04741728, + "balance_loss_mlp": 1.02003467, + "epoch": 0.7629411411050322, + "flos": 15700504550400.0, + "grad_norm": 2.131577612909581, + "language_loss": 0.73854029, + "learning_rate": 5.609476801262481e-07, + "loss": 0.76071912, + "num_input_tokens_seen": 136504715, + "step": 6345, + "time_per_iteration": 2.623641014099121 + }, + { + "auxiliary_loss_clip": 0.01189226, + "auxiliary_loss_mlp": 0.01030989, + "balance_loss_clip": 1.04303634, + "balance_loss_mlp": 1.02322221, + "epoch": 0.7630613839956712, + "flos": 13770342293760.0, + "grad_norm": 6.057011455874774, + "language_loss": 0.63850069, + "learning_rate": 5.604068204834223e-07, + "loss": 0.66070282, + "num_input_tokens_seen": 136521610, + "step": 6346, + "time_per_iteration": 2.7119293212890625 + }, + { + "auxiliary_loss_clip": 0.01201951, + "auxiliary_loss_mlp": 0.01378812, + "balance_loss_clip": 1.04271293, + "balance_loss_mlp": 1.00021148, + "epoch": 0.7631816268863103, + "flos": 14569569861120.0, + "grad_norm": 2.0762695205929442, + "language_loss": 0.76681954, + "learning_rate": 5.598661792257367e-07, + "loss": 0.79262722, + "num_input_tokens_seen": 136538655, + "step": 6347, + "time_per_iteration": 2.69242000579834 + }, + { + "auxiliary_loss_clip": 0.01174546, + "auxiliary_loss_mlp": 0.0102272, + "balance_loss_clip": 1.04450285, + "balance_loss_mlp": 1.01563859, + "epoch": 0.7633018697769495, + "flos": 19062210418560.0, + "grad_norm": 1.9334697223947888, + "language_loss": 0.75317258, + "learning_rate": 5.593257564352071e-07, + "loss": 0.77514517, + "num_input_tokens_seen": 136557095, + "step": 6348, + "time_per_iteration": 2.578584909439087 + }, + { + "auxiliary_loss_clip": 0.01172509, + "auxiliary_loss_mlp": 0.01021218, + "balance_loss_clip": 1.04416537, + "balance_loss_mlp": 1.01462555, + "epoch": 0.7634221126675885, + "flos": 22052958577920.0, + "grad_norm": 1.4999678151779716, + "language_loss": 0.75419664, + "learning_rate": 5.58785552193815e-07, + "loss": 0.77613389, + "num_input_tokens_seen": 136577340, + "step": 6349, + "time_per_iteration": 3.6222589015960693 + }, + { + "auxiliary_loss_clip": 0.011718, + "auxiliary_loss_mlp": 0.01023728, + "balance_loss_clip": 1.04981518, + "balance_loss_mlp": 1.01659071, + "epoch": 0.7635423555582276, + "flos": 29382720825600.0, + "grad_norm": 1.8059865316885388, + "language_loss": 0.75280625, + "learning_rate": 5.582455665835086e-07, + "loss": 0.77476156, + "num_input_tokens_seen": 136597635, + "step": 6350, + "time_per_iteration": 3.604640007019043 + }, + { + "auxiliary_loss_clip": 0.01192684, + "auxiliary_loss_mlp": 0.01028506, + "balance_loss_clip": 1.04322052, + "balance_loss_mlp": 1.02056634, + "epoch": 0.7636625984488667, + "flos": 17784903807360.0, + "grad_norm": 8.66927602002038, + "language_loss": 0.72636122, + "learning_rate": 5.577057996862036e-07, + "loss": 0.74857306, + "num_input_tokens_seen": 136615260, + "step": 6351, + "time_per_iteration": 2.6211671829223633 + }, + { + "auxiliary_loss_clip": 0.0116446, + "auxiliary_loss_mlp": 0.01025364, + "balance_loss_clip": 1.04587865, + "balance_loss_mlp": 1.01841354, + "epoch": 0.7637828413395058, + "flos": 23734583654400.0, + "grad_norm": 1.979576381807636, + "language_loss": 0.7617867, + "learning_rate": 5.571662515837814e-07, + "loss": 0.78368491, + "num_input_tokens_seen": 136637220, + "step": 6352, + "time_per_iteration": 2.647728681564331 + }, + { + "auxiliary_loss_clip": 0.01186964, + "auxiliary_loss_mlp": 0.01024728, + "balance_loss_clip": 1.04563987, + "balance_loss_mlp": 1.01763535, + "epoch": 0.7639030842301449, + "flos": 36283279461120.0, + "grad_norm": 1.7961556460685868, + "language_loss": 0.83588374, + "learning_rate": 5.566269223580926e-07, + "loss": 0.8580007, + "num_input_tokens_seen": 136658930, + "step": 6353, + "time_per_iteration": 2.776873826980591 + }, + { + "auxiliary_loss_clip": 0.01179576, + "auxiliary_loss_mlp": 0.01023837, + "balance_loss_clip": 1.04649222, + "balance_loss_mlp": 1.01723301, + "epoch": 0.764023327120784, + "flos": 28878104609280.0, + "grad_norm": 1.6313676914998063, + "language_loss": 0.75006044, + "learning_rate": 5.560878120909511e-07, + "loss": 0.77209461, + "num_input_tokens_seen": 136681530, + "step": 6354, + "time_per_iteration": 3.56532883644104 + }, + { + "auxiliary_loss_clip": 0.0107701, + "auxiliary_loss_mlp": 0.01002741, + "balance_loss_clip": 1.01092947, + "balance_loss_mlp": 1.00160277, + "epoch": 0.7641435700114231, + "flos": 64789711067520.0, + "grad_norm": 0.8440896659145414, + "language_loss": 0.58588821, + "learning_rate": 5.55548920864141e-07, + "loss": 0.6066857, + "num_input_tokens_seen": 136742185, + "step": 6355, + "time_per_iteration": 4.118502140045166 + }, + { + "auxiliary_loss_clip": 0.01178282, + "auxiliary_loss_mlp": 0.0102911, + "balance_loss_clip": 1.04900885, + "balance_loss_mlp": 1.02189469, + "epoch": 0.7642638129020621, + "flos": 16835784785280.0, + "grad_norm": 1.9731540844506898, + "language_loss": 0.7809152, + "learning_rate": 5.550102487594113e-07, + "loss": 0.80298913, + "num_input_tokens_seen": 136760855, + "step": 6356, + "time_per_iteration": 2.5816750526428223 + }, + { + "auxiliary_loss_clip": 0.01205279, + "auxiliary_loss_mlp": 0.01378749, + "balance_loss_clip": 1.03930604, + "balance_loss_mlp": 1.00026417, + "epoch": 0.7643840557927013, + "flos": 30408940391040.0, + "grad_norm": 2.3669586749682954, + "language_loss": 0.7144928, + "learning_rate": 5.54471795858477e-07, + "loss": 0.74033308, + "num_input_tokens_seen": 136780925, + "step": 6357, + "time_per_iteration": 2.8240439891815186 + }, + { + "auxiliary_loss_clip": 0.01191796, + "auxiliary_loss_mlp": 0.01025638, + "balance_loss_clip": 1.03969157, + "balance_loss_mlp": 1.01847386, + "epoch": 0.7645042986833404, + "flos": 16983234115200.0, + "grad_norm": 4.596030089290442, + "language_loss": 0.82755649, + "learning_rate": 5.539335622430235e-07, + "loss": 0.84973085, + "num_input_tokens_seen": 136799545, + "step": 6358, + "time_per_iteration": 2.6356513500213623 + }, + { + "auxiliary_loss_clip": 0.01169775, + "auxiliary_loss_mlp": 0.01020071, + "balance_loss_clip": 1.04352522, + "balance_loss_mlp": 1.01268947, + "epoch": 0.7646245415739794, + "flos": 17311493531520.0, + "grad_norm": 2.042012718320832, + "language_loss": 0.7471205, + "learning_rate": 5.533955479946975e-07, + "loss": 0.76901901, + "num_input_tokens_seen": 136818325, + "step": 6359, + "time_per_iteration": 2.6311604976654053 + }, + { + "auxiliary_loss_clip": 0.01106404, + "auxiliary_loss_mlp": 0.01373565, + "balance_loss_clip": 1.02428043, + "balance_loss_mlp": 0.99967301, + "epoch": 0.7647447844646186, + "flos": 70402332666240.0, + "grad_norm": 0.8576779344226081, + "language_loss": 0.6578629, + "learning_rate": 5.528577531951173e-07, + "loss": 0.68266261, + "num_input_tokens_seen": 136878730, + "step": 6360, + "time_per_iteration": 3.247246026992798 + }, + { + "auxiliary_loss_clip": 0.01189299, + "auxiliary_loss_mlp": 0.01027046, + "balance_loss_clip": 1.04604793, + "balance_loss_mlp": 1.0201261, + "epoch": 0.7648650273552576, + "flos": 17675914965120.0, + "grad_norm": 2.088672672148694, + "language_loss": 0.73967493, + "learning_rate": 5.523201779258653e-07, + "loss": 0.76183844, + "num_input_tokens_seen": 136897705, + "step": 6361, + "time_per_iteration": 2.6322827339172363 + }, + { + "auxiliary_loss_clip": 0.01165665, + "auxiliary_loss_mlp": 0.01025398, + "balance_loss_clip": 1.04483449, + "balance_loss_mlp": 1.01801288, + "epoch": 0.7649852702458967, + "flos": 22162019247360.0, + "grad_norm": 2.0505952816410242, + "language_loss": 0.83717656, + "learning_rate": 5.517828222684912e-07, + "loss": 0.85908711, + "num_input_tokens_seen": 136918360, + "step": 6362, + "time_per_iteration": 2.5350098609924316 + }, + { + "auxiliary_loss_clip": 0.01087643, + "auxiliary_loss_mlp": 0.01004348, + "balance_loss_clip": 1.01231885, + "balance_loss_mlp": 1.00326943, + "epoch": 0.7651055131365359, + "flos": 69848338227840.0, + "grad_norm": 0.7671274559025314, + "language_loss": 0.59002638, + "learning_rate": 5.512456863045117e-07, + "loss": 0.6109463, + "num_input_tokens_seen": 136979050, + "step": 6363, + "time_per_iteration": 3.2446398735046387 + }, + { + "auxiliary_loss_clip": 0.01168478, + "auxiliary_loss_mlp": 0.01026306, + "balance_loss_clip": 1.04571128, + "balance_loss_mlp": 1.0186832, + "epoch": 0.7652257560271749, + "flos": 19464014931840.0, + "grad_norm": 5.8648119532356615, + "language_loss": 0.74322605, + "learning_rate": 5.507087701154089e-07, + "loss": 0.76517391, + "num_input_tokens_seen": 136998970, + "step": 6364, + "time_per_iteration": 2.637091875076294 + }, + { + "auxiliary_loss_clip": 0.01207494, + "auxiliary_loss_mlp": 0.01024603, + "balance_loss_clip": 1.04372931, + "balance_loss_mlp": 1.01700306, + "epoch": 0.765345998917814, + "flos": 15961108700160.0, + "grad_norm": 2.7388769213812654, + "language_loss": 0.7511636, + "learning_rate": 5.50172073782634e-07, + "loss": 0.77348459, + "num_input_tokens_seen": 137016950, + "step": 6365, + "time_per_iteration": 2.6688759326934814 + }, + { + "auxiliary_loss_clip": 0.01195044, + "auxiliary_loss_mlp": 0.01027681, + "balance_loss_clip": 1.04733849, + "balance_loss_mlp": 1.02069187, + "epoch": 0.7654662418084531, + "flos": 23659853408640.0, + "grad_norm": 2.0225308070018024, + "language_loss": 0.87564969, + "learning_rate": 5.496355973876023e-07, + "loss": 0.89787698, + "num_input_tokens_seen": 137036205, + "step": 6366, + "time_per_iteration": 2.702148914337158 + }, + { + "auxiliary_loss_clip": 0.01188316, + "auxiliary_loss_mlp": 0.01379397, + "balance_loss_clip": 1.04108143, + "balance_loss_mlp": 1.00021076, + "epoch": 0.7655864846990922, + "flos": 41463608878080.0, + "grad_norm": 1.5960396644563883, + "language_loss": 0.7111817, + "learning_rate": 5.490993410116984e-07, + "loss": 0.73685884, + "num_input_tokens_seen": 137059195, + "step": 6367, + "time_per_iteration": 2.8186919689178467 + }, + { + "auxiliary_loss_clip": 0.0119029, + "auxiliary_loss_mlp": 0.01027477, + "balance_loss_clip": 1.04397273, + "balance_loss_mlp": 1.02012181, + "epoch": 0.7657067275897312, + "flos": 43142684088960.0, + "grad_norm": 2.014220578394043, + "language_loss": 0.69607949, + "learning_rate": 5.485633047362704e-07, + "loss": 0.71825719, + "num_input_tokens_seen": 137081200, + "step": 6368, + "time_per_iteration": 2.8796377182006836 + }, + { + "auxiliary_loss_clip": 0.0117364, + "auxiliary_loss_mlp": 0.01029576, + "balance_loss_clip": 1.05036783, + "balance_loss_mlp": 1.02166665, + "epoch": 0.7658269704803703, + "flos": 17311780840320.0, + "grad_norm": 3.67080357459913, + "language_loss": 0.79111457, + "learning_rate": 5.480274886426341e-07, + "loss": 0.81314677, + "num_input_tokens_seen": 137097840, + "step": 6369, + "time_per_iteration": 2.533541440963745 + }, + { + "auxiliary_loss_clip": 0.01171776, + "auxiliary_loss_mlp": 0.01023457, + "balance_loss_clip": 1.04589629, + "balance_loss_mlp": 1.01644766, + "epoch": 0.7659472133710095, + "flos": 12568160977920.0, + "grad_norm": 2.0105226479984344, + "language_loss": 0.77445155, + "learning_rate": 5.474918928120744e-07, + "loss": 0.79640388, + "num_input_tokens_seen": 137114335, + "step": 6370, + "time_per_iteration": 2.5813169479370117 + }, + { + "auxiliary_loss_clip": 0.0117562, + "auxiliary_loss_mlp": 0.01026234, + "balance_loss_clip": 1.0462656, + "balance_loss_mlp": 1.01929355, + "epoch": 0.7660674562616485, + "flos": 22707430335360.0, + "grad_norm": 1.732968653408408, + "language_loss": 0.87157661, + "learning_rate": 5.469565173258392e-07, + "loss": 0.8935951, + "num_input_tokens_seen": 137132850, + "step": 6371, + "time_per_iteration": 2.579063653945923 + }, + { + "auxiliary_loss_clip": 0.01170095, + "auxiliary_loss_mlp": 0.01024989, + "balance_loss_clip": 1.04576528, + "balance_loss_mlp": 1.01716006, + "epoch": 0.7661876991522876, + "flos": 17056455989760.0, + "grad_norm": 3.749585139640162, + "language_loss": 0.63697016, + "learning_rate": 5.464213622651454e-07, + "loss": 0.658921, + "num_input_tokens_seen": 137150665, + "step": 6372, + "time_per_iteration": 2.6454813480377197 + }, + { + "auxiliary_loss_clip": 0.01201832, + "auxiliary_loss_mlp": 0.01025717, + "balance_loss_clip": 1.04506671, + "balance_loss_mlp": 1.01808429, + "epoch": 0.7663079420429267, + "flos": 20084228092800.0, + "grad_norm": 1.6503669084873167, + "language_loss": 0.84109116, + "learning_rate": 5.458864277111753e-07, + "loss": 0.8633666, + "num_input_tokens_seen": 137168500, + "step": 6373, + "time_per_iteration": 2.640416145324707 + }, + { + "auxiliary_loss_clip": 0.01180144, + "auxiliary_loss_mlp": 0.01378332, + "balance_loss_clip": 1.04447258, + "balance_loss_mlp": 1.00014186, + "epoch": 0.7664281849335658, + "flos": 12677473042560.0, + "grad_norm": 2.477927064889199, + "language_loss": 0.69056427, + "learning_rate": 5.453517137450769e-07, + "loss": 0.71614903, + "num_input_tokens_seen": 137185075, + "step": 6374, + "time_per_iteration": 2.6734671592712402 + }, + { + "auxiliary_loss_clip": 0.01179371, + "auxiliary_loss_mlp": 0.01033802, + "balance_loss_clip": 1.0480932, + "balance_loss_mlp": 1.02620816, + "epoch": 0.7665484278242048, + "flos": 22345271458560.0, + "grad_norm": 1.5492850773808198, + "language_loss": 0.75982773, + "learning_rate": 5.448172204479684e-07, + "loss": 0.78195941, + "num_input_tokens_seen": 137204355, + "step": 6375, + "time_per_iteration": 3.7176425457000732 + }, + { + "auxiliary_loss_clip": 0.01165238, + "auxiliary_loss_mlp": 0.01021316, + "balance_loss_clip": 1.04535413, + "balance_loss_mlp": 1.01389551, + "epoch": 0.766668670714844, + "flos": 23617909301760.0, + "grad_norm": 2.0470781328330965, + "language_loss": 0.74593246, + "learning_rate": 5.442829479009294e-07, + "loss": 0.76779801, + "num_input_tokens_seen": 137223135, + "step": 6376, + "time_per_iteration": 3.4951012134552 + }, + { + "auxiliary_loss_clip": 0.01184389, + "auxiliary_loss_mlp": 0.0102352, + "balance_loss_clip": 1.04704356, + "balance_loss_mlp": 1.01592648, + "epoch": 0.7667889136054831, + "flos": 19427134642560.0, + "grad_norm": 2.0323023261029065, + "language_loss": 0.71611792, + "learning_rate": 5.437488961850103e-07, + "loss": 0.73819697, + "num_input_tokens_seen": 137242935, + "step": 6377, + "time_per_iteration": 2.610041618347168 + }, + { + "auxiliary_loss_clip": 0.01203469, + "auxiliary_loss_mlp": 0.01029467, + "balance_loss_clip": 1.04364979, + "balance_loss_mlp": 1.0227313, + "epoch": 0.7669091564961221, + "flos": 26866352609280.0, + "grad_norm": 1.7873101367638118, + "language_loss": 0.75625819, + "learning_rate": 5.432150653812258e-07, + "loss": 0.77858752, + "num_input_tokens_seen": 137262970, + "step": 6378, + "time_per_iteration": 2.7356605529785156 + }, + { + "auxiliary_loss_clip": 0.01176046, + "auxiliary_loss_mlp": 0.01023655, + "balance_loss_clip": 1.04725838, + "balance_loss_mlp": 1.01639509, + "epoch": 0.7670293993867613, + "flos": 12385303816320.0, + "grad_norm": 2.164511925608382, + "language_loss": 0.82910174, + "learning_rate": 5.42681455570557e-07, + "loss": 0.85109866, + "num_input_tokens_seen": 137279500, + "step": 6379, + "time_per_iteration": 2.589124917984009 + }, + { + "auxiliary_loss_clip": 0.01165811, + "auxiliary_loss_mlp": 0.01022779, + "balance_loss_clip": 1.04560733, + "balance_loss_mlp": 1.01577854, + "epoch": 0.7671496422774003, + "flos": 21762944167680.0, + "grad_norm": 1.702479121964883, + "language_loss": 0.64834905, + "learning_rate": 5.42148066833954e-07, + "loss": 0.67023498, + "num_input_tokens_seen": 137298745, + "step": 6380, + "time_per_iteration": 3.5126149654388428 + }, + { + "auxiliary_loss_clip": 0.01167371, + "auxiliary_loss_mlp": 0.0102496, + "balance_loss_clip": 1.04770505, + "balance_loss_mlp": 1.01780796, + "epoch": 0.7672698851680394, + "flos": 21069221823360.0, + "grad_norm": 2.0244912929199113, + "language_loss": 0.75444615, + "learning_rate": 5.416148992523289e-07, + "loss": 0.77636945, + "num_input_tokens_seen": 137317320, + "step": 6381, + "time_per_iteration": 3.4655966758728027 + }, + { + "auxiliary_loss_clip": 0.01221284, + "auxiliary_loss_mlp": 0.01021907, + "balance_loss_clip": 1.04054117, + "balance_loss_mlp": 1.0143286, + "epoch": 0.7673901280586786, + "flos": 16976697840000.0, + "grad_norm": 2.8093728491044145, + "language_loss": 0.78308016, + "learning_rate": 5.410819529065644e-07, + "loss": 0.80551201, + "num_input_tokens_seen": 137335275, + "step": 6382, + "time_per_iteration": 2.839989185333252 + }, + { + "auxiliary_loss_clip": 0.0120315, + "auxiliary_loss_mlp": 0.01030498, + "balance_loss_clip": 1.04049122, + "balance_loss_mlp": 1.02341747, + "epoch": 0.7675103709493176, + "flos": 29242669697280.0, + "grad_norm": 2.2683827701654358, + "language_loss": 0.65360391, + "learning_rate": 5.405492278775079e-07, + "loss": 0.67594039, + "num_input_tokens_seen": 137355055, + "step": 6383, + "time_per_iteration": 3.054229974746704 + }, + { + "auxiliary_loss_clip": 0.0118653, + "auxiliary_loss_mlp": 0.01025639, + "balance_loss_clip": 1.04447675, + "balance_loss_mlp": 1.01832867, + "epoch": 0.7676306138399567, + "flos": 29023004073600.0, + "grad_norm": 2.595974472321263, + "language_loss": 0.79402459, + "learning_rate": 5.400167242459732e-07, + "loss": 0.81614625, + "num_input_tokens_seen": 137374015, + "step": 6384, + "time_per_iteration": 2.6784260272979736 + }, + { + "auxiliary_loss_clip": 0.01174799, + "auxiliary_loss_mlp": 0.01022721, + "balance_loss_clip": 1.0446552, + "balance_loss_mlp": 1.01574087, + "epoch": 0.7677508567305958, + "flos": 22565116650240.0, + "grad_norm": 1.7746633582477762, + "language_loss": 0.8058517, + "learning_rate": 5.394844420927405e-07, + "loss": 0.82782686, + "num_input_tokens_seen": 137393625, + "step": 6385, + "time_per_iteration": 2.634852409362793 + }, + { + "auxiliary_loss_clip": 0.01167914, + "auxiliary_loss_mlp": 0.01027125, + "balance_loss_clip": 1.0483582, + "balance_loss_mlp": 1.01948369, + "epoch": 0.7678710996212349, + "flos": 25411432222080.0, + "grad_norm": 1.8717206700169358, + "language_loss": 0.73288894, + "learning_rate": 5.389523814985562e-07, + "loss": 0.7548393, + "num_input_tokens_seen": 137413045, + "step": 6386, + "time_per_iteration": 2.6097311973571777 + }, + { + "auxiliary_loss_clip": 0.01201355, + "auxiliary_loss_mlp": 0.01029965, + "balance_loss_clip": 1.04157996, + "balance_loss_mlp": 1.02234125, + "epoch": 0.767991342511874, + "flos": 26756825063040.0, + "grad_norm": 1.9278424739439792, + "language_loss": 0.76012087, + "learning_rate": 5.384205425441344e-07, + "loss": 0.78243405, + "num_input_tokens_seen": 137433955, + "step": 6387, + "time_per_iteration": 2.7590203285217285 + }, + { + "auxiliary_loss_clip": 0.01182773, + "auxiliary_loss_mlp": 0.01023566, + "balance_loss_clip": 1.04219651, + "balance_loss_mlp": 1.01699769, + "epoch": 0.7681115854025131, + "flos": 26359509749760.0, + "grad_norm": 1.909642741835079, + "language_loss": 0.84368479, + "learning_rate": 5.378889253101537e-07, + "loss": 0.86574817, + "num_input_tokens_seen": 137454510, + "step": 6388, + "time_per_iteration": 2.6816861629486084 + }, + { + "auxiliary_loss_clip": 0.0117748, + "auxiliary_loss_mlp": 0.01022641, + "balance_loss_clip": 1.04505026, + "balance_loss_mlp": 1.01561379, + "epoch": 0.7682318282931522, + "flos": 23257043314560.0, + "grad_norm": 1.768498257130576, + "language_loss": 0.80940616, + "learning_rate": 5.373575298772617e-07, + "loss": 0.83140743, + "num_input_tokens_seen": 137473630, + "step": 6389, + "time_per_iteration": 2.6468868255615234 + }, + { + "auxiliary_loss_clip": 0.01076128, + "auxiliary_loss_mlp": 0.01004646, + "balance_loss_clip": 1.01057458, + "balance_loss_mlp": 1.00357306, + "epoch": 0.7683520711837912, + "flos": 70072457137920.0, + "grad_norm": 0.7670127644689324, + "language_loss": 0.61307275, + "learning_rate": 5.368263563260689e-07, + "loss": 0.6338805, + "num_input_tokens_seen": 137538765, + "step": 6390, + "time_per_iteration": 3.2666876316070557 + }, + { + "auxiliary_loss_clip": 0.011734, + "auxiliary_loss_mlp": 0.01023511, + "balance_loss_clip": 1.04368448, + "balance_loss_mlp": 1.01645696, + "epoch": 0.7684723140744304, + "flos": 18624890332800.0, + "grad_norm": 1.5570784522053849, + "language_loss": 0.64393139, + "learning_rate": 5.362954047371537e-07, + "loss": 0.66590053, + "num_input_tokens_seen": 137557875, + "step": 6391, + "time_per_iteration": 2.598254919052124 + }, + { + "auxiliary_loss_clip": 0.01192835, + "auxiliary_loss_mlp": 0.01024172, + "balance_loss_clip": 1.04633081, + "balance_loss_mlp": 1.01622701, + "epoch": 0.7685925569650695, + "flos": 27452989532160.0, + "grad_norm": 2.6063208452851785, + "language_loss": 0.72004062, + "learning_rate": 5.357646751910627e-07, + "loss": 0.74221069, + "num_input_tokens_seen": 137579055, + "step": 6392, + "time_per_iteration": 2.739861011505127 + }, + { + "auxiliary_loss_clip": 0.01182123, + "auxiliary_loss_mlp": 0.01029038, + "balance_loss_clip": 1.04324877, + "balance_loss_mlp": 1.02162337, + "epoch": 0.7687127998557085, + "flos": 24535714642560.0, + "grad_norm": 2.2864927835263047, + "language_loss": 0.80173844, + "learning_rate": 5.352341677683061e-07, + "loss": 0.82385004, + "num_input_tokens_seen": 137600355, + "step": 6393, + "time_per_iteration": 2.650848150253296 + }, + { + "auxiliary_loss_clip": 0.01202354, + "auxiliary_loss_mlp": 0.01025479, + "balance_loss_clip": 1.04482603, + "balance_loss_mlp": 1.01899087, + "epoch": 0.7688330427463477, + "flos": 25155963717120.0, + "grad_norm": 1.8041523871589804, + "language_loss": 0.78640962, + "learning_rate": 5.347038825493617e-07, + "loss": 0.80868798, + "num_input_tokens_seen": 137621885, + "step": 6394, + "time_per_iteration": 2.776322841644287 + }, + { + "auxiliary_loss_clip": 0.0118072, + "auxiliary_loss_mlp": 0.01020378, + "balance_loss_clip": 1.04576683, + "balance_loss_mlp": 1.01368451, + "epoch": 0.7689532856369867, + "flos": 21211284113280.0, + "grad_norm": 2.0141531387008538, + "language_loss": 0.6869508, + "learning_rate": 5.341738196146732e-07, + "loss": 0.70896173, + "num_input_tokens_seen": 137640230, + "step": 6395, + "time_per_iteration": 2.6353507041931152 + }, + { + "auxiliary_loss_clip": 0.01172259, + "auxiliary_loss_mlp": 0.01025109, + "balance_loss_clip": 1.04440045, + "balance_loss_mlp": 1.017331, + "epoch": 0.7690735285276258, + "flos": 25119083427840.0, + "grad_norm": 2.059977757867639, + "language_loss": 0.73625183, + "learning_rate": 5.336439790446503e-07, + "loss": 0.75822556, + "num_input_tokens_seen": 137659330, + "step": 6396, + "time_per_iteration": 2.6459858417510986 + }, + { + "auxiliary_loss_clip": 0.01189284, + "auxiliary_loss_mlp": 0.01027205, + "balance_loss_clip": 1.04035056, + "balance_loss_mlp": 1.02011824, + "epoch": 0.769193771418265, + "flos": 54744020640000.0, + "grad_norm": 1.7979070664507566, + "language_loss": 0.62686002, + "learning_rate": 5.331143609196711e-07, + "loss": 0.64902484, + "num_input_tokens_seen": 137683145, + "step": 6397, + "time_per_iteration": 2.931028366088867 + }, + { + "auxiliary_loss_clip": 0.01178665, + "auxiliary_loss_mlp": 0.01026979, + "balance_loss_clip": 1.04861295, + "balance_loss_mlp": 1.0192306, + "epoch": 0.769314014308904, + "flos": 37341890115840.0, + "grad_norm": 1.8269323050503654, + "language_loss": 0.7722612, + "learning_rate": 5.325849653200758e-07, + "loss": 0.7943176, + "num_input_tokens_seen": 137707095, + "step": 6398, + "time_per_iteration": 2.7306337356567383 + }, + { + "auxiliary_loss_clip": 0.01168022, + "auxiliary_loss_mlp": 0.01024844, + "balance_loss_clip": 1.04768372, + "balance_loss_mlp": 1.01767397, + "epoch": 0.7694342571995431, + "flos": 20631686256000.0, + "grad_norm": 1.9937615536885507, + "language_loss": 0.76356614, + "learning_rate": 5.32055792326175e-07, + "loss": 0.78549474, + "num_input_tokens_seen": 137725520, + "step": 6399, + "time_per_iteration": 2.531083583831787 + }, + { + "auxiliary_loss_clip": 0.01189672, + "auxiliary_loss_mlp": 0.01028112, + "balance_loss_clip": 1.04709172, + "balance_loss_mlp": 1.02069426, + "epoch": 0.7695545000901821, + "flos": 24207706621440.0, + "grad_norm": 1.9607906368220338, + "language_loss": 0.72997606, + "learning_rate": 5.315268420182437e-07, + "loss": 0.75215393, + "num_input_tokens_seen": 137744195, + "step": 6400, + "time_per_iteration": 2.6851744651794434 + }, + { + "auxiliary_loss_clip": 0.01199626, + "auxiliary_loss_mlp": 0.01378519, + "balance_loss_clip": 1.04405534, + "balance_loss_mlp": 1.00021505, + "epoch": 0.7696747429808213, + "flos": 28001273708160.0, + "grad_norm": 2.474014300066561, + "language_loss": 0.76450491, + "learning_rate": 5.309981144765221e-07, + "loss": 0.79028636, + "num_input_tokens_seen": 137764340, + "step": 6401, + "time_per_iteration": 3.7105984687805176 + }, + { + "auxiliary_loss_clip": 0.01206764, + "auxiliary_loss_mlp": 0.01026716, + "balance_loss_clip": 1.03987479, + "balance_loss_mlp": 1.01941705, + "epoch": 0.7697949858714603, + "flos": 11509550323200.0, + "grad_norm": 3.353409686323595, + "language_loss": 0.75299269, + "learning_rate": 5.304696097812196e-07, + "loss": 0.7753275, + "num_input_tokens_seen": 137780940, + "step": 6402, + "time_per_iteration": 3.6073992252349854 + }, + { + "auxiliary_loss_clip": 0.01180872, + "auxiliary_loss_mlp": 0.01029967, + "balance_loss_clip": 1.04314494, + "balance_loss_mlp": 1.02179503, + "epoch": 0.7699152287620994, + "flos": 26688271956480.0, + "grad_norm": 2.857769862026416, + "language_loss": 0.60169053, + "learning_rate": 5.299413280125078e-07, + "loss": 0.62379897, + "num_input_tokens_seen": 137799250, + "step": 6403, + "time_per_iteration": 2.701359510421753 + }, + { + "auxiliary_loss_clip": 0.01184546, + "auxiliary_loss_mlp": 0.01027196, + "balance_loss_clip": 1.04335177, + "balance_loss_mlp": 1.01929235, + "epoch": 0.7700354716527386, + "flos": 16544944362240.0, + "grad_norm": 2.002160670501649, + "language_loss": 0.73021412, + "learning_rate": 5.294132692505284e-07, + "loss": 0.7523315, + "num_input_tokens_seen": 137817660, + "step": 6404, + "time_per_iteration": 2.606994867324829 + }, + { + "auxiliary_loss_clip": 0.01192689, + "auxiliary_loss_mlp": 0.01025587, + "balance_loss_clip": 1.0410533, + "balance_loss_mlp": 1.01813674, + "epoch": 0.7701557145433776, + "flos": 19242733196160.0, + "grad_norm": 1.8981031608601493, + "language_loss": 0.79001909, + "learning_rate": 5.288854335753861e-07, + "loss": 0.81220186, + "num_input_tokens_seen": 137835920, + "step": 6405, + "time_per_iteration": 3.692300796508789 + }, + { + "auxiliary_loss_clip": 0.01180369, + "auxiliary_loss_mlp": 0.01022549, + "balance_loss_clip": 1.04609668, + "balance_loss_mlp": 1.01503277, + "epoch": 0.7702759574340167, + "flos": 31685744211840.0, + "grad_norm": 1.9767693922825293, + "language_loss": 0.75835896, + "learning_rate": 5.283578210671551e-07, + "loss": 0.78038812, + "num_input_tokens_seen": 137858160, + "step": 6406, + "time_per_iteration": 2.767953634262085 + }, + { + "auxiliary_loss_clip": 0.01187913, + "auxiliary_loss_mlp": 0.01026708, + "balance_loss_clip": 1.04438937, + "balance_loss_mlp": 1.01921248, + "epoch": 0.7703962003246558, + "flos": 16800089644800.0, + "grad_norm": 2.0749907220056745, + "language_loss": 0.76685786, + "learning_rate": 5.278304318058719e-07, + "loss": 0.78900409, + "num_input_tokens_seen": 137876015, + "step": 6407, + "time_per_iteration": 3.550977945327759 + }, + { + "auxiliary_loss_clip": 0.01208353, + "auxiliary_loss_mlp": 0.01024042, + "balance_loss_clip": 1.03945565, + "balance_loss_mlp": 1.0170536, + "epoch": 0.7705164432152949, + "flos": 35736072693120.0, + "grad_norm": 1.7801410753792202, + "language_loss": 0.78969371, + "learning_rate": 5.273032658715411e-07, + "loss": 0.81201768, + "num_input_tokens_seen": 137898825, + "step": 6408, + "time_per_iteration": 2.903925657272339 + }, + { + "auxiliary_loss_clip": 0.01198104, + "auxiliary_loss_mlp": 0.01026345, + "balance_loss_clip": 1.04056811, + "balance_loss_mlp": 1.01832199, + "epoch": 0.7706366861059339, + "flos": 23365960329600.0, + "grad_norm": 2.2005557310934107, + "language_loss": 0.76527011, + "learning_rate": 5.267763233441347e-07, + "loss": 0.78751457, + "num_input_tokens_seen": 137919455, + "step": 6409, + "time_per_iteration": 2.797882080078125 + }, + { + "auxiliary_loss_clip": 0.01180454, + "auxiliary_loss_mlp": 0.01024703, + "balance_loss_clip": 1.04663992, + "balance_loss_mlp": 1.01659667, + "epoch": 0.7707569289965731, + "flos": 22929897219840.0, + "grad_norm": 2.4649634913662863, + "language_loss": 0.69757849, + "learning_rate": 5.26249604303588e-07, + "loss": 0.71963006, + "num_input_tokens_seen": 137937960, + "step": 6410, + "time_per_iteration": 2.6527650356292725 + }, + { + "auxiliary_loss_clip": 0.01170154, + "auxiliary_loss_mlp": 0.01028475, + "balance_loss_clip": 1.04913759, + "balance_loss_mlp": 1.02108979, + "epoch": 0.7708771718872122, + "flos": 17420661941760.0, + "grad_norm": 2.0401427566252477, + "language_loss": 0.78656071, + "learning_rate": 5.257231088298057e-07, + "loss": 0.80854702, + "num_input_tokens_seen": 137956370, + "step": 6411, + "time_per_iteration": 2.568486452102661 + }, + { + "auxiliary_loss_clip": 0.01097054, + "auxiliary_loss_mlp": 0.0100107, + "balance_loss_clip": 1.0097971, + "balance_loss_mlp": 0.99990815, + "epoch": 0.7709974147778512, + "flos": 72241316248320.0, + "grad_norm": 0.7965059318039808, + "language_loss": 0.53957963, + "learning_rate": 5.25196837002655e-07, + "loss": 0.56056088, + "num_input_tokens_seen": 138016080, + "step": 6412, + "time_per_iteration": 3.324596405029297 + }, + { + "auxiliary_loss_clip": 0.01181363, + "auxiliary_loss_mlp": 0.01026299, + "balance_loss_clip": 1.04399824, + "balance_loss_mlp": 1.01888394, + "epoch": 0.7711176576684904, + "flos": 39859694876160.0, + "grad_norm": 1.86829637905562, + "language_loss": 0.6845637, + "learning_rate": 5.24670788901971e-07, + "loss": 0.70664036, + "num_input_tokens_seen": 138039170, + "step": 6413, + "time_per_iteration": 2.8090531826019287 + }, + { + "auxiliary_loss_clip": 0.01184022, + "auxiliary_loss_mlp": 0.01023997, + "balance_loss_clip": 1.04632545, + "balance_loss_mlp": 1.01612306, + "epoch": 0.7712379005591294, + "flos": 36976391274240.0, + "grad_norm": 2.5322758568312635, + "language_loss": 0.68960774, + "learning_rate": 5.241449646075557e-07, + "loss": 0.71168786, + "num_input_tokens_seen": 138062395, + "step": 6414, + "time_per_iteration": 2.7265431880950928 + }, + { + "auxiliary_loss_clip": 0.01183825, + "auxiliary_loss_mlp": 0.01024836, + "balance_loss_clip": 1.04638815, + "balance_loss_mlp": 1.01729012, + "epoch": 0.7713581434497685, + "flos": 22776773541120.0, + "grad_norm": 1.9462688024859476, + "language_loss": 0.72163558, + "learning_rate": 5.236193641991762e-07, + "loss": 0.74372226, + "num_input_tokens_seen": 138080325, + "step": 6415, + "time_per_iteration": 2.690268039703369 + }, + { + "auxiliary_loss_clip": 0.01183005, + "auxiliary_loss_mlp": 0.01025828, + "balance_loss_clip": 1.04577422, + "balance_loss_mlp": 1.01843715, + "epoch": 0.7714783863404077, + "flos": 24097460803200.0, + "grad_norm": 2.20156644702229, + "language_loss": 0.697245, + "learning_rate": 5.23093987756565e-07, + "loss": 0.71933335, + "num_input_tokens_seen": 138099020, + "step": 6416, + "time_per_iteration": 2.642383098602295 + }, + { + "auxiliary_loss_clip": 0.01198157, + "auxiliary_loss_mlp": 0.01028209, + "balance_loss_clip": 1.0405103, + "balance_loss_mlp": 1.02063024, + "epoch": 0.7715986292310467, + "flos": 21063655215360.0, + "grad_norm": 2.229292878170313, + "language_loss": 0.75321317, + "learning_rate": 5.225688353594217e-07, + "loss": 0.77547681, + "num_input_tokens_seen": 138118650, + "step": 6417, + "time_per_iteration": 2.70760178565979 + }, + { + "auxiliary_loss_clip": 0.01191434, + "auxiliary_loss_mlp": 0.01378403, + "balance_loss_clip": 1.04629707, + "balance_loss_mlp": 1.00019717, + "epoch": 0.7717188721216858, + "flos": 20594877793920.0, + "grad_norm": 2.2501971215917433, + "language_loss": 0.77433705, + "learning_rate": 5.220439070874108e-07, + "loss": 0.80003548, + "num_input_tokens_seen": 138137890, + "step": 6418, + "time_per_iteration": 2.634140968322754 + }, + { + "auxiliary_loss_clip": 0.01178534, + "auxiliary_loss_mlp": 0.01026091, + "balance_loss_clip": 1.04951096, + "balance_loss_mlp": 1.019171, + "epoch": 0.7718391150123249, + "flos": 26250951870720.0, + "grad_norm": 1.6755991045347505, + "language_loss": 0.71010303, + "learning_rate": 5.215192030201652e-07, + "loss": 0.73214924, + "num_input_tokens_seen": 138158880, + "step": 6419, + "time_per_iteration": 2.672109603881836 + }, + { + "auxiliary_loss_clip": 0.01177129, + "auxiliary_loss_mlp": 0.01024415, + "balance_loss_clip": 1.03757894, + "balance_loss_mlp": 1.0170989, + "epoch": 0.771959357902964, + "flos": 22049762267520.0, + "grad_norm": 1.8435393833958806, + "language_loss": 0.86170113, + "learning_rate": 5.209947232372798e-07, + "loss": 0.88371652, + "num_input_tokens_seen": 138176370, + "step": 6420, + "time_per_iteration": 2.753648042678833 + }, + { + "auxiliary_loss_clip": 0.01181236, + "auxiliary_loss_mlp": 0.01378815, + "balance_loss_clip": 1.04700065, + "balance_loss_mlp": 1.00022459, + "epoch": 0.772079600793603, + "flos": 30446000248320.0, + "grad_norm": 1.7385200654503479, + "language_loss": 0.81194305, + "learning_rate": 5.204704678183196e-07, + "loss": 0.83754361, + "num_input_tokens_seen": 138195105, + "step": 6421, + "time_per_iteration": 2.686170816421509 + }, + { + "auxiliary_loss_clip": 0.01168315, + "auxiliary_loss_mlp": 0.01024623, + "balance_loss_clip": 1.04703712, + "balance_loss_mlp": 1.01708269, + "epoch": 0.7721998436842422, + "flos": 12969857750400.0, + "grad_norm": 1.958554084470307, + "language_loss": 0.85029405, + "learning_rate": 5.19946436842813e-07, + "loss": 0.87222344, + "num_input_tokens_seen": 138212235, + "step": 6422, + "time_per_iteration": 2.5690646171569824 + }, + { + "auxiliary_loss_clip": 0.01194336, + "auxiliary_loss_mlp": 0.01021185, + "balance_loss_clip": 1.04313326, + "balance_loss_mlp": 1.01412201, + "epoch": 0.7723200865748813, + "flos": 32635509678720.0, + "grad_norm": 1.7572020357875568, + "language_loss": 0.6790421, + "learning_rate": 5.194226303902546e-07, + "loss": 0.70119727, + "num_input_tokens_seen": 138231970, + "step": 6423, + "time_per_iteration": 2.7333996295928955 + }, + { + "auxiliary_loss_clip": 0.01180322, + "auxiliary_loss_mlp": 0.01023805, + "balance_loss_clip": 1.04381323, + "balance_loss_mlp": 1.0167985, + "epoch": 0.7724403294655203, + "flos": 21105707063040.0, + "grad_norm": 1.8035760102747342, + "language_loss": 0.70604819, + "learning_rate": 5.188990485401072e-07, + "loss": 0.72808945, + "num_input_tokens_seen": 138251175, + "step": 6424, + "time_per_iteration": 2.7303125858306885 + }, + { + "auxiliary_loss_clip": 0.0117939, + "auxiliary_loss_mlp": 0.01030273, + "balance_loss_clip": 1.04670143, + "balance_loss_mlp": 1.02265906, + "epoch": 0.7725605723561595, + "flos": 22090736707200.0, + "grad_norm": 1.8268402231846115, + "language_loss": 0.86426401, + "learning_rate": 5.183756913717954e-07, + "loss": 0.88636065, + "num_input_tokens_seen": 138270950, + "step": 6425, + "time_per_iteration": 2.5741100311279297 + }, + { + "auxiliary_loss_clip": 0.0118426, + "auxiliary_loss_mlp": 0.01024251, + "balance_loss_clip": 1.04590547, + "balance_loss_mlp": 1.01727152, + "epoch": 0.7726808152467985, + "flos": 34495610457600.0, + "grad_norm": 1.7087495689573466, + "language_loss": 0.73156714, + "learning_rate": 5.178525589647136e-07, + "loss": 0.75365222, + "num_input_tokens_seen": 138292590, + "step": 6426, + "time_per_iteration": 2.730556011199951 + }, + { + "auxiliary_loss_clip": 0.01190041, + "auxiliary_loss_mlp": 0.01020224, + "balance_loss_clip": 1.0449214, + "balance_loss_mlp": 1.0133698, + "epoch": 0.7728010581374376, + "flos": 22306344094080.0, + "grad_norm": 1.9109916815776473, + "language_loss": 0.79032278, + "learning_rate": 5.173296513982197e-07, + "loss": 0.81242543, + "num_input_tokens_seen": 138311115, + "step": 6427, + "time_per_iteration": 3.584026575088501 + }, + { + "auxiliary_loss_clip": 0.01206077, + "auxiliary_loss_mlp": 0.01028212, + "balance_loss_clip": 1.04697561, + "balance_loss_mlp": 1.02005541, + "epoch": 0.7729213010280768, + "flos": 27126453968640.0, + "grad_norm": 2.5925578464676198, + "language_loss": 0.64870256, + "learning_rate": 5.168069687516398e-07, + "loss": 0.67104542, + "num_input_tokens_seen": 138330885, + "step": 6428, + "time_per_iteration": 3.5870003700256348 + }, + { + "auxiliary_loss_clip": 0.01188025, + "auxiliary_loss_mlp": 0.01024648, + "balance_loss_clip": 1.04736674, + "balance_loss_mlp": 1.01726043, + "epoch": 0.7730415439187158, + "flos": 18150223080960.0, + "grad_norm": 3.5390539048984815, + "language_loss": 0.71500719, + "learning_rate": 5.16284511104263e-07, + "loss": 0.73713392, + "num_input_tokens_seen": 138350020, + "step": 6429, + "time_per_iteration": 2.628403663635254 + }, + { + "auxiliary_loss_clip": 0.01185928, + "auxiliary_loss_mlp": 0.01032722, + "balance_loss_clip": 1.04577684, + "balance_loss_mlp": 1.025033, + "epoch": 0.7731617868093549, + "flos": 11947480940160.0, + "grad_norm": 2.463150593524682, + "language_loss": 0.81053448, + "learning_rate": 5.157622785353457e-07, + "loss": 0.83272099, + "num_input_tokens_seen": 138368135, + "step": 6430, + "time_per_iteration": 2.553612470626831 + }, + { + "auxiliary_loss_clip": 0.01076599, + "auxiliary_loss_mlp": 0.01000037, + "balance_loss_clip": 1.010167, + "balance_loss_mlp": 0.99896443, + "epoch": 0.7732820296999939, + "flos": 64201027069440.0, + "grad_norm": 0.6458447312383326, + "language_loss": 0.60392529, + "learning_rate": 5.152402711241113e-07, + "loss": 0.62469167, + "num_input_tokens_seen": 138436040, + "step": 6431, + "time_per_iteration": 4.1866912841796875 + }, + { + "auxiliary_loss_clip": 0.0119001, + "auxiliary_loss_mlp": 0.01027939, + "balance_loss_clip": 1.04158759, + "balance_loss_mlp": 1.02108431, + "epoch": 0.7734022725906331, + "flos": 25302191984640.0, + "grad_norm": 1.870925108916037, + "language_loss": 0.82973164, + "learning_rate": 5.147184889497465e-07, + "loss": 0.85191119, + "num_input_tokens_seen": 138455510, + "step": 6432, + "time_per_iteration": 2.7210216522216797 + }, + { + "auxiliary_loss_clip": 0.01183408, + "auxiliary_loss_mlp": 0.01023543, + "balance_loss_clip": 1.04237652, + "balance_loss_mlp": 1.01571703, + "epoch": 0.7735225154812722, + "flos": 17347440067200.0, + "grad_norm": 2.8801497345668983, + "language_loss": 0.79768121, + "learning_rate": 5.141969320914072e-07, + "loss": 0.81975079, + "num_input_tokens_seen": 138473015, + "step": 6433, + "time_per_iteration": 3.5388107299804688 + }, + { + "auxiliary_loss_clip": 0.01169807, + "auxiliary_loss_mlp": 0.01029841, + "balance_loss_clip": 1.04571998, + "balance_loss_mlp": 1.02215481, + "epoch": 0.7736427583719112, + "flos": 32630086725120.0, + "grad_norm": 2.5434831954794648, + "language_loss": 0.62528408, + "learning_rate": 5.136756006282113e-07, + "loss": 0.64728051, + "num_input_tokens_seen": 138491680, + "step": 6434, + "time_per_iteration": 2.6957764625549316 + }, + { + "auxiliary_loss_clip": 0.01170384, + "auxiliary_loss_mlp": 0.01028568, + "balance_loss_clip": 1.04820871, + "balance_loss_mlp": 1.02115929, + "epoch": 0.7737630012625504, + "flos": 19860073269120.0, + "grad_norm": 2.640312359172627, + "language_loss": 0.84725904, + "learning_rate": 5.131544946392446e-07, + "loss": 0.86924857, + "num_input_tokens_seen": 138506960, + "step": 6435, + "time_per_iteration": 2.5481488704681396 + }, + { + "auxiliary_loss_clip": 0.01188142, + "auxiliary_loss_mlp": 0.0102722, + "balance_loss_clip": 1.04752851, + "balance_loss_mlp": 1.0193404, + "epoch": 0.7738832441531894, + "flos": 36022639397760.0, + "grad_norm": 2.111614721889553, + "language_loss": 0.6377781, + "learning_rate": 5.126336142035592e-07, + "loss": 0.65993178, + "num_input_tokens_seen": 138526995, + "step": 6436, + "time_per_iteration": 2.7755680084228516 + }, + { + "auxiliary_loss_clip": 0.0118507, + "auxiliary_loss_mlp": 0.01022988, + "balance_loss_clip": 1.043715, + "balance_loss_mlp": 1.01585889, + "epoch": 0.7740034870438285, + "flos": 13405274415360.0, + "grad_norm": 2.570924013777186, + "language_loss": 0.72307718, + "learning_rate": 5.121129594001721e-07, + "loss": 0.74515778, + "num_input_tokens_seen": 138541260, + "step": 6437, + "time_per_iteration": 2.6865744590759277 + }, + { + "auxiliary_loss_clip": 0.0117711, + "auxiliary_loss_mlp": 0.01027735, + "balance_loss_clip": 1.04869282, + "balance_loss_mlp": 1.02032304, + "epoch": 0.7741237299344677, + "flos": 22086714384000.0, + "grad_norm": 1.5678965565672438, + "language_loss": 0.81411004, + "learning_rate": 5.115925303080661e-07, + "loss": 0.83615851, + "num_input_tokens_seen": 138560970, + "step": 6438, + "time_per_iteration": 2.577410936355591 + }, + { + "auxiliary_loss_clip": 0.01188459, + "auxiliary_loss_mlp": 0.01024566, + "balance_loss_clip": 1.04501939, + "balance_loss_mlp": 1.0171572, + "epoch": 0.7742439728251067, + "flos": 19864777950720.0, + "grad_norm": 2.041430195700128, + "language_loss": 0.79355264, + "learning_rate": 5.110723270061899e-07, + "loss": 0.81568289, + "num_input_tokens_seen": 138577460, + "step": 6439, + "time_per_iteration": 2.577070713043213 + }, + { + "auxiliary_loss_clip": 0.01164281, + "auxiliary_loss_mlp": 0.01025625, + "balance_loss_clip": 1.04537106, + "balance_loss_mlp": 1.01838875, + "epoch": 0.7743642157157458, + "flos": 16690167048960.0, + "grad_norm": 1.9213504019513175, + "language_loss": 0.79539144, + "learning_rate": 5.105523495734572e-07, + "loss": 0.81729054, + "num_input_tokens_seen": 138594860, + "step": 6440, + "time_per_iteration": 2.5483133792877197 + }, + { + "auxiliary_loss_clip": 0.01170428, + "auxiliary_loss_mlp": 0.01026021, + "balance_loss_clip": 1.0488131, + "balance_loss_mlp": 1.01869845, + "epoch": 0.7744844586063849, + "flos": 20304360593280.0, + "grad_norm": 2.1389665616302933, + "language_loss": 0.75500256, + "learning_rate": 5.100325980887499e-07, + "loss": 0.77696711, + "num_input_tokens_seen": 138614785, + "step": 6441, + "time_per_iteration": 2.5254738330841064 + }, + { + "auxiliary_loss_clip": 0.01194662, + "auxiliary_loss_mlp": 0.01021583, + "balance_loss_clip": 1.04515171, + "balance_loss_mlp": 1.01462162, + "epoch": 0.774604701497024, + "flos": 22966705681920.0, + "grad_norm": 1.7238355462053192, + "language_loss": 0.83126837, + "learning_rate": 5.095130726309116e-07, + "loss": 0.85343075, + "num_input_tokens_seen": 138634960, + "step": 6442, + "time_per_iteration": 2.7221930027008057 + }, + { + "auxiliary_loss_clip": 0.010627, + "auxiliary_loss_mlp": 0.01003389, + "balance_loss_clip": 1.01040816, + "balance_loss_mlp": 1.00235796, + "epoch": 0.774724944387663, + "flos": 60288523073280.0, + "grad_norm": 0.7914334270571379, + "language_loss": 0.58985579, + "learning_rate": 5.089937732787559e-07, + "loss": 0.61051673, + "num_input_tokens_seen": 138699520, + "step": 6443, + "time_per_iteration": 3.2169687747955322 + }, + { + "auxiliary_loss_clip": 0.01194081, + "auxiliary_loss_mlp": 0.01030018, + "balance_loss_clip": 1.04355073, + "balance_loss_mlp": 1.02286267, + "epoch": 0.7748451872783022, + "flos": 26761026954240.0, + "grad_norm": 2.674495281205773, + "language_loss": 0.66450393, + "learning_rate": 5.084747001110592e-07, + "loss": 0.68674493, + "num_input_tokens_seen": 138719145, + "step": 6444, + "time_per_iteration": 2.6991541385650635 + }, + { + "auxiliary_loss_clip": 0.01176217, + "auxiliary_loss_mlp": 0.01378594, + "balance_loss_clip": 1.04947948, + "balance_loss_mlp": 1.00020504, + "epoch": 0.7749654301689413, + "flos": 30338627518080.0, + "grad_norm": 2.0010873352148923, + "language_loss": 0.70100939, + "learning_rate": 5.07955853206564e-07, + "loss": 0.72655755, + "num_input_tokens_seen": 138743850, + "step": 6445, + "time_per_iteration": 2.697298288345337 + }, + { + "auxiliary_loss_clip": 0.01183576, + "auxiliary_loss_mlp": 0.01023817, + "balance_loss_clip": 1.04759443, + "balance_loss_mlp": 1.01683736, + "epoch": 0.7750856730595803, + "flos": 43179851687040.0, + "grad_norm": 1.483467980444307, + "language_loss": 0.71003211, + "learning_rate": 5.074372326439807e-07, + "loss": 0.73210597, + "num_input_tokens_seen": 138766860, + "step": 6446, + "time_per_iteration": 2.815563201904297 + }, + { + "auxiliary_loss_clip": 0.01192762, + "auxiliary_loss_mlp": 0.01023412, + "balance_loss_clip": 1.04213583, + "balance_loss_mlp": 1.01639044, + "epoch": 0.7752059159502195, + "flos": 17640040256640.0, + "grad_norm": 2.147959862075581, + "language_loss": 0.73705214, + "learning_rate": 5.069188385019814e-07, + "loss": 0.75921392, + "num_input_tokens_seen": 138784560, + "step": 6447, + "time_per_iteration": 2.725944995880127 + }, + { + "auxiliary_loss_clip": 0.01207697, + "auxiliary_loss_mlp": 0.01027039, + "balance_loss_clip": 1.03968549, + "balance_loss_mlp": 1.01966047, + "epoch": 0.7753261588408585, + "flos": 12677688524160.0, + "grad_norm": 3.283718197260648, + "language_loss": 0.61761242, + "learning_rate": 5.064006708592077e-07, + "loss": 0.63995981, + "num_input_tokens_seen": 138800805, + "step": 6448, + "time_per_iteration": 2.719548225402832 + }, + { + "auxiliary_loss_clip": 0.01179381, + "auxiliary_loss_mlp": 0.01025096, + "balance_loss_clip": 1.04595327, + "balance_loss_mlp": 1.01848912, + "epoch": 0.7754464017314976, + "flos": 16690741666560.0, + "grad_norm": 2.456681199643425, + "language_loss": 0.75739652, + "learning_rate": 5.058827297942641e-07, + "loss": 0.77944136, + "num_input_tokens_seen": 138815910, + "step": 6449, + "time_per_iteration": 2.757769823074341 + }, + { + "auxiliary_loss_clip": 0.01193058, + "auxiliary_loss_mlp": 0.01024237, + "balance_loss_clip": 1.04619157, + "balance_loss_mlp": 1.01746011, + "epoch": 0.7755666446221368, + "flos": 19718944732800.0, + "grad_norm": 2.383125772732014, + "language_loss": 0.74880457, + "learning_rate": 5.053650153857237e-07, + "loss": 0.77097762, + "num_input_tokens_seen": 138834920, + "step": 6450, + "time_per_iteration": 2.678720235824585 + }, + { + "auxiliary_loss_clip": 0.01174327, + "auxiliary_loss_mlp": 0.01030471, + "balance_loss_clip": 1.04507565, + "balance_loss_mlp": 1.02362239, + "epoch": 0.7756868875127758, + "flos": 18693623007360.0, + "grad_norm": 1.70482612782787, + "language_loss": 0.69948846, + "learning_rate": 5.048475277121214e-07, + "loss": 0.72153646, + "num_input_tokens_seen": 138852135, + "step": 6451, + "time_per_iteration": 2.618731737136841 + }, + { + "auxiliary_loss_clip": 0.0117911, + "auxiliary_loss_mlp": 0.01026587, + "balance_loss_clip": 1.04703808, + "balance_loss_mlp": 1.01892734, + "epoch": 0.7758071304034149, + "flos": 28404191543040.0, + "grad_norm": 1.8254550801408835, + "language_loss": 0.77062333, + "learning_rate": 5.043302668519598e-07, + "loss": 0.79268026, + "num_input_tokens_seen": 138871470, + "step": 6452, + "time_per_iteration": 2.659813642501831 + }, + { + "auxiliary_loss_clip": 0.01179736, + "auxiliary_loss_mlp": 0.01023567, + "balance_loss_clip": 1.04558301, + "balance_loss_mlp": 1.01672459, + "epoch": 0.775927373294054, + "flos": 20595344670720.0, + "grad_norm": 2.3380137104037155, + "language_loss": 0.72048396, + "learning_rate": 5.038132328837079e-07, + "loss": 0.74251699, + "num_input_tokens_seen": 138889860, + "step": 6453, + "time_per_iteration": 3.498671293258667 + }, + { + "auxiliary_loss_clip": 0.01178416, + "auxiliary_loss_mlp": 0.01023825, + "balance_loss_clip": 1.04643488, + "balance_loss_mlp": 1.01732469, + "epoch": 0.7760476161846931, + "flos": 22526368853760.0, + "grad_norm": 2.0374817606383004, + "language_loss": 0.73633385, + "learning_rate": 5.032964258857993e-07, + "loss": 0.75835621, + "num_input_tokens_seen": 138909955, + "step": 6454, + "time_per_iteration": 3.6374599933624268 + }, + { + "auxiliary_loss_clip": 0.01173659, + "auxiliary_loss_mlp": 0.01024201, + "balance_loss_clip": 1.04171681, + "balance_loss_mlp": 1.01732564, + "epoch": 0.7761678590753321, + "flos": 48651488403840.0, + "grad_norm": 1.4982982810015972, + "language_loss": 0.68456411, + "learning_rate": 5.027798459366329e-07, + "loss": 0.70654273, + "num_input_tokens_seen": 138935320, + "step": 6455, + "time_per_iteration": 2.832771062850952 + }, + { + "auxiliary_loss_clip": 0.01179041, + "auxiliary_loss_mlp": 0.01023969, + "balance_loss_clip": 1.04584455, + "balance_loss_mlp": 1.01651871, + "epoch": 0.7762881019659713, + "flos": 26177047637760.0, + "grad_norm": 1.4074921783934846, + "language_loss": 0.63456142, + "learning_rate": 5.02263493114573e-07, + "loss": 0.65659142, + "num_input_tokens_seen": 138957115, + "step": 6456, + "time_per_iteration": 2.59302020072937 + }, + { + "auxiliary_loss_clip": 0.01167425, + "auxiliary_loss_mlp": 0.01024793, + "balance_loss_clip": 1.04656863, + "balance_loss_mlp": 1.01710963, + "epoch": 0.7764083448566104, + "flos": 20588341518720.0, + "grad_norm": 2.7676642588600315, + "language_loss": 0.77539563, + "learning_rate": 5.017473674979502e-07, + "loss": 0.79731786, + "num_input_tokens_seen": 138973140, + "step": 6457, + "time_per_iteration": 3.407299280166626 + }, + { + "auxiliary_loss_clip": 0.01100327, + "auxiliary_loss_mlp": 0.01003081, + "balance_loss_clip": 1.01902962, + "balance_loss_mlp": 1.00198448, + "epoch": 0.7765285877472494, + "flos": 67293078560640.0, + "grad_norm": 0.739455420533259, + "language_loss": 0.58310258, + "learning_rate": 5.01231469165061e-07, + "loss": 0.60413671, + "num_input_tokens_seen": 139028965, + "step": 6458, + "time_per_iteration": 3.1127421855926514 + }, + { + "auxiliary_loss_clip": 0.01076081, + "auxiliary_loss_mlp": 0.01002892, + "balance_loss_clip": 1.01032484, + "balance_loss_mlp": 1.00175941, + "epoch": 0.7766488306378886, + "flos": 61344476121600.0, + "grad_norm": 0.8364541952622184, + "language_loss": 0.56886572, + "learning_rate": 5.007157981941663e-07, + "loss": 0.58965546, + "num_input_tokens_seen": 139094325, + "step": 6459, + "time_per_iteration": 4.225107669830322 + }, + { + "auxiliary_loss_clip": 0.01088848, + "auxiliary_loss_mlp": 0.01000409, + "balance_loss_clip": 1.01103997, + "balance_loss_mlp": 0.9993059, + "epoch": 0.7767690735285276, + "flos": 62946199393920.0, + "grad_norm": 0.8867366384726324, + "language_loss": 0.67417306, + "learning_rate": 5.002003546634928e-07, + "loss": 0.69506562, + "num_input_tokens_seen": 139150425, + "step": 6460, + "time_per_iteration": 3.217869281768799 + }, + { + "auxiliary_loss_clip": 0.01202228, + "auxiliary_loss_mlp": 0.01024605, + "balance_loss_clip": 1.04368031, + "balance_loss_mlp": 1.01711559, + "epoch": 0.7768893164191667, + "flos": 20886400575360.0, + "grad_norm": 1.593813017901255, + "language_loss": 0.76029962, + "learning_rate": 4.996851386512331e-07, + "loss": 0.78256798, + "num_input_tokens_seen": 139169130, + "step": 6461, + "time_per_iteration": 2.691009759902954 + }, + { + "auxiliary_loss_clip": 0.01187254, + "auxiliary_loss_mlp": 0.01029182, + "balance_loss_clip": 1.04579949, + "balance_loss_mlp": 1.02111197, + "epoch": 0.7770095593098058, + "flos": 20704584908160.0, + "grad_norm": 2.6598523511033836, + "language_loss": 0.83221, + "learning_rate": 4.991701502355444e-07, + "loss": 0.85437435, + "num_input_tokens_seen": 139189595, + "step": 6462, + "time_per_iteration": 2.668607711791992 + }, + { + "auxiliary_loss_clip": 0.01181781, + "auxiliary_loss_mlp": 0.0102216, + "balance_loss_clip": 1.04656172, + "balance_loss_mlp": 1.01522231, + "epoch": 0.7771298022004449, + "flos": 24717709877760.0, + "grad_norm": 1.4961640230858233, + "language_loss": 0.7584734, + "learning_rate": 4.986553894945518e-07, + "loss": 0.78051281, + "num_input_tokens_seen": 139210805, + "step": 6463, + "time_per_iteration": 2.693453550338745 + }, + { + "auxiliary_loss_clip": 0.01205338, + "auxiliary_loss_mlp": 0.01025297, + "balance_loss_clip": 1.04100442, + "balance_loss_mlp": 1.01807928, + "epoch": 0.777250045091084, + "flos": 25009232659200.0, + "grad_norm": 1.9560511413415311, + "language_loss": 0.85770833, + "learning_rate": 4.981408565063416e-07, + "loss": 0.88001466, + "num_input_tokens_seen": 139230750, + "step": 6464, + "time_per_iteration": 2.792567491531372 + }, + { + "auxiliary_loss_clip": 0.01170087, + "auxiliary_loss_mlp": 0.01025519, + "balance_loss_clip": 1.04666734, + "balance_loss_mlp": 1.01739788, + "epoch": 0.777370287981723, + "flos": 20119887319680.0, + "grad_norm": 1.9170244643491876, + "language_loss": 0.75829101, + "learning_rate": 4.976265513489701e-07, + "loss": 0.78024703, + "num_input_tokens_seen": 139250720, + "step": 6465, + "time_per_iteration": 2.586073398590088 + }, + { + "auxiliary_loss_clip": 0.01176795, + "auxiliary_loss_mlp": 0.01023949, + "balance_loss_clip": 1.0455308, + "balance_loss_mlp": 1.01686811, + "epoch": 0.7774905308723622, + "flos": 21718809331200.0, + "grad_norm": 1.9042442319086856, + "language_loss": 0.80110049, + "learning_rate": 4.971124741004562e-07, + "loss": 0.82310796, + "num_input_tokens_seen": 139269720, + "step": 6466, + "time_per_iteration": 2.5762853622436523 + }, + { + "auxiliary_loss_clip": 0.01175581, + "auxiliary_loss_mlp": 0.01026083, + "balance_loss_clip": 1.04429734, + "balance_loss_mlp": 1.01851296, + "epoch": 0.7776107737630013, + "flos": 16034115093120.0, + "grad_norm": 1.8712190116836018, + "language_loss": 0.76488101, + "learning_rate": 4.965986248387846e-07, + "loss": 0.78689766, + "num_input_tokens_seen": 139288035, + "step": 6467, + "time_per_iteration": 2.6076512336730957 + }, + { + "auxiliary_loss_clip": 0.01187192, + "auxiliary_loss_mlp": 0.01026193, + "balance_loss_clip": 1.04391456, + "balance_loss_mlp": 1.01924312, + "epoch": 0.7777310166536403, + "flos": 24790895838720.0, + "grad_norm": 1.5364247132529385, + "language_loss": 0.77233088, + "learning_rate": 4.960850036419073e-07, + "loss": 0.79446477, + "num_input_tokens_seen": 139307135, + "step": 6468, + "time_per_iteration": 2.6490604877471924 + }, + { + "auxiliary_loss_clip": 0.01184481, + "auxiliary_loss_mlp": 0.0102391, + "balance_loss_clip": 1.04570007, + "balance_loss_mlp": 1.01662016, + "epoch": 0.7778512595442795, + "flos": 17272530253440.0, + "grad_norm": 1.672864522504928, + "language_loss": 0.77903837, + "learning_rate": 4.955716105877378e-07, + "loss": 0.80112231, + "num_input_tokens_seen": 139325905, + "step": 6469, + "time_per_iteration": 2.6150059700012207 + }, + { + "auxiliary_loss_clip": 0.01179793, + "auxiliary_loss_mlp": 0.01378664, + "balance_loss_clip": 1.04521978, + "balance_loss_mlp": 1.00013018, + "epoch": 0.7779715024349185, + "flos": 17748418567680.0, + "grad_norm": 1.6601938412703467, + "language_loss": 0.83268273, + "learning_rate": 4.950584457541598e-07, + "loss": 0.85826731, + "num_input_tokens_seen": 139344370, + "step": 6470, + "time_per_iteration": 2.6208300590515137 + }, + { + "auxiliary_loss_clip": 0.01179971, + "auxiliary_loss_mlp": 0.01024784, + "balance_loss_clip": 1.04695606, + "balance_loss_mlp": 1.01733947, + "epoch": 0.7780917453255576, + "flos": 24316875031680.0, + "grad_norm": 2.2641334306171332, + "language_loss": 0.81719172, + "learning_rate": 4.945455092190183e-07, + "loss": 0.83923924, + "num_input_tokens_seen": 139365625, + "step": 6471, + "time_per_iteration": 2.658675193786621 + }, + { + "auxiliary_loss_clip": 0.01062161, + "auxiliary_loss_mlp": 0.00999539, + "balance_loss_clip": 1.00990951, + "balance_loss_mlp": 0.99847239, + "epoch": 0.7782119882161967, + "flos": 56364601530240.0, + "grad_norm": 0.6827315615476945, + "language_loss": 0.55997384, + "learning_rate": 4.940328010601271e-07, + "loss": 0.58059084, + "num_input_tokens_seen": 139430540, + "step": 6472, + "time_per_iteration": 3.1873373985290527 + }, + { + "auxiliary_loss_clip": 0.01195018, + "auxiliary_loss_mlp": 0.01027864, + "balance_loss_clip": 1.04906094, + "balance_loss_mlp": 1.02031851, + "epoch": 0.7783322311068358, + "flos": 46789986994560.0, + "grad_norm": 2.1296044730569634, + "language_loss": 0.76703334, + "learning_rate": 4.935203213552621e-07, + "loss": 0.78926206, + "num_input_tokens_seen": 139454280, + "step": 6473, + "time_per_iteration": 2.871352195739746 + }, + { + "auxiliary_loss_clip": 0.01185684, + "auxiliary_loss_mlp": 0.01022382, + "balance_loss_clip": 1.04362082, + "balance_loss_mlp": 1.01462162, + "epoch": 0.7784524739974749, + "flos": 19057864872960.0, + "grad_norm": 2.631862998003385, + "language_loss": 0.66852969, + "learning_rate": 4.930080701821662e-07, + "loss": 0.69061035, + "num_input_tokens_seen": 139471745, + "step": 6474, + "time_per_iteration": 2.703529119491577 + }, + { + "auxiliary_loss_clip": 0.01187084, + "auxiliary_loss_mlp": 0.01026388, + "balance_loss_clip": 1.04420567, + "balance_loss_mlp": 1.01923871, + "epoch": 0.778572716888114, + "flos": 24791111320320.0, + "grad_norm": 2.067056563031982, + "language_loss": 0.76755369, + "learning_rate": 4.92496047618548e-07, + "loss": 0.78968841, + "num_input_tokens_seen": 139491505, + "step": 6475, + "time_per_iteration": 2.7094123363494873 + }, + { + "auxiliary_loss_clip": 0.01177975, + "auxiliary_loss_mlp": 0.01030986, + "balance_loss_clip": 1.04714191, + "balance_loss_mlp": 1.02358317, + "epoch": 0.7786929597787531, + "flos": 20078086867200.0, + "grad_norm": 2.1523991002433385, + "language_loss": 0.77458322, + "learning_rate": 4.919842537420811e-07, + "loss": 0.79667282, + "num_input_tokens_seen": 139508620, + "step": 6476, + "time_per_iteration": 2.6638078689575195 + }, + { + "auxiliary_loss_clip": 0.01186734, + "auxiliary_loss_mlp": 0.01026891, + "balance_loss_clip": 1.04639816, + "balance_loss_mlp": 1.01980698, + "epoch": 0.7788132026693921, + "flos": 21872220318720.0, + "grad_norm": 1.53243124168465, + "language_loss": 0.79058623, + "learning_rate": 4.91472688630404e-07, + "loss": 0.81272244, + "num_input_tokens_seen": 139529360, + "step": 6477, + "time_per_iteration": 2.615605592727661 + }, + { + "auxiliary_loss_clip": 0.01162732, + "auxiliary_loss_mlp": 0.01022195, + "balance_loss_clip": 1.04433513, + "balance_loss_mlp": 1.01535201, + "epoch": 0.7789334455600313, + "flos": 11181937351680.0, + "grad_norm": 2.261334151873999, + "language_loss": 0.74232674, + "learning_rate": 4.909613523611202e-07, + "loss": 0.76417601, + "num_input_tokens_seen": 139546240, + "step": 6478, + "time_per_iteration": 2.599658250808716 + }, + { + "auxiliary_loss_clip": 0.01196575, + "auxiliary_loss_mlp": 0.01378843, + "balance_loss_clip": 1.03915858, + "balance_loss_mlp": 1.00018954, + "epoch": 0.7790536884506704, + "flos": 28695427015680.0, + "grad_norm": 1.9242898096841963, + "language_loss": 0.7470057, + "learning_rate": 4.904502450117991e-07, + "loss": 0.77275991, + "num_input_tokens_seen": 139567200, + "step": 6479, + "time_per_iteration": 4.850749492645264 + }, + { + "auxiliary_loss_clip": 0.01182941, + "auxiliary_loss_mlp": 0.01028158, + "balance_loss_clip": 1.04476237, + "balance_loss_mlp": 1.02070761, + "epoch": 0.7791739313413094, + "flos": 11072302064640.0, + "grad_norm": 2.5183346103453847, + "language_loss": 0.72089648, + "learning_rate": 4.899393666599762e-07, + "loss": 0.74300754, + "num_input_tokens_seen": 139583775, + "step": 6480, + "time_per_iteration": 2.614466428756714 + }, + { + "auxiliary_loss_clip": 0.01166622, + "auxiliary_loss_mlp": 0.01027899, + "balance_loss_clip": 1.04496384, + "balance_loss_mlp": 1.02046061, + "epoch": 0.7792941742319486, + "flos": 14679276975360.0, + "grad_norm": 2.191913429264393, + "language_loss": 0.72913599, + "learning_rate": 4.894287173831506e-07, + "loss": 0.75108123, + "num_input_tokens_seen": 139599735, + "step": 6481, + "time_per_iteration": 2.5843677520751953 + }, + { + "auxiliary_loss_clip": 0.01184873, + "auxiliary_loss_mlp": 0.01025302, + "balance_loss_clip": 1.04295063, + "balance_loss_mlp": 1.01780391, + "epoch": 0.7794144171225876, + "flos": 23258874908160.0, + "grad_norm": 2.1410838025994843, + "language_loss": 0.84010297, + "learning_rate": 4.889182972587877e-07, + "loss": 0.86220467, + "num_input_tokens_seen": 139619030, + "step": 6482, + "time_per_iteration": 2.6819615364074707 + }, + { + "auxiliary_loss_clip": 0.01201474, + "auxiliary_loss_mlp": 0.01026621, + "balance_loss_clip": 1.04381382, + "balance_loss_mlp": 1.01980805, + "epoch": 0.7795346600132267, + "flos": 21507080613120.0, + "grad_norm": 2.345281766200352, + "language_loss": 0.66336352, + "learning_rate": 4.884081063643177e-07, + "loss": 0.68564445, + "num_input_tokens_seen": 139637690, + "step": 6483, + "time_per_iteration": 2.7005257606506348 + }, + { + "auxiliary_loss_clip": 0.01086425, + "auxiliary_loss_mlp": 0.01001271, + "balance_loss_clip": 1.01447916, + "balance_loss_mlp": 1.00020969, + "epoch": 0.7796549029038659, + "flos": 70052273694720.0, + "grad_norm": 0.8536551054328324, + "language_loss": 0.52534026, + "learning_rate": 4.878981447771353e-07, + "loss": 0.54621726, + "num_input_tokens_seen": 139692070, + "step": 6484, + "time_per_iteration": 4.116549491882324 + }, + { + "auxiliary_loss_clip": 0.01188124, + "auxiliary_loss_mlp": 0.01023734, + "balance_loss_clip": 1.04340649, + "balance_loss_mlp": 1.01647449, + "epoch": 0.7797751457945049, + "flos": 23989405714560.0, + "grad_norm": 1.756243530457855, + "language_loss": 0.72972351, + "learning_rate": 4.873884125746035e-07, + "loss": 0.75184208, + "num_input_tokens_seen": 139713745, + "step": 6485, + "time_per_iteration": 3.595701217651367 + }, + { + "auxiliary_loss_clip": 0.01180189, + "auxiliary_loss_mlp": 0.010219, + "balance_loss_clip": 1.04306078, + "balance_loss_mlp": 1.01453924, + "epoch": 0.779895388685144, + "flos": 22674751937280.0, + "grad_norm": 4.558458162285522, + "language_loss": 0.71955228, + "learning_rate": 4.868789098340456e-07, + "loss": 0.74157321, + "num_input_tokens_seen": 139731650, + "step": 6486, + "time_per_iteration": 2.6346988677978516 + }, + { + "auxiliary_loss_clip": 0.01195874, + "auxiliary_loss_mlp": 0.01024516, + "balance_loss_clip": 1.04376721, + "balance_loss_mlp": 1.01679182, + "epoch": 0.7800156315757831, + "flos": 23768698596480.0, + "grad_norm": 2.052581895237823, + "language_loss": 0.73192638, + "learning_rate": 4.863696366327543e-07, + "loss": 0.7541303, + "num_input_tokens_seen": 139750820, + "step": 6487, + "time_per_iteration": 2.655491590499878 + }, + { + "auxiliary_loss_clip": 0.01175003, + "auxiliary_loss_mlp": 0.0102331, + "balance_loss_clip": 1.04279912, + "balance_loss_mlp": 1.01596701, + "epoch": 0.7801358744664222, + "flos": 26429714881920.0, + "grad_norm": 1.6106738480006915, + "language_loss": 0.778988, + "learning_rate": 4.85860593047986e-07, + "loss": 0.80097115, + "num_input_tokens_seen": 139770885, + "step": 6488, + "time_per_iteration": 2.6947062015533447 + }, + { + "auxiliary_loss_clip": 0.01183717, + "auxiliary_loss_mlp": 0.01023438, + "balance_loss_clip": 1.03774595, + "balance_loss_mlp": 1.0165689, + "epoch": 0.7802561173570612, + "flos": 26322162583680.0, + "grad_norm": 1.690088140723485, + "language_loss": 0.74641955, + "learning_rate": 4.853517791569613e-07, + "loss": 0.76849109, + "num_input_tokens_seen": 139793065, + "step": 6489, + "time_per_iteration": 2.704256296157837 + }, + { + "auxiliary_loss_clip": 0.01188403, + "auxiliary_loss_mlp": 0.01379015, + "balance_loss_clip": 1.04312098, + "balance_loss_mlp": 1.00021279, + "epoch": 0.7803763602477004, + "flos": 40333751596800.0, + "grad_norm": 1.7988185341253689, + "language_loss": 0.66188747, + "learning_rate": 4.848431950368684e-07, + "loss": 0.68756163, + "num_input_tokens_seen": 139815625, + "step": 6490, + "time_per_iteration": 2.8354246616363525 + }, + { + "auxiliary_loss_clip": 0.01062002, + "auxiliary_loss_mlp": 0.01373697, + "balance_loss_clip": 1.00993371, + "balance_loss_mlp": 0.99974138, + "epoch": 0.7804966031383395, + "flos": 67001448038400.0, + "grad_norm": 0.7037861783583087, + "language_loss": 0.55749875, + "learning_rate": 4.843348407648569e-07, + "loss": 0.58185571, + "num_input_tokens_seen": 139876905, + "step": 6491, + "time_per_iteration": 3.1866650581359863 + }, + { + "auxiliary_loss_clip": 0.01177952, + "auxiliary_loss_mlp": 0.01027103, + "balance_loss_clip": 1.04288363, + "balance_loss_mlp": 1.01997435, + "epoch": 0.7806168460289785, + "flos": 17740733057280.0, + "grad_norm": 2.7971233519071825, + "language_loss": 0.8306343, + "learning_rate": 4.838267164180457e-07, + "loss": 0.85268486, + "num_input_tokens_seen": 139892575, + "step": 6492, + "time_per_iteration": 2.6198651790618896 + }, + { + "auxiliary_loss_clip": 0.01170383, + "auxiliary_loss_mlp": 0.0102564, + "balance_loss_clip": 1.04713583, + "balance_loss_mlp": 1.01801097, + "epoch": 0.7807370889196176, + "flos": 23946240545280.0, + "grad_norm": 3.6486596351446656, + "language_loss": 0.83856881, + "learning_rate": 4.833188220735156e-07, + "loss": 0.86052901, + "num_input_tokens_seen": 139912245, + "step": 6493, + "time_per_iteration": 2.6483492851257324 + }, + { + "auxiliary_loss_clip": 0.01175564, + "auxiliary_loss_mlp": 0.0101798, + "balance_loss_clip": 1.04548693, + "balance_loss_mlp": 1.01106536, + "epoch": 0.7808573318102567, + "flos": 18989024457600.0, + "grad_norm": 2.009701798182113, + "language_loss": 0.74487036, + "learning_rate": 4.828111578083152e-07, + "loss": 0.76680577, + "num_input_tokens_seen": 139929150, + "step": 6494, + "time_per_iteration": 2.603726387023926 + }, + { + "auxiliary_loss_clip": 0.01185572, + "auxiliary_loss_mlp": 0.01024296, + "balance_loss_clip": 1.04777503, + "balance_loss_mlp": 1.01655364, + "epoch": 0.7809775747008958, + "flos": 23980750536960.0, + "grad_norm": 1.9461206581494788, + "language_loss": 0.81464851, + "learning_rate": 4.823037236994556e-07, + "loss": 0.83674717, + "num_input_tokens_seen": 139947315, + "step": 6495, + "time_per_iteration": 2.7006146907806396 + }, + { + "auxiliary_loss_clip": 0.01075685, + "auxiliary_loss_mlp": 0.01002866, + "balance_loss_clip": 1.00987625, + "balance_loss_mlp": 1.00170994, + "epoch": 0.7810978175915348, + "flos": 68535875180160.0, + "grad_norm": 0.730492321664272, + "language_loss": 0.56331116, + "learning_rate": 4.817965198239136e-07, + "loss": 0.58409667, + "num_input_tokens_seen": 140013775, + "step": 6496, + "time_per_iteration": 3.1725668907165527 + }, + { + "auxiliary_loss_clip": 0.01192479, + "auxiliary_loss_mlp": 0.01023843, + "balance_loss_clip": 1.04171193, + "balance_loss_mlp": 1.01576066, + "epoch": 0.781218060482174, + "flos": 19642131498240.0, + "grad_norm": 2.229719349929557, + "language_loss": 0.74484694, + "learning_rate": 4.812895462586331e-07, + "loss": 0.76701021, + "num_input_tokens_seen": 140031600, + "step": 6497, + "time_per_iteration": 2.7153406143188477 + }, + { + "auxiliary_loss_clip": 0.0119923, + "auxiliary_loss_mlp": 0.01027649, + "balance_loss_clip": 1.04390001, + "balance_loss_mlp": 1.02032995, + "epoch": 0.7813383033728131, + "flos": 25627865621760.0, + "grad_norm": 1.6205785202845118, + "language_loss": 0.81779373, + "learning_rate": 4.807828030805207e-07, + "loss": 0.8400625, + "num_input_tokens_seen": 140050590, + "step": 6498, + "time_per_iteration": 2.670456886291504 + }, + { + "auxiliary_loss_clip": 0.01174537, + "auxiliary_loss_mlp": 0.01028744, + "balance_loss_clip": 1.04632473, + "balance_loss_mlp": 1.02182412, + "epoch": 0.7814585462634521, + "flos": 20485924865280.0, + "grad_norm": 2.502705982246898, + "language_loss": 0.67713797, + "learning_rate": 4.802762903664495e-07, + "loss": 0.69917077, + "num_input_tokens_seen": 140069770, + "step": 6499, + "time_per_iteration": 2.6528306007385254 + }, + { + "auxiliary_loss_clip": 0.01194755, + "auxiliary_loss_mlp": 0.01026721, + "balance_loss_clip": 1.04806554, + "balance_loss_mlp": 1.0187335, + "epoch": 0.7815787891540913, + "flos": 22304297018880.0, + "grad_norm": 2.1054579732962253, + "language_loss": 0.73317051, + "learning_rate": 4.797700081932565e-07, + "loss": 0.75538528, + "num_input_tokens_seen": 140087635, + "step": 6500, + "time_per_iteration": 2.613574981689453 + }, + { + "auxiliary_loss_clip": 0.012034, + "auxiliary_loss_mlp": 0.01025788, + "balance_loss_clip": 1.03837609, + "balance_loss_mlp": 1.01908827, + "epoch": 0.7816990320447303, + "flos": 22600668136320.0, + "grad_norm": 2.1839158599306474, + "language_loss": 0.81679738, + "learning_rate": 4.792639566377442e-07, + "loss": 0.83908927, + "num_input_tokens_seen": 140105045, + "step": 6501, + "time_per_iteration": 2.7778027057647705 + }, + { + "auxiliary_loss_clip": 0.01170045, + "auxiliary_loss_mlp": 0.01025664, + "balance_loss_clip": 1.04282856, + "balance_loss_mlp": 1.01749206, + "epoch": 0.7818192749353694, + "flos": 24935974871040.0, + "grad_norm": 1.7471530542697706, + "language_loss": 0.77297199, + "learning_rate": 4.78758135776681e-07, + "loss": 0.79492903, + "num_input_tokens_seen": 140124900, + "step": 6502, + "time_per_iteration": 2.6232738494873047 + }, + { + "auxiliary_loss_clip": 0.01189992, + "auxiliary_loss_mlp": 0.01026574, + "balance_loss_clip": 1.04691243, + "balance_loss_mlp": 1.01918042, + "epoch": 0.7819395178260086, + "flos": 23733039369600.0, + "grad_norm": 2.0886322361444116, + "language_loss": 0.78836823, + "learning_rate": 4.782525456867989e-07, + "loss": 0.81053382, + "num_input_tokens_seen": 140143755, + "step": 6503, + "time_per_iteration": 2.6616265773773193 + }, + { + "auxiliary_loss_clip": 0.0119924, + "auxiliary_loss_mlp": 0.01028731, + "balance_loss_clip": 1.04426289, + "balance_loss_mlp": 1.02079177, + "epoch": 0.7820597607166476, + "flos": 23221671396480.0, + "grad_norm": 1.9290456091081838, + "language_loss": 0.83072317, + "learning_rate": 4.777471864447959e-07, + "loss": 0.85300291, + "num_input_tokens_seen": 140164495, + "step": 6504, + "time_per_iteration": 3.581712007522583 + }, + { + "auxiliary_loss_clip": 0.01183191, + "auxiliary_loss_mlp": 0.01028237, + "balance_loss_clip": 1.04167986, + "balance_loss_mlp": 1.02067924, + "epoch": 0.7821800036072867, + "flos": 22309540404480.0, + "grad_norm": 1.940157758096377, + "language_loss": 0.80771554, + "learning_rate": 4.772420581273344e-07, + "loss": 0.82982981, + "num_input_tokens_seen": 140181980, + "step": 6505, + "time_per_iteration": 3.6504087448120117 + }, + { + "auxiliary_loss_clip": 0.0117404, + "auxiliary_loss_mlp": 0.01024094, + "balance_loss_clip": 1.04661977, + "balance_loss_mlp": 1.01702523, + "epoch": 0.7823002464979258, + "flos": 21544176384000.0, + "grad_norm": 2.3511157308403816, + "language_loss": 0.76470345, + "learning_rate": 4.7673716081104134e-07, + "loss": 0.78668481, + "num_input_tokens_seen": 140202155, + "step": 6506, + "time_per_iteration": 2.7539713382720947 + }, + { + "auxiliary_loss_clip": 0.01176937, + "auxiliary_loss_mlp": 0.01025203, + "balance_loss_clip": 1.04749179, + "balance_loss_mlp": 1.0176872, + "epoch": 0.7824204893885649, + "flos": 24535642815360.0, + "grad_norm": 1.809805224353339, + "language_loss": 0.84440452, + "learning_rate": 4.762324945725109e-07, + "loss": 0.86642593, + "num_input_tokens_seen": 140221600, + "step": 6507, + "time_per_iteration": 2.655078887939453 + }, + { + "auxiliary_loss_clip": 0.01184238, + "auxiliary_loss_mlp": 0.01025511, + "balance_loss_clip": 1.04720032, + "balance_loss_mlp": 1.01816213, + "epoch": 0.782540732279204, + "flos": 27415211402880.0, + "grad_norm": 1.654393790154856, + "language_loss": 0.75577563, + "learning_rate": 4.7572805948829844e-07, + "loss": 0.7778731, + "num_input_tokens_seen": 140241860, + "step": 6508, + "time_per_iteration": 2.647233486175537 + }, + { + "auxiliary_loss_clip": 0.01208643, + "auxiliary_loss_mlp": 0.01024758, + "balance_loss_clip": 1.04271996, + "balance_loss_mlp": 1.01791883, + "epoch": 0.7826609751698431, + "flos": 24353216616960.0, + "grad_norm": 6.945013328572358, + "language_loss": 0.71314788, + "learning_rate": 4.7522385563492795e-07, + "loss": 0.73548192, + "num_input_tokens_seen": 140262160, + "step": 6509, + "time_per_iteration": 3.6741676330566406 + }, + { + "auxiliary_loss_clip": 0.01199192, + "auxiliary_loss_mlp": 0.01026234, + "balance_loss_clip": 1.04458582, + "balance_loss_mlp": 1.01850891, + "epoch": 0.7827812180604822, + "flos": 23988543788160.0, + "grad_norm": 1.8806976213208086, + "language_loss": 0.70508152, + "learning_rate": 4.747198830888863e-07, + "loss": 0.72733581, + "num_input_tokens_seen": 140282030, + "step": 6510, + "time_per_iteration": 3.565354824066162 + }, + { + "auxiliary_loss_clip": 0.01178934, + "auxiliary_loss_mlp": 0.01025921, + "balance_loss_clip": 1.04324281, + "balance_loss_mlp": 1.01832771, + "epoch": 0.7829014609511212, + "flos": 27454318335360.0, + "grad_norm": 3.1289116438269007, + "language_loss": 0.68214703, + "learning_rate": 4.742161419266251e-07, + "loss": 0.70419556, + "num_input_tokens_seen": 140301190, + "step": 6511, + "time_per_iteration": 2.6582729816436768 + }, + { + "auxiliary_loss_clip": 0.01179887, + "auxiliary_loss_mlp": 0.01021689, + "balance_loss_clip": 1.04690456, + "balance_loss_mlp": 1.01445341, + "epoch": 0.7830217038417604, + "flos": 29204532432000.0, + "grad_norm": 3.0630456117673446, + "language_loss": 0.652125, + "learning_rate": 4.7371263222456304e-07, + "loss": 0.67414081, + "num_input_tokens_seen": 140318510, + "step": 6512, + "time_per_iteration": 2.6536569595336914 + }, + { + "auxiliary_loss_clip": 0.01071525, + "auxiliary_loss_mlp": 0.0100349, + "balance_loss_clip": 1.01024008, + "balance_loss_mlp": 1.0023936, + "epoch": 0.7831419467323995, + "flos": 60950895822720.0, + "grad_norm": 0.7995366217263676, + "language_loss": 0.61388999, + "learning_rate": 4.7320935405908004e-07, + "loss": 0.63464016, + "num_input_tokens_seen": 140379380, + "step": 6513, + "time_per_iteration": 3.1528139114379883 + }, + { + "auxiliary_loss_clip": 0.01174047, + "auxiliary_loss_mlp": 0.01027289, + "balance_loss_clip": 1.04912972, + "balance_loss_mlp": 1.01930213, + "epoch": 0.7832621896230385, + "flos": 19682531320320.0, + "grad_norm": 2.5048065534657864, + "language_loss": 0.84388328, + "learning_rate": 4.7270630750652475e-07, + "loss": 0.8658967, + "num_input_tokens_seen": 140395335, + "step": 6514, + "time_per_iteration": 2.551710844039917 + }, + { + "auxiliary_loss_clip": 0.01175427, + "auxiliary_loss_mlp": 0.01022295, + "balance_loss_clip": 1.04569864, + "balance_loss_mlp": 1.01512194, + "epoch": 0.7833824325136777, + "flos": 25009232659200.0, + "grad_norm": 1.6353335652188994, + "language_loss": 0.80209708, + "learning_rate": 4.7220349264320746e-07, + "loss": 0.82407439, + "num_input_tokens_seen": 140414420, + "step": 6515, + "time_per_iteration": 2.649829864501953 + }, + { + "auxiliary_loss_clip": 0.01074467, + "auxiliary_loss_mlp": 0.01003261, + "balance_loss_clip": 1.01080251, + "balance_loss_mlp": 1.00222361, + "epoch": 0.7835026754043167, + "flos": 68800142517120.0, + "grad_norm": 0.7339323807098255, + "language_loss": 0.5490067, + "learning_rate": 4.71700909545407e-07, + "loss": 0.56978393, + "num_input_tokens_seen": 140477365, + "step": 6516, + "time_per_iteration": 3.1929166316986084 + }, + { + "auxiliary_loss_clip": 0.01177172, + "auxiliary_loss_mlp": 0.01021257, + "balance_loss_clip": 1.04461908, + "balance_loss_mlp": 1.014256, + "epoch": 0.7836229182949558, + "flos": 19864598382720.0, + "grad_norm": 2.025736464088908, + "language_loss": 0.76885909, + "learning_rate": 4.711985582893627e-07, + "loss": 0.79084337, + "num_input_tokens_seen": 140495885, + "step": 6517, + "time_per_iteration": 2.5405473709106445 + }, + { + "auxiliary_loss_clip": 0.01202291, + "auxiliary_loss_mlp": 0.0102693, + "balance_loss_clip": 1.0389204, + "balance_loss_mlp": 1.01927924, + "epoch": 0.783743161185595, + "flos": 22965843755520.0, + "grad_norm": 1.7141179623498923, + "language_loss": 0.71580821, + "learning_rate": 4.706964389512811e-07, + "loss": 0.73810047, + "num_input_tokens_seen": 140515920, + "step": 6518, + "time_per_iteration": 2.7357771396636963 + }, + { + "auxiliary_loss_clip": 0.01166863, + "auxiliary_loss_mlp": 0.01024535, + "balance_loss_clip": 1.04744267, + "balance_loss_mlp": 1.01748991, + "epoch": 0.783863404076234, + "flos": 12458489777280.0, + "grad_norm": 2.0105939535127937, + "language_loss": 0.87554771, + "learning_rate": 4.701945516073345e-07, + "loss": 0.89746165, + "num_input_tokens_seen": 140533395, + "step": 6519, + "time_per_iteration": 2.5322999954223633 + }, + { + "auxiliary_loss_clip": 0.01194298, + "auxiliary_loss_mlp": 0.01022805, + "balance_loss_clip": 1.04583216, + "balance_loss_mlp": 1.01584899, + "epoch": 0.7839836469668731, + "flos": 24243940465920.0, + "grad_norm": 1.854124065142136, + "language_loss": 0.75217283, + "learning_rate": 4.696928963336577e-07, + "loss": 0.77434385, + "num_input_tokens_seen": 140552825, + "step": 6520, + "time_per_iteration": 2.7385470867156982 + }, + { + "auxiliary_loss_clip": 0.010717, + "auxiliary_loss_mlp": 0.01001415, + "balance_loss_clip": 1.01038671, + "balance_loss_mlp": 1.00036645, + "epoch": 0.7841038898575122, + "flos": 62121978938880.0, + "grad_norm": 0.8550106773846108, + "language_loss": 0.60976452, + "learning_rate": 4.6919147320635224e-07, + "loss": 0.63049573, + "num_input_tokens_seen": 140615535, + "step": 6521, + "time_per_iteration": 3.145781993865967 + }, + { + "auxiliary_loss_clip": 0.0117739, + "auxiliary_loss_mlp": 0.01024378, + "balance_loss_clip": 1.04510081, + "balance_loss_mlp": 1.01727581, + "epoch": 0.7842241327481513, + "flos": 20193899293440.0, + "grad_norm": 2.680612371431939, + "language_loss": 0.7317735, + "learning_rate": 4.6869028230148286e-07, + "loss": 0.75379121, + "num_input_tokens_seen": 140633330, + "step": 6522, + "time_per_iteration": 2.556544065475464 + }, + { + "auxiliary_loss_clip": 0.01185162, + "auxiliary_loss_mlp": 0.01028285, + "balance_loss_clip": 1.03998029, + "balance_loss_mlp": 1.02023268, + "epoch": 0.7843443756387903, + "flos": 28074531496320.0, + "grad_norm": 2.5253877097666546, + "language_loss": 0.59874034, + "learning_rate": 4.6818932369507957e-07, + "loss": 0.62087482, + "num_input_tokens_seen": 140652830, + "step": 6523, + "time_per_iteration": 2.7205002307891846 + }, + { + "auxiliary_loss_clip": 0.01176755, + "auxiliary_loss_mlp": 0.01022192, + "balance_loss_clip": 1.04620028, + "balance_loss_mlp": 1.01438355, + "epoch": 0.7844646185294295, + "flos": 21323397438720.0, + "grad_norm": 1.8746286947138862, + "language_loss": 0.89250332, + "learning_rate": 4.676885974631386e-07, + "loss": 0.91449285, + "num_input_tokens_seen": 140671190, + "step": 6524, + "time_per_iteration": 2.5805256366729736 + }, + { + "auxiliary_loss_clip": 0.01178006, + "auxiliary_loss_mlp": 0.01026542, + "balance_loss_clip": 1.04639959, + "balance_loss_mlp": 1.0190146, + "epoch": 0.7845848614200686, + "flos": 23656585271040.0, + "grad_norm": 9.174834064220091, + "language_loss": 0.81336975, + "learning_rate": 4.67188103681619e-07, + "loss": 0.83541524, + "num_input_tokens_seen": 140690975, + "step": 6525, + "time_per_iteration": 2.6041951179504395 + }, + { + "auxiliary_loss_clip": 0.011745, + "auxiliary_loss_mlp": 0.01378627, + "balance_loss_clip": 1.04832888, + "balance_loss_mlp": 1.00013888, + "epoch": 0.7847051043107076, + "flos": 23402194174080.0, + "grad_norm": 2.160017189918886, + "language_loss": 0.68902206, + "learning_rate": 4.666878424264453e-07, + "loss": 0.7145533, + "num_input_tokens_seen": 140710930, + "step": 6526, + "time_per_iteration": 2.6170620918273926 + }, + { + "auxiliary_loss_clip": 0.01174621, + "auxiliary_loss_mlp": 0.01030257, + "balance_loss_clip": 1.04340744, + "balance_loss_mlp": 1.02371562, + "epoch": 0.7848253472013467, + "flos": 19022277473280.0, + "grad_norm": 1.6166346089666999, + "language_loss": 0.73979807, + "learning_rate": 4.661878137735069e-07, + "loss": 0.76184684, + "num_input_tokens_seen": 140729120, + "step": 6527, + "time_per_iteration": 2.6801581382751465 + }, + { + "auxiliary_loss_clip": 0.01187344, + "auxiliary_loss_mlp": 0.01022813, + "balance_loss_clip": 1.04585195, + "balance_loss_mlp": 1.01600957, + "epoch": 0.7849455900919858, + "flos": 21179180332800.0, + "grad_norm": 1.849941708481619, + "language_loss": 0.74564308, + "learning_rate": 4.656880177986571e-07, + "loss": 0.76774466, + "num_input_tokens_seen": 140747665, + "step": 6528, + "time_per_iteration": 2.5912561416625977 + }, + { + "auxiliary_loss_clip": 0.01188218, + "auxiliary_loss_mlp": 0.01027854, + "balance_loss_clip": 1.04372382, + "balance_loss_mlp": 1.02023935, + "epoch": 0.7850658329826249, + "flos": 19536482620800.0, + "grad_norm": 1.8737071475600975, + "language_loss": 0.8107723, + "learning_rate": 4.6518845457771607e-07, + "loss": 0.83293301, + "num_input_tokens_seen": 140766525, + "step": 6529, + "time_per_iteration": 2.6723849773406982 + }, + { + "auxiliary_loss_clip": 0.01169248, + "auxiliary_loss_mlp": 0.01378718, + "balance_loss_clip": 1.0456351, + "balance_loss_mlp": 1.00017548, + "epoch": 0.7851860758732639, + "flos": 12495334152960.0, + "grad_norm": 4.0816227640846074, + "language_loss": 0.79013455, + "learning_rate": 4.646891241864652e-07, + "loss": 0.81561428, + "num_input_tokens_seen": 140785090, + "step": 6530, + "time_per_iteration": 3.5128448009490967 + }, + { + "auxiliary_loss_clip": 0.01172215, + "auxiliary_loss_mlp": 0.01025013, + "balance_loss_clip": 1.04353213, + "balance_loss_mlp": 1.01715088, + "epoch": 0.7853063187639031, + "flos": 22960959505920.0, + "grad_norm": 1.7725407880291928, + "language_loss": 0.73128593, + "learning_rate": 4.6419002670065397e-07, + "loss": 0.75325823, + "num_input_tokens_seen": 140804670, + "step": 6531, + "time_per_iteration": 2.5912418365478516 + }, + { + "auxiliary_loss_clip": 0.01200436, + "auxiliary_loss_mlp": 0.01026004, + "balance_loss_clip": 1.04630518, + "balance_loss_mlp": 1.01830912, + "epoch": 0.7854265616545422, + "flos": 17347260499200.0, + "grad_norm": 2.239026933805067, + "language_loss": 0.86548305, + "learning_rate": 4.6369116219599445e-07, + "loss": 0.88774747, + "num_input_tokens_seen": 140820655, + "step": 6532, + "time_per_iteration": 3.6124420166015625 + }, + { + "auxiliary_loss_clip": 0.01194972, + "auxiliary_loss_mlp": 0.01022177, + "balance_loss_clip": 1.04273558, + "balance_loss_mlp": 1.01528335, + "epoch": 0.7855468045451812, + "flos": 23838293197440.0, + "grad_norm": 1.721193379499141, + "language_loss": 0.79387981, + "learning_rate": 4.631925307481637e-07, + "loss": 0.8160513, + "num_input_tokens_seen": 140840470, + "step": 6533, + "time_per_iteration": 2.6731767654418945 + }, + { + "auxiliary_loss_clip": 0.0118595, + "auxiliary_loss_mlp": 0.01025622, + "balance_loss_clip": 1.04662228, + "balance_loss_mlp": 1.01838017, + "epoch": 0.7856670474358204, + "flos": 25666792986240.0, + "grad_norm": 2.1583112998888, + "language_loss": 0.75311577, + "learning_rate": 4.6269413243280533e-07, + "loss": 0.77523148, + "num_input_tokens_seen": 140859890, + "step": 6534, + "time_per_iteration": 2.6787638664245605 + }, + { + "auxiliary_loss_clip": 0.01190986, + "auxiliary_loss_mlp": 0.01023101, + "balance_loss_clip": 1.04674911, + "balance_loss_mlp": 1.0155369, + "epoch": 0.7857872903264594, + "flos": 18144656472960.0, + "grad_norm": 2.4281077370515725, + "language_loss": 0.7494514, + "learning_rate": 4.621959673255236e-07, + "loss": 0.77159226, + "num_input_tokens_seen": 140876190, + "step": 6535, + "time_per_iteration": 3.532418966293335 + }, + { + "auxiliary_loss_clip": 0.01200189, + "auxiliary_loss_mlp": 0.0103016, + "balance_loss_clip": 1.04109812, + "balance_loss_mlp": 1.02323055, + "epoch": 0.7859075332170985, + "flos": 14386138081920.0, + "grad_norm": 1.898920771273052, + "language_loss": 0.90372968, + "learning_rate": 4.6169803550189135e-07, + "loss": 0.9260332, + "num_input_tokens_seen": 140891885, + "step": 6536, + "time_per_iteration": 3.5900745391845703 + }, + { + "auxiliary_loss_clip": 0.01195465, + "auxiliary_loss_mlp": 0.01027912, + "balance_loss_clip": 1.04218173, + "balance_loss_mlp": 1.02006805, + "epoch": 0.7860277761077377, + "flos": 19864059678720.0, + "grad_norm": 3.291537113465824, + "language_loss": 0.7737931, + "learning_rate": 4.6120033703744355e-07, + "loss": 0.79602683, + "num_input_tokens_seen": 140910780, + "step": 6537, + "time_per_iteration": 2.765529155731201 + }, + { + "auxiliary_loss_clip": 0.01174312, + "auxiliary_loss_mlp": 0.01019369, + "balance_loss_clip": 1.04204559, + "balance_loss_mlp": 1.01269364, + "epoch": 0.7861480189983767, + "flos": 26396174557440.0, + "grad_norm": 1.7724214357521415, + "language_loss": 0.78337693, + "learning_rate": 4.607028720076822e-07, + "loss": 0.80531371, + "num_input_tokens_seen": 140927460, + "step": 6538, + "time_per_iteration": 2.6271867752075195 + }, + { + "auxiliary_loss_clip": 0.0117916, + "auxiliary_loss_mlp": 0.01024766, + "balance_loss_clip": 1.04771638, + "balance_loss_mlp": 1.01722026, + "epoch": 0.7862682618890158, + "flos": 24236578177920.0, + "grad_norm": 1.8467311074426318, + "language_loss": 0.73558581, + "learning_rate": 4.6020564048807074e-07, + "loss": 0.75762504, + "num_input_tokens_seen": 140945135, + "step": 6539, + "time_per_iteration": 2.6108484268188477 + }, + { + "auxiliary_loss_clip": 0.0117906, + "auxiliary_loss_mlp": 0.01028567, + "balance_loss_clip": 1.04628944, + "balance_loss_mlp": 1.02101481, + "epoch": 0.7863885047796549, + "flos": 47551508259840.0, + "grad_norm": 2.270394954346918, + "language_loss": 0.72006035, + "learning_rate": 4.5970864255403883e-07, + "loss": 0.74213666, + "num_input_tokens_seen": 140966660, + "step": 6540, + "time_per_iteration": 2.7858376502990723 + }, + { + "auxiliary_loss_clip": 0.0116249, + "auxiliary_loss_mlp": 0.01024138, + "balance_loss_clip": 1.04245925, + "balance_loss_mlp": 1.01731646, + "epoch": 0.786508747670294, + "flos": 24389234979840.0, + "grad_norm": 2.199026678592603, + "language_loss": 0.82550097, + "learning_rate": 4.59211878280982e-07, + "loss": 0.84736717, + "num_input_tokens_seen": 140986175, + "step": 6541, + "time_per_iteration": 2.6079487800598145 + }, + { + "auxiliary_loss_clip": 0.01187335, + "auxiliary_loss_mlp": 0.01024371, + "balance_loss_clip": 1.04484701, + "balance_loss_mlp": 1.01710486, + "epoch": 0.786628990560933, + "flos": 18041234238720.0, + "grad_norm": 2.474607503370842, + "language_loss": 0.6965273, + "learning_rate": 4.587153477442578e-07, + "loss": 0.71864438, + "num_input_tokens_seen": 141002490, + "step": 6542, + "time_per_iteration": 2.6154162883758545 + }, + { + "auxiliary_loss_clip": 0.01173121, + "auxiliary_loss_mlp": 0.01022542, + "balance_loss_clip": 1.04883099, + "balance_loss_mlp": 1.01520503, + "epoch": 0.7867492334515722, + "flos": 25848860048640.0, + "grad_norm": 2.1295968563286216, + "language_loss": 0.81235379, + "learning_rate": 4.582190510191899e-07, + "loss": 0.83431041, + "num_input_tokens_seen": 141021150, + "step": 6543, + "time_per_iteration": 2.5971508026123047 + }, + { + "auxiliary_loss_clip": 0.0119097, + "auxiliary_loss_mlp": 0.01021784, + "balance_loss_clip": 1.04369879, + "balance_loss_mlp": 1.01473832, + "epoch": 0.7868694763422113, + "flos": 16580819070720.0, + "grad_norm": 2.0427763092068596, + "language_loss": 0.87206292, + "learning_rate": 4.5772298818106625e-07, + "loss": 0.89419043, + "num_input_tokens_seen": 141036940, + "step": 6544, + "time_per_iteration": 2.62319278717041 + }, + { + "auxiliary_loss_clip": 0.01201806, + "auxiliary_loss_mlp": 0.01028362, + "balance_loss_clip": 1.04467463, + "balance_loss_mlp": 1.02084303, + "epoch": 0.7869897192328503, + "flos": 29386276272000.0, + "grad_norm": 2.4808199886765028, + "language_loss": 0.72150576, + "learning_rate": 4.572271593051384e-07, + "loss": 0.74380749, + "num_input_tokens_seen": 141054295, + "step": 6545, + "time_per_iteration": 2.699183464050293 + }, + { + "auxiliary_loss_clip": 0.01195535, + "auxiliary_loss_mlp": 0.01028597, + "balance_loss_clip": 1.04228079, + "balance_loss_mlp": 1.02095819, + "epoch": 0.7871099621234895, + "flos": 17128923678720.0, + "grad_norm": 1.5680957976492245, + "language_loss": 0.78344077, + "learning_rate": 4.567315644666245e-07, + "loss": 0.80568206, + "num_input_tokens_seen": 141073090, + "step": 6546, + "time_per_iteration": 2.7189559936523438 + }, + { + "auxiliary_loss_clip": 0.01191273, + "auxiliary_loss_mlp": 0.01029484, + "balance_loss_clip": 1.04594731, + "balance_loss_mlp": 1.02234364, + "epoch": 0.7872302050141285, + "flos": 23440187784960.0, + "grad_norm": 2.405764514080676, + "language_loss": 0.85026419, + "learning_rate": 4.5623620374070507e-07, + "loss": 0.87247181, + "num_input_tokens_seen": 141092405, + "step": 6547, + "time_per_iteration": 2.6723945140838623 + }, + { + "auxiliary_loss_clip": 0.01098306, + "auxiliary_loss_mlp": 0.01001317, + "balance_loss_clip": 1.01080799, + "balance_loss_mlp": 1.00010705, + "epoch": 0.7873504479047676, + "flos": 65959752689280.0, + "grad_norm": 0.763017851287178, + "language_loss": 0.58417678, + "learning_rate": 4.557410772025263e-07, + "loss": 0.60517299, + "num_input_tokens_seen": 141154355, + "step": 6548, + "time_per_iteration": 3.3908371925354004 + }, + { + "auxiliary_loss_clip": 0.01182959, + "auxiliary_loss_mlp": 0.01026698, + "balance_loss_clip": 1.04416299, + "balance_loss_mlp": 1.01964128, + "epoch": 0.7874706907954068, + "flos": 23258336204160.0, + "grad_norm": 1.856231294197791, + "language_loss": 0.66029072, + "learning_rate": 4.5524618492719803e-07, + "loss": 0.68238723, + "num_input_tokens_seen": 141173575, + "step": 6549, + "time_per_iteration": 2.6640703678131104 + }, + { + "auxiliary_loss_clip": 0.01179154, + "auxiliary_loss_mlp": 0.01027879, + "balance_loss_clip": 1.04614925, + "balance_loss_mlp": 1.02130425, + "epoch": 0.7875909336860458, + "flos": 28767786963840.0, + "grad_norm": 1.5775833740907275, + "language_loss": 0.79053229, + "learning_rate": 4.54751526989795e-07, + "loss": 0.81260264, + "num_input_tokens_seen": 141195415, + "step": 6550, + "time_per_iteration": 2.7189548015594482 + }, + { + "auxiliary_loss_clip": 0.01178364, + "auxiliary_loss_mlp": 0.01024552, + "balance_loss_clip": 1.04457402, + "balance_loss_mlp": 1.01719713, + "epoch": 0.7877111765766849, + "flos": 18697286194560.0, + "grad_norm": 2.7927240089177783, + "language_loss": 0.78957999, + "learning_rate": 4.5425710346535775e-07, + "loss": 0.81160921, + "num_input_tokens_seen": 141213360, + "step": 6551, + "time_per_iteration": 2.5435359477996826 + }, + { + "auxiliary_loss_clip": 0.0117835, + "auxiliary_loss_mlp": 0.01025967, + "balance_loss_clip": 1.04537463, + "balance_loss_mlp": 1.01882625, + "epoch": 0.787831419467324, + "flos": 27592968833280.0, + "grad_norm": 2.0477324245703454, + "language_loss": 0.81751585, + "learning_rate": 4.537629144288877e-07, + "loss": 0.83955896, + "num_input_tokens_seen": 141230815, + "step": 6552, + "time_per_iteration": 2.668997049331665 + }, + { + "auxiliary_loss_clip": 0.01209664, + "auxiliary_loss_mlp": 0.01023442, + "balance_loss_clip": 1.04063797, + "balance_loss_mlp": 1.01576757, + "epoch": 0.7879516623579631, + "flos": 18150187167360.0, + "grad_norm": 1.9485350505461387, + "language_loss": 0.75296658, + "learning_rate": 4.5326895995535477e-07, + "loss": 0.77529764, + "num_input_tokens_seen": 141249715, + "step": 6553, + "time_per_iteration": 2.696263074874878 + }, + { + "auxiliary_loss_clip": 0.01173194, + "auxiliary_loss_mlp": 0.0102393, + "balance_loss_clip": 1.04380012, + "balance_loss_mlp": 1.01655674, + "epoch": 0.7880719052486022, + "flos": 20339193807360.0, + "grad_norm": 2.2464141432693707, + "language_loss": 0.8403852, + "learning_rate": 4.527752401196907e-07, + "loss": 0.86235648, + "num_input_tokens_seen": 141267730, + "step": 6554, + "time_per_iteration": 2.587174892425537 + }, + { + "auxiliary_loss_clip": 0.01179425, + "auxiliary_loss_mlp": 0.01028015, + "balance_loss_clip": 1.04327476, + "balance_loss_mlp": 1.02042484, + "epoch": 0.7881921481392413, + "flos": 21653237053440.0, + "grad_norm": 1.7480331838117127, + "language_loss": 0.66978115, + "learning_rate": 4.5228175499679254e-07, + "loss": 0.69185555, + "num_input_tokens_seen": 141287315, + "step": 6555, + "time_per_iteration": 2.651226043701172 + }, + { + "auxiliary_loss_clip": 0.01075419, + "auxiliary_loss_mlp": 0.01002252, + "balance_loss_clip": 1.01042557, + "balance_loss_mlp": 1.00116718, + "epoch": 0.7883123910298804, + "flos": 68565860058240.0, + "grad_norm": 0.8587976954415776, + "language_loss": 0.54595917, + "learning_rate": 4.5178850466152174e-07, + "loss": 0.56673586, + "num_input_tokens_seen": 141346145, + "step": 6556, + "time_per_iteration": 4.128547668457031 + }, + { + "auxiliary_loss_clip": 0.01179297, + "auxiliary_loss_mlp": 0.01027343, + "balance_loss_clip": 1.04208946, + "balance_loss_mlp": 1.01992786, + "epoch": 0.7884326339205194, + "flos": 19318217627520.0, + "grad_norm": 2.3006411459409435, + "language_loss": 0.81944311, + "learning_rate": 4.512954891887031e-07, + "loss": 0.84150958, + "num_input_tokens_seen": 141364445, + "step": 6557, + "time_per_iteration": 3.498788356781006 + }, + { + "auxiliary_loss_clip": 0.01177893, + "auxiliary_loss_mlp": 0.01026408, + "balance_loss_clip": 1.04384685, + "balance_loss_mlp": 1.01881433, + "epoch": 0.7885528768111585, + "flos": 17784903807360.0, + "grad_norm": 2.4598799204765203, + "language_loss": 0.83156359, + "learning_rate": 4.5080270865312806e-07, + "loss": 0.85360658, + "num_input_tokens_seen": 141381640, + "step": 6558, + "time_per_iteration": 2.630612373352051 + }, + { + "auxiliary_loss_clip": 0.01175734, + "auxiliary_loss_mlp": 0.01023815, + "balance_loss_clip": 1.0449394, + "balance_loss_mlp": 1.01628661, + "epoch": 0.7886731197017977, + "flos": 18807639753600.0, + "grad_norm": 2.2307185238611265, + "language_loss": 0.71158445, + "learning_rate": 4.5031016312954985e-07, + "loss": 0.73357993, + "num_input_tokens_seen": 141399955, + "step": 6559, + "time_per_iteration": 2.5650925636291504 + }, + { + "auxiliary_loss_clip": 0.0118582, + "auxiliary_loss_mlp": 0.01028467, + "balance_loss_clip": 1.04768753, + "balance_loss_mlp": 1.02062309, + "epoch": 0.7887933625924367, + "flos": 33365358126720.0, + "grad_norm": 4.701101494413159, + "language_loss": 0.74503767, + "learning_rate": 4.498178526926886e-07, + "loss": 0.76718056, + "num_input_tokens_seen": 141420820, + "step": 6560, + "time_per_iteration": 2.7590010166168213 + }, + { + "auxiliary_loss_clip": 0.01166836, + "auxiliary_loss_mlp": 0.01025418, + "balance_loss_clip": 1.04705548, + "balance_loss_mlp": 1.01823545, + "epoch": 0.7889136054830758, + "flos": 17019360218880.0, + "grad_norm": 2.0778659239382176, + "language_loss": 0.7215513, + "learning_rate": 4.4932577741722635e-07, + "loss": 0.74347389, + "num_input_tokens_seen": 141439350, + "step": 6561, + "time_per_iteration": 3.4208054542541504 + }, + { + "auxiliary_loss_clip": 0.01183464, + "auxiliary_loss_mlp": 0.01027138, + "balance_loss_clip": 1.04451919, + "balance_loss_mlp": 1.02018762, + "epoch": 0.7890338483737149, + "flos": 29424629018880.0, + "grad_norm": 1.6914578917186245, + "language_loss": 0.74195993, + "learning_rate": 4.4883393737780985e-07, + "loss": 0.76406598, + "num_input_tokens_seen": 141460300, + "step": 6562, + "time_per_iteration": 3.569195032119751 + }, + { + "auxiliary_loss_clip": 0.01173353, + "auxiliary_loss_mlp": 0.01028103, + "balance_loss_clip": 1.04592931, + "balance_loss_mlp": 1.02095973, + "epoch": 0.789154091264354, + "flos": 19971576063360.0, + "grad_norm": 2.6927739441018876, + "language_loss": 0.78254169, + "learning_rate": 4.4834233264905254e-07, + "loss": 0.80455625, + "num_input_tokens_seen": 141477315, + "step": 6563, + "time_per_iteration": 2.6386008262634277 + }, + { + "auxiliary_loss_clip": 0.01184513, + "auxiliary_loss_mlp": 0.01024336, + "balance_loss_clip": 1.04023111, + "balance_loss_mlp": 1.01684952, + "epoch": 0.789274334154993, + "flos": 14537825216640.0, + "grad_norm": 2.716035972132253, + "language_loss": 0.71367776, + "learning_rate": 4.478509633055294e-07, + "loss": 0.73576629, + "num_input_tokens_seen": 141495025, + "step": 6564, + "time_per_iteration": 2.623721122741699 + }, + { + "auxiliary_loss_clip": 0.0119012, + "auxiliary_loss_mlp": 0.01030296, + "balance_loss_clip": 1.04543853, + "balance_loss_mlp": 1.02237475, + "epoch": 0.7893945770456322, + "flos": 21827403123840.0, + "grad_norm": 2.2937157999046907, + "language_loss": 0.7991972, + "learning_rate": 4.473598294217813e-07, + "loss": 0.8214013, + "num_input_tokens_seen": 141510450, + "step": 6565, + "time_per_iteration": 2.6421926021575928 + }, + { + "auxiliary_loss_clip": 0.01173634, + "auxiliary_loss_mlp": 0.01023336, + "balance_loss_clip": 1.04545414, + "balance_loss_mlp": 1.01656461, + "epoch": 0.7895148199362713, + "flos": 20740639184640.0, + "grad_norm": 2.052051875893698, + "language_loss": 0.71563965, + "learning_rate": 4.468689310723124e-07, + "loss": 0.73760939, + "num_input_tokens_seen": 141528265, + "step": 6566, + "time_per_iteration": 2.590073347091675 + }, + { + "auxiliary_loss_clip": 0.01200126, + "auxiliary_loss_mlp": 0.0102896, + "balance_loss_clip": 1.04398406, + "balance_loss_mlp": 1.02199793, + "epoch": 0.7896350628269103, + "flos": 16690669839360.0, + "grad_norm": 1.775199848034881, + "language_loss": 0.78659016, + "learning_rate": 4.463782683315913e-07, + "loss": 0.80888104, + "num_input_tokens_seen": 141547270, + "step": 6567, + "time_per_iteration": 2.669591188430786 + }, + { + "auxiliary_loss_clip": 0.01167212, + "auxiliary_loss_mlp": 0.01021473, + "balance_loss_clip": 1.04696381, + "balance_loss_mlp": 1.01445186, + "epoch": 0.7897553057175495, + "flos": 22638374438400.0, + "grad_norm": 1.6385714605844184, + "language_loss": 0.72978753, + "learning_rate": 4.458878412740523e-07, + "loss": 0.75167441, + "num_input_tokens_seen": 141566050, + "step": 6568, + "time_per_iteration": 2.5307164192199707 + }, + { + "auxiliary_loss_clip": 0.01171721, + "auxiliary_loss_mlp": 0.01022989, + "balance_loss_clip": 1.0445466, + "balance_loss_mlp": 1.01558053, + "epoch": 0.7898755486081885, + "flos": 14537573821440.0, + "grad_norm": 2.195475415786094, + "language_loss": 0.78314853, + "learning_rate": 4.453976499740919e-07, + "loss": 0.80509555, + "num_input_tokens_seen": 141583695, + "step": 6569, + "time_per_iteration": 2.60705828666687 + }, + { + "auxiliary_loss_clip": 0.01174204, + "auxiliary_loss_mlp": 0.01030312, + "balance_loss_clip": 1.04669833, + "balance_loss_mlp": 1.02237248, + "epoch": 0.7899957914988276, + "flos": 17238487138560.0, + "grad_norm": 1.7672181147706614, + "language_loss": 0.77702451, + "learning_rate": 4.4490769450607215e-07, + "loss": 0.7990697, + "num_input_tokens_seen": 141601320, + "step": 6570, + "time_per_iteration": 2.5261123180389404 + }, + { + "auxiliary_loss_clip": 0.0118751, + "auxiliary_loss_mlp": 0.01027346, + "balance_loss_clip": 1.03876054, + "balance_loss_mlp": 1.02004433, + "epoch": 0.7901160343894668, + "flos": 41279351086080.0, + "grad_norm": 4.6734164507503095, + "language_loss": 0.72793269, + "learning_rate": 4.4441797494431845e-07, + "loss": 0.7500813, + "num_input_tokens_seen": 141623125, + "step": 6571, + "time_per_iteration": 2.895033121109009 + }, + { + "auxiliary_loss_clip": 0.01176588, + "auxiliary_loss_mlp": 0.0102233, + "balance_loss_clip": 1.0477277, + "balance_loss_mlp": 1.01502252, + "epoch": 0.7902362772801058, + "flos": 16837005847680.0, + "grad_norm": 1.9046196800713582, + "language_loss": 0.77729893, + "learning_rate": 4.439284913631207e-07, + "loss": 0.79928809, + "num_input_tokens_seen": 141640335, + "step": 6572, + "time_per_iteration": 2.58984637260437 + }, + { + "auxiliary_loss_clip": 0.0119932, + "auxiliary_loss_mlp": 0.01023744, + "balance_loss_clip": 1.04667616, + "balance_loss_mlp": 1.01625156, + "epoch": 0.7903565201707449, + "flos": 27125987091840.0, + "grad_norm": 2.078590835339096, + "language_loss": 0.83911085, + "learning_rate": 4.434392438367347e-07, + "loss": 0.86134148, + "num_input_tokens_seen": 141659760, + "step": 6573, + "time_per_iteration": 2.760164499282837 + }, + { + "auxiliary_loss_clip": 0.01182924, + "auxiliary_loss_mlp": 0.01023124, + "balance_loss_clip": 1.04614449, + "balance_loss_mlp": 1.01579595, + "epoch": 0.790476763061384, + "flos": 31025167142400.0, + "grad_norm": 1.7390457594881292, + "language_loss": 0.74168706, + "learning_rate": 4.4295023243937677e-07, + "loss": 0.76374745, + "num_input_tokens_seen": 141679965, + "step": 6574, + "time_per_iteration": 2.656514883041382 + }, + { + "auxiliary_loss_clip": 0.01179357, + "auxiliary_loss_mlp": 0.0102732, + "balance_loss_clip": 1.04796791, + "balance_loss_mlp": 1.01967239, + "epoch": 0.7905970059520231, + "flos": 22089084681600.0, + "grad_norm": 2.3743069217852915, + "language_loss": 0.8045311, + "learning_rate": 4.4246145724523123e-07, + "loss": 0.82659781, + "num_input_tokens_seen": 141697710, + "step": 6575, + "time_per_iteration": 2.638209581375122 + }, + { + "auxiliary_loss_clip": 0.01194111, + "auxiliary_loss_mlp": 0.0102043, + "balance_loss_clip": 1.04440761, + "balance_loss_mlp": 1.01307189, + "epoch": 0.7907172488426621, + "flos": 20558141159040.0, + "grad_norm": 2.7775408082914375, + "language_loss": 0.77571392, + "learning_rate": 4.41972918328444e-07, + "loss": 0.79785931, + "num_input_tokens_seen": 141715145, + "step": 6576, + "time_per_iteration": 2.6436564922332764 + }, + { + "auxiliary_loss_clip": 0.01174662, + "auxiliary_loss_mlp": 0.01029121, + "balance_loss_clip": 1.04637575, + "balance_loss_mlp": 1.02173257, + "epoch": 0.7908374917333013, + "flos": 30081542901120.0, + "grad_norm": 2.1900254918695197, + "language_loss": 0.7706486, + "learning_rate": 4.4148461576312646e-07, + "loss": 0.79268646, + "num_input_tokens_seen": 141734810, + "step": 6577, + "time_per_iteration": 2.6712112426757812 + }, + { + "auxiliary_loss_clip": 0.01179505, + "auxiliary_loss_mlp": 0.01023074, + "balance_loss_clip": 1.04906702, + "balance_loss_mlp": 1.01661265, + "epoch": 0.7909577346239404, + "flos": 20996359084800.0, + "grad_norm": 1.4249932509909307, + "language_loss": 0.74464035, + "learning_rate": 4.4099654962335343e-07, + "loss": 0.76666617, + "num_input_tokens_seen": 141755260, + "step": 6578, + "time_per_iteration": 2.625883102416992 + }, + { + "auxiliary_loss_clip": 0.01194007, + "auxiliary_loss_mlp": 0.01027638, + "balance_loss_clip": 1.0476526, + "balance_loss_mlp": 1.02018785, + "epoch": 0.7910779775145794, + "flos": 26247935128320.0, + "grad_norm": 1.794965503907888, + "language_loss": 0.75203103, + "learning_rate": 4.405087199831636e-07, + "loss": 0.77424753, + "num_input_tokens_seen": 141775500, + "step": 6579, + "time_per_iteration": 2.695620536804199 + }, + { + "auxiliary_loss_clip": 0.0118872, + "auxiliary_loss_mlp": 0.01378501, + "balance_loss_clip": 1.04450834, + "balance_loss_mlp": 1.00015378, + "epoch": 0.7911982204052186, + "flos": 22564434291840.0, + "grad_norm": 2.398580533613748, + "language_loss": 0.67326951, + "learning_rate": 4.400211269165619e-07, + "loss": 0.69894171, + "num_input_tokens_seen": 141791955, + "step": 6580, + "time_per_iteration": 2.73203444480896 + }, + { + "auxiliary_loss_clip": 0.01169853, + "auxiliary_loss_mlp": 0.01023309, + "balance_loss_clip": 1.04930031, + "balance_loss_mlp": 1.01676106, + "epoch": 0.7913184632958576, + "flos": 23112538899840.0, + "grad_norm": 1.5188539745241845, + "language_loss": 0.76879823, + "learning_rate": 4.3953377049751416e-07, + "loss": 0.79072988, + "num_input_tokens_seen": 141812380, + "step": 6581, + "time_per_iteration": 2.593618631362915 + }, + { + "auxiliary_loss_clip": 0.01192617, + "auxiliary_loss_mlp": 0.01025662, + "balance_loss_clip": 1.04726446, + "balance_loss_mlp": 1.01903141, + "epoch": 0.7914387061864967, + "flos": 12311758719360.0, + "grad_norm": 2.410745509972357, + "language_loss": 0.78161341, + "learning_rate": 4.390466507999537e-07, + "loss": 0.80379617, + "num_input_tokens_seen": 141828130, + "step": 6582, + "time_per_iteration": 3.561107873916626 + }, + { + "auxiliary_loss_clip": 0.01192364, + "auxiliary_loss_mlp": 0.01029617, + "balance_loss_clip": 1.04226208, + "balance_loss_mlp": 1.02223206, + "epoch": 0.7915589490771359, + "flos": 17603267708160.0, + "grad_norm": 3.0777834214794977, + "language_loss": 0.76574302, + "learning_rate": 4.385597678977748e-07, + "loss": 0.78796285, + "num_input_tokens_seen": 141846965, + "step": 6583, + "time_per_iteration": 2.6998610496520996 + }, + { + "auxiliary_loss_clip": 0.01183577, + "auxiliary_loss_mlp": 0.01022922, + "balance_loss_clip": 1.04257631, + "balance_loss_mlp": 1.01545966, + "epoch": 0.7916791919677749, + "flos": 25591272641280.0, + "grad_norm": 1.7167403904350182, + "language_loss": 0.75638878, + "learning_rate": 4.3807312186483726e-07, + "loss": 0.77845377, + "num_input_tokens_seen": 141867685, + "step": 6584, + "time_per_iteration": 3.649811267852783 + }, + { + "auxiliary_loss_clip": 0.01176897, + "auxiliary_loss_mlp": 0.01028123, + "balance_loss_clip": 1.05045223, + "balance_loss_mlp": 1.02063656, + "epoch": 0.791799434858414, + "flos": 18844340474880.0, + "grad_norm": 2.2510343992928266, + "language_loss": 0.78645098, + "learning_rate": 4.375867127749655e-07, + "loss": 0.80850112, + "num_input_tokens_seen": 141885960, + "step": 6585, + "time_per_iteration": 2.641653537750244 + }, + { + "auxiliary_loss_clip": 0.01198815, + "auxiliary_loss_mlp": 0.01029376, + "balance_loss_clip": 1.04666543, + "balance_loss_mlp": 1.02236676, + "epoch": 0.7919196777490531, + "flos": 25812015672960.0, + "grad_norm": 1.8210714196325661, + "language_loss": 0.6733588, + "learning_rate": 4.3710054070194744e-07, + "loss": 0.69564074, + "num_input_tokens_seen": 141905655, + "step": 6586, + "time_per_iteration": 3.5810437202453613 + }, + { + "auxiliary_loss_clip": 0.01172296, + "auxiliary_loss_mlp": 0.01379083, + "balance_loss_clip": 1.04866874, + "balance_loss_mlp": 1.00018406, + "epoch": 0.7920399206396922, + "flos": 11947624594560.0, + "grad_norm": 3.418295567097442, + "language_loss": 0.66005528, + "learning_rate": 4.3661460571953455e-07, + "loss": 0.68556905, + "num_input_tokens_seen": 141922390, + "step": 6587, + "time_per_iteration": 2.591684579849243 + }, + { + "auxiliary_loss_clip": 0.01176396, + "auxiliary_loss_mlp": 0.01025918, + "balance_loss_clip": 1.04293776, + "balance_loss_mlp": 1.01868522, + "epoch": 0.7921601635303313, + "flos": 21579907438080.0, + "grad_norm": 1.774689470645247, + "language_loss": 0.68676823, + "learning_rate": 4.36128907901443e-07, + "loss": 0.70879138, + "num_input_tokens_seen": 141941985, + "step": 6588, + "time_per_iteration": 3.4562795162200928 + }, + { + "auxiliary_loss_clip": 0.01195958, + "auxiliary_loss_mlp": 0.01024586, + "balance_loss_clip": 1.04201102, + "balance_loss_mlp": 1.01768696, + "epoch": 0.7922804064209703, + "flos": 18113989236480.0, + "grad_norm": 2.820393126653126, + "language_loss": 0.72505432, + "learning_rate": 4.356434473213519e-07, + "loss": 0.7472598, + "num_input_tokens_seen": 141959435, + "step": 6589, + "time_per_iteration": 2.654092788696289 + }, + { + "auxiliary_loss_clip": 0.01187291, + "auxiliary_loss_mlp": 0.01023456, + "balance_loss_clip": 1.04676104, + "balance_loss_mlp": 1.01600504, + "epoch": 0.7924006493116095, + "flos": 21652806090240.0, + "grad_norm": 1.7447366209140691, + "language_loss": 0.79556984, + "learning_rate": 4.351582240529068e-07, + "loss": 0.81767732, + "num_input_tokens_seen": 141980265, + "step": 6590, + "time_per_iteration": 2.691513776779175 + }, + { + "auxiliary_loss_clip": 0.01089093, + "auxiliary_loss_mlp": 0.01002222, + "balance_loss_clip": 1.00988436, + "balance_loss_mlp": 1.00110722, + "epoch": 0.7925208922022485, + "flos": 64242755694720.0, + "grad_norm": 0.6797048662216247, + "language_loss": 0.58182847, + "learning_rate": 4.346732381697149e-07, + "loss": 0.60274166, + "num_input_tokens_seen": 142044395, + "step": 6591, + "time_per_iteration": 3.2721893787384033 + }, + { + "auxiliary_loss_clip": 0.01183501, + "auxiliary_loss_mlp": 0.01025449, + "balance_loss_clip": 1.04690838, + "balance_loss_mlp": 1.01767898, + "epoch": 0.7926411350928876, + "flos": 16941541403520.0, + "grad_norm": 1.8176464425311845, + "language_loss": 0.81017047, + "learning_rate": 4.3418848974534825e-07, + "loss": 0.83225995, + "num_input_tokens_seen": 142061335, + "step": 6592, + "time_per_iteration": 2.63669490814209 + }, + { + "auxiliary_loss_clip": 0.01198343, + "auxiliary_loss_mlp": 0.01027481, + "balance_loss_clip": 1.04430699, + "balance_loss_mlp": 1.0205338, + "epoch": 0.7927613779835267, + "flos": 34459987144320.0, + "grad_norm": 1.9120092130420174, + "language_loss": 0.68974435, + "learning_rate": 4.3370397885334276e-07, + "loss": 0.71200258, + "num_input_tokens_seen": 142081965, + "step": 6593, + "time_per_iteration": 2.7541682720184326 + }, + { + "auxiliary_loss_clip": 0.01166905, + "auxiliary_loss_mlp": 0.01027578, + "balance_loss_clip": 1.04416931, + "balance_loss_mlp": 1.01983583, + "epoch": 0.7928816208741658, + "flos": 18951174501120.0, + "grad_norm": 2.371817393770312, + "language_loss": 0.75321782, + "learning_rate": 4.3321970556719777e-07, + "loss": 0.7751627, + "num_input_tokens_seen": 142100260, + "step": 6594, + "time_per_iteration": 2.532299518585205 + }, + { + "auxiliary_loss_clip": 0.01168215, + "auxiliary_loss_mlp": 0.01025601, + "balance_loss_clip": 1.04668784, + "balance_loss_mlp": 1.01841259, + "epoch": 0.7930018637648049, + "flos": 18623022825600.0, + "grad_norm": 2.2557174348332607, + "language_loss": 0.71978092, + "learning_rate": 4.3273566996037856e-07, + "loss": 0.74171907, + "num_input_tokens_seen": 142116955, + "step": 6595, + "time_per_iteration": 2.4938206672668457 + }, + { + "auxiliary_loss_clip": 0.0118491, + "auxiliary_loss_mlp": 0.01024035, + "balance_loss_clip": 1.04311121, + "balance_loss_mlp": 1.01693654, + "epoch": 0.793122106655444, + "flos": 24530650824960.0, + "grad_norm": 2.522704949213356, + "language_loss": 0.80753684, + "learning_rate": 4.322518721063113e-07, + "loss": 0.82962632, + "num_input_tokens_seen": 142135505, + "step": 6596, + "time_per_iteration": 2.5842292308807373 + }, + { + "auxiliary_loss_clip": 0.0117786, + "auxiliary_loss_mlp": 0.0102736, + "balance_loss_clip": 1.04753828, + "balance_loss_mlp": 1.01974881, + "epoch": 0.7932423495460831, + "flos": 34421203434240.0, + "grad_norm": 1.9179086007526043, + "language_loss": 0.70291579, + "learning_rate": 4.3176831207838906e-07, + "loss": 0.72496802, + "num_input_tokens_seen": 142158915, + "step": 6597, + "time_per_iteration": 2.600175619125366 + }, + { + "auxiliary_loss_clip": 0.01178209, + "auxiliary_loss_mlp": 0.01026542, + "balance_loss_clip": 1.04949391, + "balance_loss_mlp": 1.01901948, + "epoch": 0.7933625924367221, + "flos": 26980333441920.0, + "grad_norm": 1.7108631574565893, + "language_loss": 0.74572384, + "learning_rate": 4.3128498994996685e-07, + "loss": 0.7677713, + "num_input_tokens_seen": 142178390, + "step": 6598, + "time_per_iteration": 2.666581153869629 + }, + { + "auxiliary_loss_clip": 0.01183667, + "auxiliary_loss_mlp": 0.01024129, + "balance_loss_clip": 1.04767489, + "balance_loss_mlp": 1.01656234, + "epoch": 0.7934828353273613, + "flos": 29568630643200.0, + "grad_norm": 2.0635898118113714, + "language_loss": 0.7126531, + "learning_rate": 4.308019057943646e-07, + "loss": 0.73473102, + "num_input_tokens_seen": 142200115, + "step": 6599, + "time_per_iteration": 2.6312830448150635 + }, + { + "auxiliary_loss_clip": 0.01213026, + "auxiliary_loss_mlp": 0.01024333, + "balance_loss_clip": 1.04408073, + "balance_loss_mlp": 1.0170728, + "epoch": 0.7936030782180004, + "flos": 28615381557120.0, + "grad_norm": 1.7432064497755635, + "language_loss": 0.74564683, + "learning_rate": 4.3031905968486535e-07, + "loss": 0.76802039, + "num_input_tokens_seen": 142220945, + "step": 6600, + "time_per_iteration": 2.866373300552368 + }, + { + "auxiliary_loss_clip": 0.01208243, + "auxiliary_loss_mlp": 0.0102999, + "balance_loss_clip": 1.04808915, + "balance_loss_mlp": 1.02310824, + "epoch": 0.7937233211086394, + "flos": 16392574869120.0, + "grad_norm": 2.845298431713241, + "language_loss": 0.68782264, + "learning_rate": 4.298364516947162e-07, + "loss": 0.71020502, + "num_input_tokens_seen": 142238175, + "step": 6601, + "time_per_iteration": 2.8154759407043457 + }, + { + "auxiliary_loss_clip": 0.01196592, + "auxiliary_loss_mlp": 0.01022974, + "balance_loss_clip": 1.03976607, + "balance_loss_mlp": 1.01568401, + "epoch": 0.7938435639992786, + "flos": 22013420682240.0, + "grad_norm": 7.259824487637824, + "language_loss": 0.65702534, + "learning_rate": 4.293540818971295e-07, + "loss": 0.67922097, + "num_input_tokens_seen": 142255980, + "step": 6602, + "time_per_iteration": 2.7059261798858643 + }, + { + "auxiliary_loss_clip": 0.01184294, + "auxiliary_loss_mlp": 0.01028129, + "balance_loss_clip": 1.04773378, + "balance_loss_mlp": 1.0206573, + "epoch": 0.7939638068899176, + "flos": 22197032029440.0, + "grad_norm": 2.1944821863257062, + "language_loss": 0.7682305, + "learning_rate": 4.2887195036527934e-07, + "loss": 0.79035473, + "num_input_tokens_seen": 142274785, + "step": 6603, + "time_per_iteration": 2.60182523727417 + }, + { + "auxiliary_loss_clip": 0.01165131, + "auxiliary_loss_mlp": 0.01026243, + "balance_loss_clip": 1.04201245, + "balance_loss_mlp": 1.01885819, + "epoch": 0.7940840497805567, + "flos": 17745186343680.0, + "grad_norm": 2.5751592769213123, + "language_loss": 0.73159671, + "learning_rate": 4.28390057172306e-07, + "loss": 0.75351042, + "num_input_tokens_seen": 142291290, + "step": 6604, + "time_per_iteration": 2.51995587348938 + }, + { + "auxiliary_loss_clip": 0.01188299, + "auxiliary_loss_mlp": 0.01029028, + "balance_loss_clip": 1.04159164, + "balance_loss_mlp": 1.02152371, + "epoch": 0.7942042926711959, + "flos": 23805435231360.0, + "grad_norm": 2.595000198209462, + "language_loss": 0.7265783, + "learning_rate": 4.279084023913111e-07, + "loss": 0.74875158, + "num_input_tokens_seen": 142309165, + "step": 6605, + "time_per_iteration": 2.701171398162842 + }, + { + "auxiliary_loss_clip": 0.01176819, + "auxiliary_loss_mlp": 0.01022225, + "balance_loss_clip": 1.04742086, + "balance_loss_mlp": 1.01474476, + "epoch": 0.7943245355618349, + "flos": 19244959839360.0, + "grad_norm": 1.7402377515862053, + "language_loss": 0.69369709, + "learning_rate": 4.2742698609536096e-07, + "loss": 0.71568751, + "num_input_tokens_seen": 142327475, + "step": 6606, + "time_per_iteration": 2.591660261154175 + }, + { + "auxiliary_loss_clip": 0.01189524, + "auxiliary_loss_mlp": 0.01026417, + "balance_loss_clip": 1.04584074, + "balance_loss_mlp": 1.01919329, + "epoch": 0.794444778452474, + "flos": 25007616547200.0, + "grad_norm": 2.463387752404835, + "language_loss": 0.78474283, + "learning_rate": 4.2694580835748706e-07, + "loss": 0.80690217, + "num_input_tokens_seen": 142347335, + "step": 6607, + "time_per_iteration": 3.590470314025879 + }, + { + "auxiliary_loss_clip": 0.0118611, + "auxiliary_loss_mlp": 0.01027727, + "balance_loss_clip": 1.04509735, + "balance_loss_mlp": 1.02026999, + "epoch": 0.7945650213431131, + "flos": 23221491828480.0, + "grad_norm": 2.051470550480012, + "language_loss": 0.74577367, + "learning_rate": 4.264648692506836e-07, + "loss": 0.76791203, + "num_input_tokens_seen": 142366125, + "step": 6608, + "time_per_iteration": 2.626291513442993 + }, + { + "auxiliary_loss_clip": 0.01179282, + "auxiliary_loss_mlp": 0.01029252, + "balance_loss_clip": 1.04303896, + "balance_loss_mlp": 1.02111614, + "epoch": 0.7946852642337522, + "flos": 26062887237120.0, + "grad_norm": 1.7371701757602582, + "language_loss": 0.72162104, + "learning_rate": 4.2598416884790824e-07, + "loss": 0.74370635, + "num_input_tokens_seen": 142385175, + "step": 6609, + "time_per_iteration": 2.6634483337402344 + }, + { + "auxiliary_loss_clip": 0.01196616, + "auxiliary_loss_mlp": 0.01026211, + "balance_loss_clip": 1.04586995, + "balance_loss_mlp": 1.01871276, + "epoch": 0.7948055071243912, + "flos": 23769704177280.0, + "grad_norm": 1.948155969086376, + "language_loss": 0.80738652, + "learning_rate": 4.255037072220828e-07, + "loss": 0.82961476, + "num_input_tokens_seen": 142406545, + "step": 6610, + "time_per_iteration": 2.6426384449005127 + }, + { + "auxiliary_loss_clip": 0.01165413, + "auxiliary_loss_mlp": 0.01026634, + "balance_loss_clip": 1.04555631, + "balance_loss_mlp": 1.01997924, + "epoch": 0.7949257500150304, + "flos": 21980814111360.0, + "grad_norm": 1.52395559607123, + "language_loss": 0.71461004, + "learning_rate": 4.2502348444609293e-07, + "loss": 0.73653054, + "num_input_tokens_seen": 142426165, + "step": 6611, + "time_per_iteration": 3.5324366092681885 + }, + { + "auxiliary_loss_clip": 0.01198268, + "auxiliary_loss_mlp": 0.01025731, + "balance_loss_clip": 1.03871047, + "balance_loss_mlp": 1.01858401, + "epoch": 0.7950459929056695, + "flos": 25774129802880.0, + "grad_norm": 2.676484789852508, + "language_loss": 0.69852412, + "learning_rate": 4.2454350059278844e-07, + "loss": 0.7207641, + "num_input_tokens_seen": 142447225, + "step": 6612, + "time_per_iteration": 3.633481025695801 + }, + { + "auxiliary_loss_clip": 0.01175663, + "auxiliary_loss_mlp": 0.01031377, + "balance_loss_clip": 1.04001117, + "balance_loss_mlp": 1.0242815, + "epoch": 0.7951662357963085, + "flos": 22158068751360.0, + "grad_norm": 1.7307383020181206, + "language_loss": 0.84537339, + "learning_rate": 4.240637557349824e-07, + "loss": 0.8674438, + "num_input_tokens_seen": 142464440, + "step": 6613, + "time_per_iteration": 2.6198441982269287 + }, + { + "auxiliary_loss_clip": 0.01170268, + "auxiliary_loss_mlp": 0.01025175, + "balance_loss_clip": 1.04280508, + "balance_loss_mlp": 1.01767659, + "epoch": 0.7952864786869477, + "flos": 24641938137600.0, + "grad_norm": 1.9386601074441554, + "language_loss": 0.66041666, + "learning_rate": 4.235842499454516e-07, + "loss": 0.68237114, + "num_input_tokens_seen": 142484355, + "step": 6614, + "time_per_iteration": 3.592477560043335 + }, + { + "auxiliary_loss_clip": 0.01187695, + "auxiliary_loss_mlp": 0.01025258, + "balance_loss_clip": 1.04575205, + "balance_loss_mlp": 1.01864791, + "epoch": 0.7954067215775867, + "flos": 21830922656640.0, + "grad_norm": 1.7037821854264663, + "language_loss": 0.83274591, + "learning_rate": 4.2310498329693687e-07, + "loss": 0.85487545, + "num_input_tokens_seen": 142505255, + "step": 6615, + "time_per_iteration": 2.7129902839660645 + }, + { + "auxiliary_loss_clip": 0.01180834, + "auxiliary_loss_mlp": 0.01023136, + "balance_loss_clip": 1.04714608, + "balance_loss_mlp": 1.01533937, + "epoch": 0.7955269644682258, + "flos": 24060652341120.0, + "grad_norm": 1.5314605982965872, + "language_loss": 0.80777562, + "learning_rate": 4.2262595586214164e-07, + "loss": 0.82981539, + "num_input_tokens_seen": 142526350, + "step": 6616, + "time_per_iteration": 2.6535861492156982 + }, + { + "auxiliary_loss_clip": 0.01184262, + "auxiliary_loss_mlp": 0.01026683, + "balance_loss_clip": 1.04901505, + "balance_loss_mlp": 1.01882744, + "epoch": 0.795647207358865, + "flos": 25010741030400.0, + "grad_norm": 1.5774007806718935, + "language_loss": 0.76806372, + "learning_rate": 4.221471677137358e-07, + "loss": 0.79017317, + "num_input_tokens_seen": 142547165, + "step": 6617, + "time_per_iteration": 2.6865036487579346 + }, + { + "auxiliary_loss_clip": 0.0117324, + "auxiliary_loss_mlp": 0.01024846, + "balance_loss_clip": 1.04401112, + "balance_loss_mlp": 1.01796436, + "epoch": 0.795767450249504, + "flos": 14648358343680.0, + "grad_norm": 2.000068828335967, + "language_loss": 0.70135748, + "learning_rate": 4.216686189243492e-07, + "loss": 0.72333837, + "num_input_tokens_seen": 142565955, + "step": 6618, + "time_per_iteration": 2.676116943359375 + }, + { + "auxiliary_loss_clip": 0.01193381, + "auxiliary_loss_mlp": 0.01024041, + "balance_loss_clip": 1.04383576, + "balance_loss_mlp": 1.01609874, + "epoch": 0.7958876931401431, + "flos": 18547897530240.0, + "grad_norm": 1.7778601861389152, + "language_loss": 0.72782463, + "learning_rate": 4.211903095665785e-07, + "loss": 0.74999887, + "num_input_tokens_seen": 142585340, + "step": 6619, + "time_per_iteration": 2.74676251411438 + }, + { + "auxiliary_loss_clip": 0.01173645, + "auxiliary_loss_mlp": 0.0102541, + "balance_loss_clip": 1.04667222, + "balance_loss_mlp": 1.01837039, + "epoch": 0.7960079360307821, + "flos": 21543960902400.0, + "grad_norm": 1.8030515652476562, + "language_loss": 0.75202608, + "learning_rate": 4.2071223971298277e-07, + "loss": 0.77401674, + "num_input_tokens_seen": 142602525, + "step": 6620, + "time_per_iteration": 2.5912652015686035 + }, + { + "auxiliary_loss_clip": 0.01176941, + "auxiliary_loss_mlp": 0.01023181, + "balance_loss_clip": 1.04512572, + "balance_loss_mlp": 1.01539648, + "epoch": 0.7961281789214213, + "flos": 25481745095040.0, + "grad_norm": 2.3917230296741017, + "language_loss": 0.61329323, + "learning_rate": 4.2023440943608433e-07, + "loss": 0.63529444, + "num_input_tokens_seen": 142622490, + "step": 6621, + "time_per_iteration": 2.666325807571411 + }, + { + "auxiliary_loss_clip": 0.01176334, + "auxiliary_loss_mlp": 0.01022243, + "balance_loss_clip": 1.04470372, + "balance_loss_mlp": 1.01530528, + "epoch": 0.7962484218120603, + "flos": 21944436612480.0, + "grad_norm": 1.6251245236961174, + "language_loss": 0.78508013, + "learning_rate": 4.1975681880837023e-07, + "loss": 0.8070659, + "num_input_tokens_seen": 142642495, + "step": 6622, + "time_per_iteration": 2.5973291397094727 + }, + { + "auxiliary_loss_clip": 0.01186546, + "auxiliary_loss_mlp": 0.01024411, + "balance_loss_clip": 1.03967679, + "balance_loss_mlp": 1.01687062, + "epoch": 0.7963686647026994, + "flos": 18876264687360.0, + "grad_norm": 2.099675269287625, + "language_loss": 0.82211328, + "learning_rate": 4.192794679022895e-07, + "loss": 0.8442229, + "num_input_tokens_seen": 142660820, + "step": 6623, + "time_per_iteration": 2.6907875537872314 + }, + { + "auxiliary_loss_clip": 0.01177758, + "auxiliary_loss_mlp": 0.01024111, + "balance_loss_clip": 1.04478741, + "balance_loss_mlp": 1.01708055, + "epoch": 0.7964889075933386, + "flos": 29716582763520.0, + "grad_norm": 1.873055445223322, + "language_loss": 0.71923786, + "learning_rate": 4.1880235679025743e-07, + "loss": 0.74125659, + "num_input_tokens_seen": 142680915, + "step": 6624, + "time_per_iteration": 2.6930031776428223 + }, + { + "auxiliary_loss_clip": 0.01211173, + "auxiliary_loss_mlp": 0.01031062, + "balance_loss_clip": 1.03900719, + "balance_loss_mlp": 1.02316177, + "epoch": 0.7966091504839776, + "flos": 29491458272640.0, + "grad_norm": 1.7947486550921101, + "language_loss": 0.63710177, + "learning_rate": 4.1832548554464986e-07, + "loss": 0.65952408, + "num_input_tokens_seen": 142699210, + "step": 6625, + "time_per_iteration": 2.833320140838623 + }, + { + "auxiliary_loss_clip": 0.0107087, + "auxiliary_loss_mlp": 0.01000625, + "balance_loss_clip": 1.01051116, + "balance_loss_mlp": 0.99949819, + "epoch": 0.7967293933746167, + "flos": 67288697101440.0, + "grad_norm": 0.7430004354267146, + "language_loss": 0.58767289, + "learning_rate": 4.178488542378098e-07, + "loss": 0.60838783, + "num_input_tokens_seen": 142756790, + "step": 6626, + "time_per_iteration": 3.1610500812530518 + }, + { + "auxiliary_loss_clip": 0.0117392, + "auxiliary_loss_mlp": 0.01026862, + "balance_loss_clip": 1.04890633, + "balance_loss_mlp": 1.01931596, + "epoch": 0.7968496362652558, + "flos": 25554679660800.0, + "grad_norm": 1.883394012795927, + "language_loss": 0.89086366, + "learning_rate": 4.173724629420401e-07, + "loss": 0.91287148, + "num_input_tokens_seen": 142778150, + "step": 6627, + "time_per_iteration": 2.6353111267089844 + }, + { + "auxiliary_loss_clip": 0.0118924, + "auxiliary_loss_mlp": 0.01024307, + "balance_loss_clip": 1.04420137, + "balance_loss_mlp": 1.01652288, + "epoch": 0.7969698791558949, + "flos": 14501088581760.0, + "grad_norm": 2.714473903844997, + "language_loss": 0.68418413, + "learning_rate": 4.168963117296087e-07, + "loss": 0.70631957, + "num_input_tokens_seen": 142795485, + "step": 6628, + "time_per_iteration": 2.665984869003296 + }, + { + "auxiliary_loss_clip": 0.01168432, + "auxiliary_loss_mlp": 0.0103216, + "balance_loss_clip": 1.04661381, + "balance_loss_mlp": 1.02426267, + "epoch": 0.797090122046534, + "flos": 22127545169280.0, + "grad_norm": 3.722805713508335, + "language_loss": 0.7593621, + "learning_rate": 4.1642040067274876e-07, + "loss": 0.78136808, + "num_input_tokens_seen": 142815155, + "step": 6629, + "time_per_iteration": 2.610027313232422 + }, + { + "auxiliary_loss_clip": 0.01191749, + "auxiliary_loss_mlp": 0.01023323, + "balance_loss_clip": 1.04503846, + "balance_loss_mlp": 1.01632535, + "epoch": 0.7972103649371731, + "flos": 19897671830400.0, + "grad_norm": 1.6292120512897355, + "language_loss": 0.72696477, + "learning_rate": 4.1594472984365493e-07, + "loss": 0.74911547, + "num_input_tokens_seen": 142833840, + "step": 6630, + "time_per_iteration": 2.738489866256714 + }, + { + "auxiliary_loss_clip": 0.01172461, + "auxiliary_loss_mlp": 0.01021734, + "balance_loss_clip": 1.04431152, + "balance_loss_mlp": 1.01447392, + "epoch": 0.7973306078278122, + "flos": 36058621847040.0, + "grad_norm": 2.0085560969757594, + "language_loss": 0.77744633, + "learning_rate": 4.154692993144862e-07, + "loss": 0.79938829, + "num_input_tokens_seen": 142853610, + "step": 6631, + "time_per_iteration": 2.7672877311706543 + }, + { + "auxiliary_loss_clip": 0.01169989, + "auxiliary_loss_mlp": 0.01378591, + "balance_loss_clip": 1.04840398, + "balance_loss_mlp": 1.00016999, + "epoch": 0.7974508507184512, + "flos": 21360600950400.0, + "grad_norm": 2.2677679438394844, + "language_loss": 0.71605259, + "learning_rate": 4.1499410915736476e-07, + "loss": 0.74153841, + "num_input_tokens_seen": 142872540, + "step": 6632, + "time_per_iteration": 2.6503422260284424 + }, + { + "auxiliary_loss_clip": 0.01076892, + "auxiliary_loss_mlp": 0.01002763, + "balance_loss_clip": 1.0113039, + "balance_loss_mlp": 1.00171399, + "epoch": 0.7975710936090904, + "flos": 68253115317120.0, + "grad_norm": 0.7962151825290252, + "language_loss": 0.6422708, + "learning_rate": 4.145191594443762e-07, + "loss": 0.66306734, + "num_input_tokens_seen": 142936895, + "step": 6633, + "time_per_iteration": 3.3787004947662354 + }, + { + "auxiliary_loss_clip": 0.01189575, + "auxiliary_loss_mlp": 0.01032756, + "balance_loss_clip": 1.04367781, + "balance_loss_mlp": 1.02456629, + "epoch": 0.7976913364997295, + "flos": 22492433479680.0, + "grad_norm": 2.087528926051694, + "language_loss": 0.70532548, + "learning_rate": 4.140444502475713e-07, + "loss": 0.72754878, + "num_input_tokens_seen": 142956445, + "step": 6634, + "time_per_iteration": 3.496570110321045 + }, + { + "auxiliary_loss_clip": 0.01170244, + "auxiliary_loss_mlp": 0.01029399, + "balance_loss_clip": 1.04224181, + "balance_loss_mlp": 1.02162647, + "epoch": 0.7978115793903685, + "flos": 15263220378240.0, + "grad_norm": 1.7610432394763473, + "language_loss": 0.69895208, + "learning_rate": 4.1356998163896216e-07, + "loss": 0.7209484, + "num_input_tokens_seen": 142973495, + "step": 6635, + "time_per_iteration": 2.614805221557617 + }, + { + "auxiliary_loss_clip": 0.01201505, + "auxiliary_loss_mlp": 0.01025266, + "balance_loss_clip": 1.04493213, + "balance_loss_mlp": 1.01812577, + "epoch": 0.7979318222810077, + "flos": 19719232041600.0, + "grad_norm": 1.9790648828300323, + "language_loss": 0.75019252, + "learning_rate": 4.130957536905255e-07, + "loss": 0.77246022, + "num_input_tokens_seen": 142991510, + "step": 6636, + "time_per_iteration": 3.609125852584839 + }, + { + "auxiliary_loss_clip": 0.01197523, + "auxiliary_loss_mlp": 0.01026606, + "balance_loss_clip": 1.0469954, + "balance_loss_mlp": 1.01860762, + "epoch": 0.7980520651716467, + "flos": 15560273854080.0, + "grad_norm": 3.706311948289481, + "language_loss": 0.70972872, + "learning_rate": 4.1262176647420134e-07, + "loss": 0.73196995, + "num_input_tokens_seen": 143009675, + "step": 6637, + "time_per_iteration": 2.62455677986145 + }, + { + "auxiliary_loss_clip": 0.01191959, + "auxiliary_loss_mlp": 0.010238, + "balance_loss_clip": 1.04627752, + "balance_loss_mlp": 1.01670063, + "epoch": 0.7981723080622858, + "flos": 22309432663680.0, + "grad_norm": 1.5869992116467089, + "language_loss": 0.79586518, + "learning_rate": 4.121480200618923e-07, + "loss": 0.81802273, + "num_input_tokens_seen": 143029330, + "step": 6638, + "time_per_iteration": 3.6088101863861084 + }, + { + "auxiliary_loss_clip": 0.01180061, + "auxiliary_loss_mlp": 0.01031761, + "balance_loss_clip": 1.04420733, + "balance_loss_mlp": 1.02462053, + "epoch": 0.798292550952925, + "flos": 22929573997440.0, + "grad_norm": 3.4981375675557818, + "language_loss": 0.8013376, + "learning_rate": 4.116745145254674e-07, + "loss": 0.82345581, + "num_input_tokens_seen": 143048865, + "step": 6639, + "time_per_iteration": 2.6387252807617188 + }, + { + "auxiliary_loss_clip": 0.0108354, + "auxiliary_loss_mlp": 0.01002972, + "balance_loss_clip": 1.01061463, + "balance_loss_mlp": 1.00187516, + "epoch": 0.798412793843564, + "flos": 64497936890880.0, + "grad_norm": 0.7662256411268438, + "language_loss": 0.58114511, + "learning_rate": 4.1120124993675476e-07, + "loss": 0.60201025, + "num_input_tokens_seen": 143113295, + "step": 6640, + "time_per_iteration": 4.223943710327148 + }, + { + "auxiliary_loss_clip": 0.01191421, + "auxiliary_loss_mlp": 0.01030189, + "balance_loss_clip": 1.04397798, + "balance_loss_mlp": 1.02189493, + "epoch": 0.7985330367342031, + "flos": 13586910514560.0, + "grad_norm": 1.9114112292312828, + "language_loss": 0.61633348, + "learning_rate": 4.107282263675498e-07, + "loss": 0.63854957, + "num_input_tokens_seen": 143130965, + "step": 6641, + "time_per_iteration": 2.6646525859832764 + }, + { + "auxiliary_loss_clip": 0.01084816, + "auxiliary_loss_mlp": 0.01373946, + "balance_loss_clip": 1.01230133, + "balance_loss_mlp": 0.99974984, + "epoch": 0.7986532796248422, + "flos": 67698797656320.0, + "grad_norm": 0.7698228774106549, + "language_loss": 0.52505529, + "learning_rate": 4.1025544388960907e-07, + "loss": 0.54964292, + "num_input_tokens_seen": 143192005, + "step": 6642, + "time_per_iteration": 3.2413196563720703 + }, + { + "auxiliary_loss_clip": 0.01173779, + "auxiliary_loss_mlp": 0.01025147, + "balance_loss_clip": 1.04585457, + "balance_loss_mlp": 1.01734817, + "epoch": 0.7987735225154813, + "flos": 22455373622400.0, + "grad_norm": 2.1747717309599084, + "language_loss": 0.7149089, + "learning_rate": 4.097829025746538e-07, + "loss": 0.73689824, + "num_input_tokens_seen": 143213550, + "step": 6643, + "time_per_iteration": 2.686357259750366 + }, + { + "auxiliary_loss_clip": 0.01075086, + "auxiliary_loss_mlp": 0.01003659, + "balance_loss_clip": 1.01078916, + "balance_loss_mlp": 1.00252652, + "epoch": 0.7988937654061203, + "flos": 68864098682880.0, + "grad_norm": 0.6596905525554955, + "language_loss": 0.6101886, + "learning_rate": 4.0931060249436757e-07, + "loss": 0.63097596, + "num_input_tokens_seen": 143277390, + "step": 6644, + "time_per_iteration": 3.2335422039031982 + }, + { + "auxiliary_loss_clip": 0.01176346, + "auxiliary_loss_mlp": 0.01029181, + "balance_loss_clip": 1.04691446, + "balance_loss_mlp": 1.02135468, + "epoch": 0.7990140082967595, + "flos": 20806893820800.0, + "grad_norm": 4.131203547362341, + "language_loss": 0.69331133, + "learning_rate": 4.088385437203978e-07, + "loss": 0.71536654, + "num_input_tokens_seen": 143294400, + "step": 6645, + "time_per_iteration": 2.63961124420166 + }, + { + "auxiliary_loss_clip": 0.0116904, + "auxiliary_loss_mlp": 0.01022081, + "balance_loss_clip": 1.04696918, + "balance_loss_mlp": 1.01495826, + "epoch": 0.7991342511873986, + "flos": 18985289443200.0, + "grad_norm": 2.03584441657714, + "language_loss": 0.77426004, + "learning_rate": 4.083667263243564e-07, + "loss": 0.79617131, + "num_input_tokens_seen": 143312745, + "step": 6646, + "time_per_iteration": 2.6096558570861816 + }, + { + "auxiliary_loss_clip": 0.01173367, + "auxiliary_loss_mlp": 0.0102784, + "balance_loss_clip": 1.04635096, + "balance_loss_mlp": 1.02037776, + "epoch": 0.7992544940780376, + "flos": 20816805974400.0, + "grad_norm": 1.6436115830897984, + "language_loss": 0.71953112, + "learning_rate": 4.0789515037781653e-07, + "loss": 0.74154317, + "num_input_tokens_seen": 143333470, + "step": 6647, + "time_per_iteration": 2.641554117202759 + }, + { + "auxiliary_loss_clip": 0.01181226, + "auxiliary_loss_mlp": 0.01027672, + "balance_loss_clip": 1.04639423, + "balance_loss_mlp": 1.0195713, + "epoch": 0.7993747369686768, + "flos": 12640772321280.0, + "grad_norm": 2.085093732643389, + "language_loss": 0.8276912, + "learning_rate": 4.0742381595231755e-07, + "loss": 0.8497802, + "num_input_tokens_seen": 143350195, + "step": 6648, + "time_per_iteration": 2.6040217876434326 + }, + { + "auxiliary_loss_clip": 0.01198966, + "auxiliary_loss_mlp": 0.01027477, + "balance_loss_clip": 1.04352713, + "balance_loss_mlp": 1.01989794, + "epoch": 0.7994949798593158, + "flos": 20078769225600.0, + "grad_norm": 2.3845055150044954, + "language_loss": 0.78234285, + "learning_rate": 4.06952723119359e-07, + "loss": 0.80460733, + "num_input_tokens_seen": 143370070, + "step": 6649, + "time_per_iteration": 2.7111332416534424 + }, + { + "auxiliary_loss_clip": 0.01177272, + "auxiliary_loss_mlp": 0.01023968, + "balance_loss_clip": 1.04474545, + "balance_loss_mlp": 1.01639509, + "epoch": 0.7996152227499549, + "flos": 38654209509120.0, + "grad_norm": 2.3945351769918255, + "language_loss": 0.67277026, + "learning_rate": 4.0648187195040504e-07, + "loss": 0.69478267, + "num_input_tokens_seen": 143392275, + "step": 6650, + "time_per_iteration": 2.822180986404419 + }, + { + "auxiliary_loss_clip": 0.01070515, + "auxiliary_loss_mlp": 0.01000851, + "balance_loss_clip": 1.0102489, + "balance_loss_mlp": 0.99972481, + "epoch": 0.799735465640594, + "flos": 70243821947520.0, + "grad_norm": 0.809557226809407, + "language_loss": 0.6759876, + "learning_rate": 4.060112625168848e-07, + "loss": 0.69670129, + "num_input_tokens_seen": 143457385, + "step": 6651, + "time_per_iteration": 3.3107452392578125 + }, + { + "auxiliary_loss_clip": 0.011706, + "auxiliary_loss_mlp": 0.01024331, + "balance_loss_clip": 1.0487982, + "balance_loss_mlp": 1.01694632, + "epoch": 0.7998557085312331, + "flos": 24240995550720.0, + "grad_norm": 1.878179942785847, + "language_loss": 0.73971301, + "learning_rate": 4.055408948901886e-07, + "loss": 0.7616623, + "num_input_tokens_seen": 143478785, + "step": 6652, + "time_per_iteration": 2.610583782196045 + }, + { + "auxiliary_loss_clip": 0.01183777, + "auxiliary_loss_mlp": 0.0102504, + "balance_loss_clip": 1.04821038, + "balance_loss_mlp": 1.01738667, + "epoch": 0.7999759514218722, + "flos": 27564025449600.0, + "grad_norm": 2.040411237300802, + "language_loss": 0.71417493, + "learning_rate": 4.050707691416708e-07, + "loss": 0.73626304, + "num_input_tokens_seen": 143500095, + "step": 6653, + "time_per_iteration": 2.6376469135284424 + }, + { + "auxiliary_loss_clip": 0.01070031, + "auxiliary_loss_mlp": 0.01000977, + "balance_loss_clip": 1.00969911, + "balance_loss_mlp": 0.99987471, + "epoch": 0.8000961943125112, + "flos": 67337428878720.0, + "grad_norm": 0.6919983863868606, + "language_loss": 0.59800345, + "learning_rate": 4.046008853426495e-07, + "loss": 0.6187135, + "num_input_tokens_seen": 143563410, + "step": 6654, + "time_per_iteration": 3.286766290664673 + }, + { + "auxiliary_loss_clip": 0.01191053, + "auxiliary_loss_mlp": 0.01022951, + "balance_loss_clip": 1.04332805, + "balance_loss_mlp": 1.0153513, + "epoch": 0.8002164372031504, + "flos": 28733815676160.0, + "grad_norm": 1.7799164009124988, + "language_loss": 0.62179065, + "learning_rate": 4.0413124356440464e-07, + "loss": 0.64393067, + "num_input_tokens_seen": 143587455, + "step": 6655, + "time_per_iteration": 2.800647735595703 + }, + { + "auxiliary_loss_clip": 0.0120461, + "auxiliary_loss_mlp": 0.01023961, + "balance_loss_clip": 1.0407815, + "balance_loss_mlp": 1.01657856, + "epoch": 0.8003366800937894, + "flos": 17639429725440.0, + "grad_norm": 2.265879698454381, + "language_loss": 0.82592767, + "learning_rate": 4.0366184387818223e-07, + "loss": 0.84821337, + "num_input_tokens_seen": 143605915, + "step": 6656, + "time_per_iteration": 2.696911096572876 + }, + { + "auxiliary_loss_clip": 0.0117386, + "auxiliary_loss_mlp": 0.01028957, + "balance_loss_clip": 1.04862547, + "balance_loss_mlp": 1.02088702, + "epoch": 0.8004569229844285, + "flos": 25995303797760.0, + "grad_norm": 1.7468174574396393, + "language_loss": 0.85273582, + "learning_rate": 4.0319268635518797e-07, + "loss": 0.87476397, + "num_input_tokens_seen": 143626490, + "step": 6657, + "time_per_iteration": 307.4063341617584 + }, + { + "auxiliary_loss_clip": 0.011785, + "auxiliary_loss_mlp": 0.01022451, + "balance_loss_clip": 1.04580951, + "balance_loss_mlp": 1.01489854, + "epoch": 0.8005771658750677, + "flos": 20812352688000.0, + "grad_norm": 1.7229570013442161, + "language_loss": 0.75044245, + "learning_rate": 4.027237710665943e-07, + "loss": 0.772452, + "num_input_tokens_seen": 143644955, + "step": 6658, + "time_per_iteration": 2.7492990493774414 + }, + { + "auxiliary_loss_clip": 0.01197859, + "auxiliary_loss_mlp": 0.01025913, + "balance_loss_clip": 1.0410738, + "balance_loss_mlp": 1.01853132, + "epoch": 0.8006974087657067, + "flos": 25812626204160.0, + "grad_norm": 1.9863399978797802, + "language_loss": 0.69342321, + "learning_rate": 4.022550980835344e-07, + "loss": 0.71566093, + "num_input_tokens_seen": 143667200, + "step": 6659, + "time_per_iteration": 3.8513407707214355 + }, + { + "auxiliary_loss_clip": 0.01193242, + "auxiliary_loss_mlp": 0.01024432, + "balance_loss_clip": 1.0400908, + "balance_loss_mlp": 1.01702321, + "epoch": 0.8008176516563458, + "flos": 17164690646400.0, + "grad_norm": 1.9692029496979737, + "language_loss": 0.79713416, + "learning_rate": 4.017866674771051e-07, + "loss": 0.8193109, + "num_input_tokens_seen": 143684685, + "step": 6660, + "time_per_iteration": 2.8338260650634766 + }, + { + "auxiliary_loss_clip": 0.01193024, + "auxiliary_loss_mlp": 0.01025218, + "balance_loss_clip": 1.03864563, + "balance_loss_mlp": 1.01783311, + "epoch": 0.8009378945469849, + "flos": 24207311571840.0, + "grad_norm": 2.134610572973598, + "language_loss": 0.74740505, + "learning_rate": 4.013184793183688e-07, + "loss": 0.76958746, + "num_input_tokens_seen": 143706780, + "step": 6661, + "time_per_iteration": 2.8819313049316406 + }, + { + "auxiliary_loss_clip": 0.01175612, + "auxiliary_loss_mlp": 0.0102282, + "balance_loss_clip": 1.04470372, + "balance_loss_mlp": 1.01568866, + "epoch": 0.801058137437624, + "flos": 19787318271360.0, + "grad_norm": 1.7693287857636495, + "language_loss": 0.72420835, + "learning_rate": 4.008505336783472e-07, + "loss": 0.74619269, + "num_input_tokens_seen": 143724505, + "step": 6662, + "time_per_iteration": 3.624565362930298 + }, + { + "auxiliary_loss_clip": 0.01167499, + "auxiliary_loss_mlp": 0.01022403, + "balance_loss_clip": 1.04416764, + "balance_loss_mlp": 1.01533937, + "epoch": 0.801178380328263, + "flos": 18659400324480.0, + "grad_norm": 1.7749257457562904, + "language_loss": 0.80549562, + "learning_rate": 4.003828306280284e-07, + "loss": 0.8273946, + "num_input_tokens_seen": 143742180, + "step": 6663, + "time_per_iteration": 2.7068324089050293 + }, + { + "auxiliary_loss_clip": 0.01178441, + "auxiliary_loss_mlp": 0.01023138, + "balance_loss_clip": 1.04601097, + "balance_loss_mlp": 1.01629221, + "epoch": 0.8012986232189022, + "flos": 15706573948800.0, + "grad_norm": 1.6633821460830123, + "language_loss": 0.77802742, + "learning_rate": 3.999153702383626e-07, + "loss": 0.80004323, + "num_input_tokens_seen": 143760070, + "step": 6664, + "time_per_iteration": 3.5669264793395996 + }, + { + "auxiliary_loss_clip": 0.01180454, + "auxiliary_loss_mlp": 0.01026286, + "balance_loss_clip": 1.04453087, + "balance_loss_mlp": 1.01921344, + "epoch": 0.8014188661095413, + "flos": 28584139703040.0, + "grad_norm": 1.7305170711800884, + "language_loss": 0.7384125, + "learning_rate": 3.9944815258026263e-07, + "loss": 0.76047993, + "num_input_tokens_seen": 143781890, + "step": 6665, + "time_per_iteration": 3.813135862350464 + }, + { + "auxiliary_loss_clip": 0.0118296, + "auxiliary_loss_mlp": 0.01025201, + "balance_loss_clip": 1.04843211, + "balance_loss_mlp": 1.01768196, + "epoch": 0.8015391090001803, + "flos": 29310360877440.0, + "grad_norm": 1.6701914856006665, + "language_loss": 0.82966387, + "learning_rate": 3.989811777246057e-07, + "loss": 0.85174549, + "num_input_tokens_seen": 143802060, + "step": 6666, + "time_per_iteration": 2.753694534301758 + }, + { + "auxiliary_loss_clip": 0.01061686, + "auxiliary_loss_mlp": 0.01002099, + "balance_loss_clip": 1.00987649, + "balance_loss_mlp": 1.00104356, + "epoch": 0.8016593518908195, + "flos": 70397340675840.0, + "grad_norm": 0.850182755428802, + "language_loss": 0.66203392, + "learning_rate": 3.985144457422305e-07, + "loss": 0.68267179, + "num_input_tokens_seen": 143856345, + "step": 6667, + "time_per_iteration": 3.1760482788085938 + }, + { + "auxiliary_loss_clip": 0.01170133, + "auxiliary_loss_mlp": 0.01024977, + "balance_loss_clip": 1.04866433, + "balance_loss_mlp": 1.01741302, + "epoch": 0.8017795947814585, + "flos": 26026114688640.0, + "grad_norm": 2.126673019196033, + "language_loss": 0.76925009, + "learning_rate": 3.9804795670394096e-07, + "loss": 0.79120117, + "num_input_tokens_seen": 143876470, + "step": 6668, + "time_per_iteration": 2.866560459136963 + }, + { + "auxiliary_loss_clip": 0.01177397, + "auxiliary_loss_mlp": 0.01023349, + "balance_loss_clip": 1.0449543, + "balance_loss_mlp": 1.01630664, + "epoch": 0.8018998376720976, + "flos": 22087181260800.0, + "grad_norm": 1.5356172715519476, + "language_loss": 0.70500463, + "learning_rate": 3.975817106805022e-07, + "loss": 0.72701204, + "num_input_tokens_seen": 143895170, + "step": 6669, + "time_per_iteration": 2.891523838043213 + }, + { + "auxiliary_loss_clip": 0.01194154, + "auxiliary_loss_mlp": 0.0102359, + "balance_loss_clip": 1.04303908, + "balance_loss_mlp": 1.0162046, + "epoch": 0.8020200805627368, + "flos": 34568545023360.0, + "grad_norm": 1.9575795473990625, + "language_loss": 0.65076447, + "learning_rate": 3.97115707742645e-07, + "loss": 0.67294192, + "num_input_tokens_seen": 143915845, + "step": 6670, + "time_per_iteration": 2.9130656719207764 + }, + { + "auxiliary_loss_clip": 0.01189812, + "auxiliary_loss_mlp": 0.01027803, + "balance_loss_clip": 1.04666185, + "balance_loss_mlp": 1.02057552, + "epoch": 0.8021403234533758, + "flos": 20120354196480.0, + "grad_norm": 2.2652028414978282, + "language_loss": 0.6521287, + "learning_rate": 3.966499479610599e-07, + "loss": 0.67430484, + "num_input_tokens_seen": 143933940, + "step": 6671, + "time_per_iteration": 2.745119333267212 + }, + { + "auxiliary_loss_clip": 0.01193452, + "auxiliary_loss_mlp": 0.0102572, + "balance_loss_clip": 1.0472126, + "balance_loss_mlp": 1.01881766, + "epoch": 0.8022605663440149, + "flos": 27746200252800.0, + "grad_norm": 1.9106705792142233, + "language_loss": 0.6540482, + "learning_rate": 3.9618443140640225e-07, + "loss": 0.67623991, + "num_input_tokens_seen": 143952850, + "step": 6672, + "time_per_iteration": 2.8048222064971924 + }, + { + "auxiliary_loss_clip": 0.0111031, + "auxiliary_loss_mlp": 0.01001628, + "balance_loss_clip": 1.00928819, + "balance_loss_mlp": 1.00054348, + "epoch": 0.802380809234654, + "flos": 60244998768000.0, + "grad_norm": 0.6815217992012726, + "language_loss": 0.51407731, + "learning_rate": 3.957191581492918e-07, + "loss": 0.53519666, + "num_input_tokens_seen": 144013610, + "step": 6673, + "time_per_iteration": 3.4538447856903076 + }, + { + "auxiliary_loss_clip": 0.01180572, + "auxiliary_loss_mlp": 0.01024162, + "balance_loss_clip": 1.04354179, + "balance_loss_mlp": 1.01653552, + "epoch": 0.8025010521252931, + "flos": 15080722352640.0, + "grad_norm": 4.436350549764229, + "language_loss": 0.71162289, + "learning_rate": 3.952541282603097e-07, + "loss": 0.73367023, + "num_input_tokens_seen": 144028715, + "step": 6674, + "time_per_iteration": 2.856165647506714 + }, + { + "auxiliary_loss_clip": 0.01176903, + "auxiliary_loss_mlp": 0.01029505, + "balance_loss_clip": 1.04702187, + "balance_loss_mlp": 1.02193213, + "epoch": 0.8026212950159322, + "flos": 22163527618560.0, + "grad_norm": 1.8613093563642165, + "language_loss": 0.83691883, + "learning_rate": 3.9478934181000013e-07, + "loss": 0.85898292, + "num_input_tokens_seen": 144048740, + "step": 6675, + "time_per_iteration": 2.684737205505371 + }, + { + "auxiliary_loss_clip": 0.01172193, + "auxiliary_loss_mlp": 0.01029109, + "balance_loss_clip": 1.04813039, + "balance_loss_mlp": 1.02130365, + "epoch": 0.8027415379065713, + "flos": 17675986792320.0, + "grad_norm": 2.3486319741332347, + "language_loss": 0.84135413, + "learning_rate": 3.943247988688714e-07, + "loss": 0.86336708, + "num_input_tokens_seen": 144067435, + "step": 6676, + "time_per_iteration": 2.7193856239318848 + }, + { + "auxiliary_loss_clip": 0.01177261, + "auxiliary_loss_mlp": 0.01024726, + "balance_loss_clip": 1.04523134, + "balance_loss_mlp": 1.01766038, + "epoch": 0.8028617807972104, + "flos": 21979593048960.0, + "grad_norm": 2.9300497984366154, + "language_loss": 0.72121263, + "learning_rate": 3.938604995073933e-07, + "loss": 0.74323255, + "num_input_tokens_seen": 144085905, + "step": 6677, + "time_per_iteration": 2.677034616470337 + }, + { + "auxiliary_loss_clip": 0.01188256, + "auxiliary_loss_mlp": 0.01027429, + "balance_loss_clip": 1.04415131, + "balance_loss_mlp": 1.0194304, + "epoch": 0.8029820236878494, + "flos": 26428457905920.0, + "grad_norm": 1.6382563319557568, + "language_loss": 0.65123481, + "learning_rate": 3.9339644379600157e-07, + "loss": 0.67339164, + "num_input_tokens_seen": 144105735, + "step": 6678, + "time_per_iteration": 2.689809799194336 + }, + { + "auxiliary_loss_clip": 0.01179801, + "auxiliary_loss_mlp": 0.01030123, + "balance_loss_clip": 1.0477246, + "balance_loss_mlp": 1.02265787, + "epoch": 0.8031022665784886, + "flos": 17676489582720.0, + "grad_norm": 2.0556143756131497, + "language_loss": 0.71000493, + "learning_rate": 3.929326318050907e-07, + "loss": 0.73210418, + "num_input_tokens_seen": 144123405, + "step": 6679, + "time_per_iteration": 2.674198865890503 + }, + { + "auxiliary_loss_clip": 0.01164935, + "auxiliary_loss_mlp": 0.01019718, + "balance_loss_clip": 1.04483175, + "balance_loss_mlp": 1.01245832, + "epoch": 0.8032225094691277, + "flos": 15450279431040.0, + "grad_norm": 1.918581955357316, + "language_loss": 0.79030538, + "learning_rate": 3.924690636050225e-07, + "loss": 0.81215191, + "num_input_tokens_seen": 144140815, + "step": 6680, + "time_per_iteration": 2.611178159713745 + }, + { + "auxiliary_loss_clip": 0.01178709, + "auxiliary_loss_mlp": 0.01024566, + "balance_loss_clip": 1.04655576, + "balance_loss_mlp": 1.01698458, + "epoch": 0.8033427523597667, + "flos": 26179202453760.0, + "grad_norm": 2.1706352359625427, + "language_loss": 0.7298032, + "learning_rate": 3.9200573926611915e-07, + "loss": 0.75183594, + "num_input_tokens_seen": 144162230, + "step": 6681, + "time_per_iteration": 2.6927876472473145 + }, + { + "auxiliary_loss_clip": 0.01175683, + "auxiliary_loss_mlp": 0.01023255, + "balance_loss_clip": 1.04761171, + "balance_loss_mlp": 1.01578617, + "epoch": 0.8034629952504058, + "flos": 21324905809920.0, + "grad_norm": 1.8298362586492758, + "language_loss": 0.72961789, + "learning_rate": 3.9154265885866613e-07, + "loss": 0.7516073, + "num_input_tokens_seen": 144181540, + "step": 6682, + "time_per_iteration": 2.6644155979156494 + }, + { + "auxiliary_loss_clip": 0.01175798, + "auxiliary_loss_mlp": 0.01028058, + "balance_loss_clip": 1.0468384, + "balance_loss_mlp": 1.02025533, + "epoch": 0.8035832381410449, + "flos": 21651585027840.0, + "grad_norm": 2.822590207695377, + "language_loss": 0.74665624, + "learning_rate": 3.9107982245291394e-07, + "loss": 0.76869476, + "num_input_tokens_seen": 144199665, + "step": 6683, + "time_per_iteration": 2.6910362243652344 + }, + { + "auxiliary_loss_clip": 0.01197476, + "auxiliary_loss_mlp": 0.01026663, + "balance_loss_clip": 1.04704261, + "balance_loss_mlp": 1.01887238, + "epoch": 0.803703481031684, + "flos": 20518818744960.0, + "grad_norm": 2.518874754802626, + "language_loss": 0.77406859, + "learning_rate": 3.9061723011907245e-07, + "loss": 0.79630995, + "num_input_tokens_seen": 144219020, + "step": 6684, + "time_per_iteration": 2.8464081287384033 + }, + { + "auxiliary_loss_clip": 0.01183285, + "auxiliary_loss_mlp": 0.01026404, + "balance_loss_clip": 1.0442791, + "balance_loss_mlp": 1.01870561, + "epoch": 0.803823723922323, + "flos": 22854807838080.0, + "grad_norm": 1.6550360634541825, + "language_loss": 0.794191, + "learning_rate": 3.901548819273179e-07, + "loss": 0.81628788, + "num_input_tokens_seen": 144239035, + "step": 6685, + "time_per_iteration": 2.726590871810913 + }, + { + "auxiliary_loss_clip": 0.0117937, + "auxiliary_loss_mlp": 0.0102299, + "balance_loss_clip": 1.0480485, + "balance_loss_mlp": 1.01541138, + "epoch": 0.8039439668129622, + "flos": 21362145235200.0, + "grad_norm": 2.2198293643166984, + "language_loss": 0.69319504, + "learning_rate": 3.896927779477881e-07, + "loss": 0.71521866, + "num_input_tokens_seen": 144258295, + "step": 6686, + "time_per_iteration": 3.641242742538452 + }, + { + "auxiliary_loss_clip": 0.01196881, + "auxiliary_loss_mlp": 0.01027436, + "balance_loss_clip": 1.04406667, + "balance_loss_mlp": 1.02070665, + "epoch": 0.8040642097036013, + "flos": 23802382575360.0, + "grad_norm": 1.9997926554895902, + "language_loss": 0.67022091, + "learning_rate": 3.892309182505833e-07, + "loss": 0.69246411, + "num_input_tokens_seen": 144276110, + "step": 6687, + "time_per_iteration": 2.8539016246795654 + }, + { + "auxiliary_loss_clip": 0.01167834, + "auxiliary_loss_mlp": 0.01024161, + "balance_loss_clip": 1.04604912, + "balance_loss_mlp": 1.01674008, + "epoch": 0.8041844525942403, + "flos": 25922046009600.0, + "grad_norm": 2.0467995044983316, + "language_loss": 0.86144149, + "learning_rate": 3.887693029057675e-07, + "loss": 0.8833614, + "num_input_tokens_seen": 144295620, + "step": 6688, + "time_per_iteration": 3.576885223388672 + }, + { + "auxiliary_loss_clip": 0.01188573, + "auxiliary_loss_mlp": 0.01029227, + "balance_loss_clip": 1.04509163, + "balance_loss_mlp": 1.02238989, + "epoch": 0.8043046954848795, + "flos": 25191120153600.0, + "grad_norm": 1.7033613585774283, + "language_loss": 0.81551623, + "learning_rate": 3.8830793198336684e-07, + "loss": 0.83769429, + "num_input_tokens_seen": 144315210, + "step": 6689, + "time_per_iteration": 2.781257152557373 + }, + { + "auxiliary_loss_clip": 0.01183199, + "auxiliary_loss_mlp": 0.01026595, + "balance_loss_clip": 1.04733276, + "balance_loss_mlp": 1.01907849, + "epoch": 0.8044249383755185, + "flos": 41719185123840.0, + "grad_norm": 1.7111153680562978, + "language_loss": 0.70232201, + "learning_rate": 3.878468055533721e-07, + "loss": 0.72441995, + "num_input_tokens_seen": 144337750, + "step": 6690, + "time_per_iteration": 3.722419500350952 + }, + { + "auxiliary_loss_clip": 0.01204101, + "auxiliary_loss_mlp": 0.01026163, + "balance_loss_clip": 1.04607522, + "balance_loss_mlp": 1.01883781, + "epoch": 0.8045451812661576, + "flos": 20631434860800.0, + "grad_norm": 2.726487110104339, + "language_loss": 0.8479048, + "learning_rate": 3.8738592368573464e-07, + "loss": 0.87020743, + "num_input_tokens_seen": 144355305, + "step": 6691, + "time_per_iteration": 4.169808626174927 + }, + { + "auxiliary_loss_clip": 0.01183208, + "auxiliary_loss_mlp": 0.01026137, + "balance_loss_clip": 1.04287314, + "balance_loss_mlp": 1.01815021, + "epoch": 0.8046654241567968, + "flos": 29711806254720.0, + "grad_norm": 1.9553449670526881, + "language_loss": 0.88148975, + "learning_rate": 3.8692528645037137e-07, + "loss": 0.90358329, + "num_input_tokens_seen": 144374485, + "step": 6692, + "time_per_iteration": 2.816972017288208 + }, + { + "auxiliary_loss_clip": 0.01168891, + "auxiliary_loss_mlp": 0.01025415, + "balance_loss_clip": 1.0477829, + "balance_loss_mlp": 1.01792264, + "epoch": 0.8047856670474358, + "flos": 17671389851520.0, + "grad_norm": 2.775963079016379, + "language_loss": 0.77278733, + "learning_rate": 3.8646489391715907e-07, + "loss": 0.79473037, + "num_input_tokens_seen": 144388780, + "step": 6693, + "time_per_iteration": 2.7103848457336426 + }, + { + "auxiliary_loss_clip": 0.01187067, + "auxiliary_loss_mlp": 0.01026434, + "balance_loss_clip": 1.04488492, + "balance_loss_mlp": 1.01884019, + "epoch": 0.8049059099380749, + "flos": 17120699464320.0, + "grad_norm": 2.874523221546739, + "language_loss": 0.8779974, + "learning_rate": 3.8600474615593903e-07, + "loss": 0.90013236, + "num_input_tokens_seen": 144403395, + "step": 6694, + "time_per_iteration": 2.730607032775879 + }, + { + "auxiliary_loss_clip": 0.01101235, + "auxiliary_loss_mlp": 0.01002475, + "balance_loss_clip": 1.01047611, + "balance_loss_mlp": 1.00145578, + "epoch": 0.805026152828714, + "flos": 62212903240320.0, + "grad_norm": 0.7872742303112844, + "language_loss": 0.59724963, + "learning_rate": 3.8554484323651605e-07, + "loss": 0.61828673, + "num_input_tokens_seen": 144465265, + "step": 6695, + "time_per_iteration": 3.364107131958008 + }, + { + "auxiliary_loss_clip": 0.01176855, + "auxiliary_loss_mlp": 0.01378612, + "balance_loss_clip": 1.04757357, + "balance_loss_mlp": 1.00018835, + "epoch": 0.8051463957193531, + "flos": 21688608971520.0, + "grad_norm": 1.618805441405211, + "language_loss": 0.79333866, + "learning_rate": 3.85085185228657e-07, + "loss": 0.81889331, + "num_input_tokens_seen": 144484235, + "step": 6696, + "time_per_iteration": 2.703706979751587 + }, + { + "auxiliary_loss_clip": 0.01178696, + "auxiliary_loss_mlp": 0.01024365, + "balance_loss_clip": 1.04270649, + "balance_loss_mlp": 1.01751316, + "epoch": 0.8052666386099921, + "flos": 32051458535040.0, + "grad_norm": 2.040498312791058, + "language_loss": 0.7323246, + "learning_rate": 3.8462577220209114e-07, + "loss": 0.75435525, + "num_input_tokens_seen": 144504610, + "step": 6697, + "time_per_iteration": 2.853018045425415 + }, + { + "auxiliary_loss_clip": 0.01061667, + "auxiliary_loss_mlp": 0.01001411, + "balance_loss_clip": 1.00993729, + "balance_loss_mlp": 1.00032032, + "epoch": 0.8053868815006313, + "flos": 67157875768320.0, + "grad_norm": 0.7157614234459486, + "language_loss": 0.59025323, + "learning_rate": 3.8416660422651127e-07, + "loss": 0.61088407, + "num_input_tokens_seen": 144574260, + "step": 6698, + "time_per_iteration": 3.3085501194000244 + }, + { + "auxiliary_loss_clip": 0.01198337, + "auxiliary_loss_mlp": 0.01025217, + "balance_loss_clip": 1.04223418, + "balance_loss_mlp": 1.01768339, + "epoch": 0.8055071243912704, + "flos": 23837000307840.0, + "grad_norm": 1.8907755949829919, + "language_loss": 0.681292, + "learning_rate": 3.837076813715723e-07, + "loss": 0.70352757, + "num_input_tokens_seen": 144594145, + "step": 6699, + "time_per_iteration": 2.7370262145996094 + }, + { + "auxiliary_loss_clip": 0.01186499, + "auxiliary_loss_mlp": 0.01026082, + "balance_loss_clip": 1.03966832, + "balance_loss_mlp": 1.01825619, + "epoch": 0.8056273672819094, + "flos": 21324510760320.0, + "grad_norm": 1.9121611437758343, + "language_loss": 0.75083005, + "learning_rate": 3.832490037068941e-07, + "loss": 0.77295589, + "num_input_tokens_seen": 144612935, + "step": 6700, + "time_per_iteration": 2.707075595855713 + }, + { + "auxiliary_loss_clip": 0.01211243, + "auxiliary_loss_mlp": 0.01022655, + "balance_loss_clip": 1.04179239, + "balance_loss_mlp": 1.01530004, + "epoch": 0.8057476101725486, + "flos": 25768383626880.0, + "grad_norm": 1.8873833472743327, + "language_loss": 0.75823951, + "learning_rate": 3.827905713020554e-07, + "loss": 0.78057849, + "num_input_tokens_seen": 144630580, + "step": 6701, + "time_per_iteration": 2.8350324630737305 + }, + { + "auxiliary_loss_clip": 0.0119332, + "auxiliary_loss_mlp": 0.01029628, + "balance_loss_clip": 1.03977263, + "balance_loss_mlp": 1.02137852, + "epoch": 0.8058678530631876, + "flos": 24535283679360.0, + "grad_norm": 1.9499236551492072, + "language_loss": 0.68982673, + "learning_rate": 3.823323842266017e-07, + "loss": 0.71205616, + "num_input_tokens_seen": 144649975, + "step": 6702, + "time_per_iteration": 2.8080222606658936 + }, + { + "auxiliary_loss_clip": 0.01176885, + "auxiliary_loss_mlp": 0.0102601, + "balance_loss_clip": 1.04377007, + "balance_loss_mlp": 1.01840985, + "epoch": 0.8059880959538267, + "flos": 24753728240640.0, + "grad_norm": 2.7979901325062513, + "language_loss": 0.72786361, + "learning_rate": 3.818744425500393e-07, + "loss": 0.74989259, + "num_input_tokens_seen": 144667990, + "step": 6703, + "time_per_iteration": 2.6565802097320557 + }, + { + "auxiliary_loss_clip": 0.01186476, + "auxiliary_loss_mlp": 0.01025525, + "balance_loss_clip": 1.04133785, + "balance_loss_mlp": 1.01816654, + "epoch": 0.8061083388444659, + "flos": 22196349671040.0, + "grad_norm": 1.7536817987698978, + "language_loss": 0.80409873, + "learning_rate": 3.8141674634183675e-07, + "loss": 0.82621872, + "num_input_tokens_seen": 144687020, + "step": 6704, + "time_per_iteration": 2.71921706199646 + }, + { + "auxiliary_loss_clip": 0.01202717, + "auxiliary_loss_mlp": 0.0102764, + "balance_loss_clip": 1.04281712, + "balance_loss_mlp": 1.02048767, + "epoch": 0.8062285817351049, + "flos": 30044195735040.0, + "grad_norm": 1.9357401136825914, + "language_loss": 0.66201049, + "learning_rate": 3.809592956714278e-07, + "loss": 0.68431413, + "num_input_tokens_seen": 144710255, + "step": 6705, + "time_per_iteration": 2.8006253242492676 + }, + { + "auxiliary_loss_clip": 0.01180213, + "auxiliary_loss_mlp": 0.01027492, + "balance_loss_clip": 1.04728723, + "balance_loss_mlp": 1.02041435, + "epoch": 0.806348824625744, + "flos": 22782591544320.0, + "grad_norm": 2.054529265073439, + "language_loss": 0.74869347, + "learning_rate": 3.805020906082057e-07, + "loss": 0.77077055, + "num_input_tokens_seen": 144728830, + "step": 6706, + "time_per_iteration": 2.644230842590332 + }, + { + "auxiliary_loss_clip": 0.01189984, + "auxiliary_loss_mlp": 0.01029535, + "balance_loss_clip": 1.04538345, + "balance_loss_mlp": 1.02174211, + "epoch": 0.8064690675163831, + "flos": 23404600385280.0, + "grad_norm": 3.1711331181146485, + "language_loss": 0.80997217, + "learning_rate": 3.8004513122152917e-07, + "loss": 0.83216739, + "num_input_tokens_seen": 144747140, + "step": 6707, + "time_per_iteration": 2.6177916526794434 + }, + { + "auxiliary_loss_clip": 0.01175881, + "auxiliary_loss_mlp": 0.01031594, + "balance_loss_clip": 1.04401731, + "balance_loss_mlp": 1.02465916, + "epoch": 0.8065893104070222, + "flos": 24060903736320.0, + "grad_norm": 1.7130475813715904, + "language_loss": 0.67098659, + "learning_rate": 3.79588417580718e-07, + "loss": 0.69306135, + "num_input_tokens_seen": 144765250, + "step": 6708, + "time_per_iteration": 2.73946475982666 + }, + { + "auxiliary_loss_clip": 0.01180601, + "auxiliary_loss_mlp": 0.0102755, + "balance_loss_clip": 1.04804504, + "balance_loss_mlp": 1.0196104, + "epoch": 0.8067095532976613, + "flos": 22305410340480.0, + "grad_norm": 1.899720841683008, + "language_loss": 0.76202643, + "learning_rate": 3.791319497550558e-07, + "loss": 0.78410792, + "num_input_tokens_seen": 144783080, + "step": 6709, + "time_per_iteration": 2.583484411239624 + }, + { + "auxiliary_loss_clip": 0.01201439, + "auxiliary_loss_mlp": 0.01378622, + "balance_loss_clip": 1.04361284, + "balance_loss_mlp": 1.00020087, + "epoch": 0.8068297961883004, + "flos": 17129498296320.0, + "grad_norm": 2.041657762191672, + "language_loss": 0.70990795, + "learning_rate": 3.78675727813788e-07, + "loss": 0.73570853, + "num_input_tokens_seen": 144800645, + "step": 6710, + "time_per_iteration": 2.7450015544891357 + }, + { + "auxiliary_loss_clip": 0.01186923, + "auxiliary_loss_mlp": 0.0102626, + "balance_loss_clip": 1.04620278, + "balance_loss_mlp": 1.01842761, + "epoch": 0.8069500390789395, + "flos": 22018843635840.0, + "grad_norm": 1.595267089465208, + "language_loss": 0.73299491, + "learning_rate": 3.782197518261225e-07, + "loss": 0.75512671, + "num_input_tokens_seen": 144820085, + "step": 6711, + "time_per_iteration": 3.6461050510406494 + }, + { + "auxiliary_loss_clip": 0.01190398, + "auxiliary_loss_mlp": 0.01026043, + "balance_loss_clip": 1.04495513, + "balance_loss_mlp": 1.01880741, + "epoch": 0.8070702819695785, + "flos": 19244241567360.0, + "grad_norm": 2.0939056813260564, + "language_loss": 0.95425653, + "learning_rate": 3.777640218612319e-07, + "loss": 0.97642088, + "num_input_tokens_seen": 144838070, + "step": 6712, + "time_per_iteration": 2.661470890045166 + }, + { + "auxiliary_loss_clip": 0.01170667, + "auxiliary_loss_mlp": 0.01021033, + "balance_loss_clip": 1.04449058, + "balance_loss_mlp": 1.01412749, + "epoch": 0.8071905248602176, + "flos": 21544320038400.0, + "grad_norm": 2.1385575453982324, + "language_loss": 0.72488815, + "learning_rate": 3.773085379882488e-07, + "loss": 0.74680519, + "num_input_tokens_seen": 144857125, + "step": 6713, + "time_per_iteration": 3.522981643676758 + }, + { + "auxiliary_loss_clip": 0.01178251, + "auxiliary_loss_mlp": 0.01378989, + "balance_loss_clip": 1.04567182, + "balance_loss_mlp": 1.00020897, + "epoch": 0.8073107677508568, + "flos": 37268309105280.0, + "grad_norm": 2.0309640355657503, + "language_loss": 0.75501132, + "learning_rate": 3.768533002762715e-07, + "loss": 0.78058374, + "num_input_tokens_seen": 144880660, + "step": 6714, + "time_per_iteration": 2.751783847808838 + }, + { + "auxiliary_loss_clip": 0.01186982, + "auxiliary_loss_mlp": 0.01026151, + "balance_loss_clip": 1.04296136, + "balance_loss_mlp": 1.01889157, + "epoch": 0.8074310106414958, + "flos": 28366269759360.0, + "grad_norm": 2.254232688551718, + "language_loss": 0.77008343, + "learning_rate": 3.763983087943572e-07, + "loss": 0.79221481, + "num_input_tokens_seen": 144900050, + "step": 6715, + "time_per_iteration": 3.643699884414673 + }, + { + "auxiliary_loss_clip": 0.01164271, + "auxiliary_loss_mlp": 0.01378455, + "balance_loss_clip": 1.04214525, + "balance_loss_mlp": 1.00017667, + "epoch": 0.8075512535321349, + "flos": 24281646768000.0, + "grad_norm": 1.640089796358983, + "language_loss": 0.81149834, + "learning_rate": 3.759435636115282e-07, + "loss": 0.83692563, + "num_input_tokens_seen": 144920835, + "step": 6716, + "time_per_iteration": 2.6827890872955322 + }, + { + "auxiliary_loss_clip": 0.01207571, + "auxiliary_loss_mlp": 0.01378657, + "balance_loss_clip": 1.04081392, + "balance_loss_mlp": 1.00019372, + "epoch": 0.807671496422774, + "flos": 26030855283840.0, + "grad_norm": 1.7649368406984485, + "language_loss": 0.73175514, + "learning_rate": 3.7548906479676967e-07, + "loss": 0.75761741, + "num_input_tokens_seen": 144940430, + "step": 6717, + "time_per_iteration": 2.774108409881592 + }, + { + "auxiliary_loss_clip": 0.01177921, + "auxiliary_loss_mlp": 0.0102283, + "balance_loss_clip": 1.04299927, + "balance_loss_mlp": 1.01530242, + "epoch": 0.8077917393134131, + "flos": 23730740899200.0, + "grad_norm": 2.6348743314670364, + "language_loss": 0.71431762, + "learning_rate": 3.7503481241902855e-07, + "loss": 0.7363252, + "num_input_tokens_seen": 144960405, + "step": 6718, + "time_per_iteration": 3.724104642868042 + }, + { + "auxiliary_loss_clip": 0.01183578, + "auxiliary_loss_mlp": 0.01378646, + "balance_loss_clip": 1.04303205, + "balance_loss_mlp": 1.00021219, + "epoch": 0.8079119822040521, + "flos": 18402028398720.0, + "grad_norm": 2.996324496282965, + "language_loss": 0.79974794, + "learning_rate": 3.745808065472145e-07, + "loss": 0.82537019, + "num_input_tokens_seen": 144977700, + "step": 6719, + "time_per_iteration": 2.6432206630706787 + }, + { + "auxiliary_loss_clip": 0.0117403, + "auxiliary_loss_mlp": 0.01024082, + "balance_loss_clip": 1.04860008, + "balance_loss_mlp": 1.01708138, + "epoch": 0.8080322250946913, + "flos": 23621787970560.0, + "grad_norm": 1.5993410775272359, + "language_loss": 0.7625103, + "learning_rate": 3.741270472501994e-07, + "loss": 0.78449142, + "num_input_tokens_seen": 144998340, + "step": 6720, + "time_per_iteration": 2.697406053543091 + }, + { + "auxiliary_loss_clip": 0.01185479, + "auxiliary_loss_mlp": 0.01027743, + "balance_loss_clip": 1.04631925, + "balance_loss_mlp": 1.02055156, + "epoch": 0.8081524679853304, + "flos": 22820692896000.0, + "grad_norm": 1.706918027198962, + "language_loss": 0.72551739, + "learning_rate": 3.736735345968183e-07, + "loss": 0.74764955, + "num_input_tokens_seen": 145017950, + "step": 6721, + "time_per_iteration": 2.777252435684204 + }, + { + "auxiliary_loss_clip": 0.01179442, + "auxiliary_loss_mlp": 0.01027763, + "balance_loss_clip": 1.04708076, + "balance_loss_mlp": 1.01980543, + "epoch": 0.8082727108759694, + "flos": 17640004343040.0, + "grad_norm": 1.6016982959758694, + "language_loss": 0.78766632, + "learning_rate": 3.7322026865586986e-07, + "loss": 0.8097384, + "num_input_tokens_seen": 145036985, + "step": 6722, + "time_per_iteration": 2.678046703338623 + }, + { + "auxiliary_loss_clip": 0.01185062, + "auxiliary_loss_mlp": 0.01028034, + "balance_loss_clip": 1.04826927, + "balance_loss_mlp": 1.02065527, + "epoch": 0.8083929537666086, + "flos": 25958172113280.0, + "grad_norm": 1.948268680248792, + "language_loss": 0.72991186, + "learning_rate": 3.7276724949611206e-07, + "loss": 0.75204283, + "num_input_tokens_seen": 145057095, + "step": 6723, + "time_per_iteration": 2.693758487701416 + }, + { + "auxiliary_loss_clip": 0.01191229, + "auxiliary_loss_mlp": 0.01032083, + "balance_loss_clip": 1.04664898, + "balance_loss_mlp": 1.02455544, + "epoch": 0.8085131966572476, + "flos": 27089178629760.0, + "grad_norm": 1.7635380928611826, + "language_loss": 0.75188541, + "learning_rate": 3.723144771862694e-07, + "loss": 0.77411854, + "num_input_tokens_seen": 145077735, + "step": 6724, + "time_per_iteration": 2.6665568351745605 + }, + { + "auxiliary_loss_clip": 0.01198483, + "auxiliary_loss_mlp": 0.01030012, + "balance_loss_clip": 1.04265499, + "balance_loss_mlp": 1.02257288, + "epoch": 0.8086334395478867, + "flos": 23988543788160.0, + "grad_norm": 1.6443262824305431, + "language_loss": 0.76816225, + "learning_rate": 3.718619517950263e-07, + "loss": 0.79044712, + "num_input_tokens_seen": 145098330, + "step": 6725, + "time_per_iteration": 2.742957830429077 + }, + { + "auxiliary_loss_clip": 0.01171164, + "auxiliary_loss_mlp": 0.01027171, + "balance_loss_clip": 1.05038786, + "balance_loss_mlp": 1.02027178, + "epoch": 0.8087536824385259, + "flos": 20405879406720.0, + "grad_norm": 1.9758975946894146, + "language_loss": 0.77036512, + "learning_rate": 3.714096733910301e-07, + "loss": 0.79234844, + "num_input_tokens_seen": 145115855, + "step": 6726, + "time_per_iteration": 2.5627596378326416 + }, + { + "auxiliary_loss_clip": 0.01183638, + "auxiliary_loss_mlp": 0.01028692, + "balance_loss_clip": 1.04633594, + "balance_loss_mlp": 1.02079725, + "epoch": 0.8088739253291649, + "flos": 25919639798400.0, + "grad_norm": 1.917312894484014, + "language_loss": 0.70450163, + "learning_rate": 3.709576420428926e-07, + "loss": 0.72662497, + "num_input_tokens_seen": 145136655, + "step": 6727, + "time_per_iteration": 2.6903958320617676 + }, + { + "auxiliary_loss_clip": 0.01187588, + "auxiliary_loss_mlp": 0.01024078, + "balance_loss_clip": 1.0433805, + "balance_loss_mlp": 1.01664519, + "epoch": 0.808994168219804, + "flos": 28402072640640.0, + "grad_norm": 2.9409774859445195, + "language_loss": 0.73571235, + "learning_rate": 3.7050585781918463e-07, + "loss": 0.75782901, + "num_input_tokens_seen": 145156955, + "step": 6728, + "time_per_iteration": 2.686544418334961 + }, + { + "auxiliary_loss_clip": 0.01180723, + "auxiliary_loss_mlp": 0.01021963, + "balance_loss_clip": 1.04528594, + "balance_loss_mlp": 1.01386833, + "epoch": 0.8091144111104431, + "flos": 17421056991360.0, + "grad_norm": 2.151207910580529, + "language_loss": 0.69111043, + "learning_rate": 3.700543207884428e-07, + "loss": 0.71313727, + "num_input_tokens_seen": 145173865, + "step": 6729, + "time_per_iteration": 2.657184362411499 + }, + { + "auxiliary_loss_clip": 0.01174838, + "auxiliary_loss_mlp": 0.01023033, + "balance_loss_clip": 1.04599214, + "balance_loss_mlp": 1.01619315, + "epoch": 0.8092346540010822, + "flos": 32153803361280.0, + "grad_norm": 2.146551726247815, + "language_loss": 0.70997608, + "learning_rate": 3.6960303101916466e-07, + "loss": 0.73195469, + "num_input_tokens_seen": 145193780, + "step": 6730, + "time_per_iteration": 2.725419521331787 + }, + { + "auxiliary_loss_clip": 0.01061311, + "auxiliary_loss_mlp": 0.01373674, + "balance_loss_clip": 1.00950289, + "balance_loss_mlp": 0.99966735, + "epoch": 0.8093548968917212, + "flos": 58035093390720.0, + "grad_norm": 0.7437503926691309, + "language_loss": 0.55588228, + "learning_rate": 3.6915198857981047e-07, + "loss": 0.58023208, + "num_input_tokens_seen": 145258980, + "step": 6731, + "time_per_iteration": 3.2206203937530518 + }, + { + "auxiliary_loss_clip": 0.0119102, + "auxiliary_loss_mlp": 0.01026751, + "balance_loss_clip": 1.04252493, + "balance_loss_mlp": 1.01876354, + "epoch": 0.8094751397823604, + "flos": 27381599251200.0, + "grad_norm": 1.6587135150570513, + "language_loss": 0.67772949, + "learning_rate": 3.687011935388027e-07, + "loss": 0.69990724, + "num_input_tokens_seen": 145281875, + "step": 6732, + "time_per_iteration": 2.7843375205993652 + }, + { + "auxiliary_loss_clip": 0.01177637, + "auxiliary_loss_mlp": 0.01023523, + "balance_loss_clip": 1.04656339, + "balance_loss_mlp": 1.01606631, + "epoch": 0.8095953826729995, + "flos": 24061083304320.0, + "grad_norm": 1.8218816256301043, + "language_loss": 0.73020113, + "learning_rate": 3.6825064596452646e-07, + "loss": 0.75221276, + "num_input_tokens_seen": 145302220, + "step": 6733, + "time_per_iteration": 2.665332317352295 + }, + { + "auxiliary_loss_clip": 0.01176559, + "auxiliary_loss_mlp": 0.01025493, + "balance_loss_clip": 1.04528737, + "balance_loss_mlp": 1.01861453, + "epoch": 0.8097156255636385, + "flos": 23951412103680.0, + "grad_norm": 1.837899416443552, + "language_loss": 0.70888567, + "learning_rate": 3.678003459253305e-07, + "loss": 0.73090613, + "num_input_tokens_seen": 145323070, + "step": 6734, + "time_per_iteration": 2.7115371227264404 + }, + { + "auxiliary_loss_clip": 0.01194419, + "auxiliary_loss_mlp": 0.01027312, + "balance_loss_clip": 1.044155, + "balance_loss_mlp": 1.01934254, + "epoch": 0.8098358684542777, + "flos": 21799142098560.0, + "grad_norm": 2.0766246157908075, + "language_loss": 0.74246526, + "learning_rate": 3.673502934895236e-07, + "loss": 0.76468253, + "num_input_tokens_seen": 145342575, + "step": 6735, + "time_per_iteration": 2.684704303741455 + }, + { + "auxiliary_loss_clip": 0.01060798, + "auxiliary_loss_mlp": 0.01001467, + "balance_loss_clip": 1.00919211, + "balance_loss_mlp": 1.0003581, + "epoch": 0.8099561113449167, + "flos": 68809515966720.0, + "grad_norm": 0.6834630502142522, + "language_loss": 0.57962614, + "learning_rate": 3.669004887253802e-07, + "loss": 0.60024875, + "num_input_tokens_seen": 145408865, + "step": 6736, + "time_per_iteration": 3.3269622325897217 + }, + { + "auxiliary_loss_clip": 0.01190867, + "auxiliary_loss_mlp": 0.01023754, + "balance_loss_clip": 1.04634309, + "balance_loss_mlp": 1.01693487, + "epoch": 0.8100763542355558, + "flos": 23586056916480.0, + "grad_norm": 1.6188472999278725, + "language_loss": 0.79029542, + "learning_rate": 3.664509317011335e-07, + "loss": 0.81244165, + "num_input_tokens_seen": 145429200, + "step": 6737, + "time_per_iteration": 3.5980141162872314 + }, + { + "auxiliary_loss_clip": 0.01179083, + "auxiliary_loss_mlp": 0.01030383, + "balance_loss_clip": 1.05000973, + "balance_loss_mlp": 1.02165699, + "epoch": 0.810196597126195, + "flos": 31650408207360.0, + "grad_norm": 2.006162711744843, + "language_loss": 0.73870927, + "learning_rate": 3.6600162248498134e-07, + "loss": 0.76080394, + "num_input_tokens_seen": 145452830, + "step": 6738, + "time_per_iteration": 2.6932904720306396 + }, + { + "auxiliary_loss_clip": 0.01194562, + "auxiliary_loss_mlp": 0.01022613, + "balance_loss_clip": 1.03839338, + "balance_loss_mlp": 1.01555932, + "epoch": 0.810316840016834, + "flos": 24900459298560.0, + "grad_norm": 1.7387583624898002, + "language_loss": 0.76116508, + "learning_rate": 3.6555256114508426e-07, + "loss": 0.78333682, + "num_input_tokens_seen": 145472625, + "step": 6739, + "time_per_iteration": 3.701145648956299 + }, + { + "auxiliary_loss_clip": 0.01184765, + "auxiliary_loss_mlp": 0.01026991, + "balance_loss_clip": 1.04218721, + "balance_loss_mlp": 1.01982641, + "epoch": 0.8104370829074731, + "flos": 27965003950080.0, + "grad_norm": 1.8243446737625204, + "language_loss": 0.72603166, + "learning_rate": 3.651037477495642e-07, + "loss": 0.74814922, + "num_input_tokens_seen": 145494075, + "step": 6740, + "time_per_iteration": 2.841630697250366 + }, + { + "auxiliary_loss_clip": 0.01168906, + "auxiliary_loss_mlp": 0.01029145, + "balance_loss_clip": 1.04781628, + "balance_loss_mlp": 1.02136099, + "epoch": 0.8105573257981122, + "flos": 24640752988800.0, + "grad_norm": 2.0296447649828866, + "language_loss": 0.68397653, + "learning_rate": 3.6465518236650584e-07, + "loss": 0.705957, + "num_input_tokens_seen": 145514220, + "step": 6741, + "time_per_iteration": 2.6423258781433105 + }, + { + "auxiliary_loss_clip": 0.01191309, + "auxiliary_loss_mlp": 0.01022808, + "balance_loss_clip": 1.04144323, + "balance_loss_mlp": 1.01562262, + "epoch": 0.8106775686887513, + "flos": 26358935132160.0, + "grad_norm": 1.9808425587849716, + "language_loss": 0.78405643, + "learning_rate": 3.642068650639558e-07, + "loss": 0.80619758, + "num_input_tokens_seen": 145533965, + "step": 6742, + "time_per_iteration": 3.6227593421936035 + }, + { + "auxiliary_loss_clip": 0.01176827, + "auxiliary_loss_mlp": 0.01023458, + "balance_loss_clip": 1.03979278, + "balance_loss_mlp": 1.01620078, + "epoch": 0.8107978115793903, + "flos": 27271892136960.0, + "grad_norm": 1.7187116744086237, + "language_loss": 0.64352381, + "learning_rate": 3.6375879590992334e-07, + "loss": 0.66552663, + "num_input_tokens_seen": 145554310, + "step": 6743, + "time_per_iteration": 3.663438081741333 + }, + { + "auxiliary_loss_clip": 0.01181386, + "auxiliary_loss_mlp": 0.0102204, + "balance_loss_clip": 1.04396296, + "balance_loss_mlp": 1.01418984, + "epoch": 0.8109180544700295, + "flos": 24934322845440.0, + "grad_norm": 1.6958962937694855, + "language_loss": 0.80884522, + "learning_rate": 3.6331097497238173e-07, + "loss": 0.83087945, + "num_input_tokens_seen": 145573755, + "step": 6744, + "time_per_iteration": 2.671671152114868 + }, + { + "auxiliary_loss_clip": 0.01192705, + "auxiliary_loss_mlp": 0.01023621, + "balance_loss_clip": 1.04221106, + "balance_loss_mlp": 1.01658154, + "epoch": 0.8110382973606686, + "flos": 21105383840640.0, + "grad_norm": 2.751432755712013, + "language_loss": 0.79963696, + "learning_rate": 3.628634023192627e-07, + "loss": 0.82180023, + "num_input_tokens_seen": 145594000, + "step": 6745, + "time_per_iteration": 2.752230167388916 + }, + { + "auxiliary_loss_clip": 0.01177908, + "auxiliary_loss_mlp": 0.01027437, + "balance_loss_clip": 1.04414058, + "balance_loss_mlp": 1.01992726, + "epoch": 0.8111585402513076, + "flos": 15414081500160.0, + "grad_norm": 2.1381144492296555, + "language_loss": 0.75248134, + "learning_rate": 3.624160780184644e-07, + "loss": 0.77453476, + "num_input_tokens_seen": 145611215, + "step": 6746, + "time_per_iteration": 2.6439640522003174 + }, + { + "auxiliary_loss_clip": 0.01178563, + "auxiliary_loss_mlp": 0.01022536, + "balance_loss_clip": 1.04160786, + "balance_loss_mlp": 1.01487124, + "epoch": 0.8112787831419467, + "flos": 24095736950400.0, + "grad_norm": 1.8107025202083291, + "language_loss": 0.74539173, + "learning_rate": 3.6196900213784496e-07, + "loss": 0.76740265, + "num_input_tokens_seen": 145630530, + "step": 6747, + "time_per_iteration": 2.6895008087158203 + }, + { + "auxiliary_loss_clip": 0.01178338, + "auxiliary_loss_mlp": 0.01027622, + "balance_loss_clip": 1.04611659, + "balance_loss_mlp": 1.02009976, + "epoch": 0.8113990260325858, + "flos": 20483374999680.0, + "grad_norm": 1.8394142351798117, + "language_loss": 0.86593878, + "learning_rate": 3.6152217474522527e-07, + "loss": 0.88799834, + "num_input_tokens_seen": 145647345, + "step": 6748, + "time_per_iteration": 2.61057186126709 + }, + { + "auxiliary_loss_clip": 0.01179667, + "auxiliary_loss_mlp": 0.01027055, + "balance_loss_clip": 1.0497576, + "balance_loss_mlp": 1.01992083, + "epoch": 0.8115192689232249, + "flos": 24901141656960.0, + "grad_norm": 1.6472571646446563, + "language_loss": 0.72848713, + "learning_rate": 3.6107559590838975e-07, + "loss": 0.75055438, + "num_input_tokens_seen": 145666330, + "step": 6749, + "time_per_iteration": 2.667677640914917 + }, + { + "auxiliary_loss_clip": 0.01203712, + "auxiliary_loss_mlp": 0.01024444, + "balance_loss_clip": 1.0389564, + "balance_loss_mlp": 1.01686883, + "epoch": 0.811639511813864, + "flos": 24057204635520.0, + "grad_norm": 3.096697262801433, + "language_loss": 0.66553712, + "learning_rate": 3.606292656950822e-07, + "loss": 0.68781871, + "num_input_tokens_seen": 145684740, + "step": 6750, + "time_per_iteration": 2.7862775325775146 + }, + { + "auxiliary_loss_clip": 0.01180985, + "auxiliary_loss_mlp": 0.01024579, + "balance_loss_clip": 1.04245448, + "balance_loss_mlp": 1.01693761, + "epoch": 0.8117597547045031, + "flos": 23185150243200.0, + "grad_norm": 6.204242285313564, + "language_loss": 0.86452842, + "learning_rate": 3.601831841730121e-07, + "loss": 0.88658404, + "num_input_tokens_seen": 145702660, + "step": 6751, + "time_per_iteration": 2.725001811981201 + }, + { + "auxiliary_loss_clip": 0.01175271, + "auxiliary_loss_mlp": 0.01025012, + "balance_loss_clip": 1.04600167, + "balance_loss_mlp": 1.01778793, + "epoch": 0.8118799975951422, + "flos": 23040250778880.0, + "grad_norm": 1.7619238649156088, + "language_loss": 0.72619134, + "learning_rate": 3.5973735140984916e-07, + "loss": 0.74819416, + "num_input_tokens_seen": 145722830, + "step": 6752, + "time_per_iteration": 2.608283758163452 + }, + { + "auxiliary_loss_clip": 0.0120026, + "auxiliary_loss_mlp": 0.01379, + "balance_loss_clip": 1.04021418, + "balance_loss_mlp": 1.00014114, + "epoch": 0.8120002404857812, + "flos": 24639962889600.0, + "grad_norm": 2.271021410568505, + "language_loss": 0.79047036, + "learning_rate": 3.5929176747322607e-07, + "loss": 0.81626296, + "num_input_tokens_seen": 145741935, + "step": 6753, + "time_per_iteration": 2.7652857303619385 + }, + { + "auxiliary_loss_clip": 0.01089233, + "auxiliary_loss_mlp": 0.01001354, + "balance_loss_clip": 1.01007283, + "balance_loss_mlp": 1.00026309, + "epoch": 0.8121204833764204, + "flos": 57415742156160.0, + "grad_norm": 0.8079103512860845, + "language_loss": 0.56255603, + "learning_rate": 3.588464324307372e-07, + "loss": 0.58346188, + "num_input_tokens_seen": 145805560, + "step": 6754, + "time_per_iteration": 3.34719181060791 + }, + { + "auxiliary_loss_clip": 0.0117886, + "auxiliary_loss_mlp": 0.01028646, + "balance_loss_clip": 1.04540324, + "balance_loss_mlp": 1.02121639, + "epoch": 0.8122407262670595, + "flos": 19464589549440.0, + "grad_norm": 1.81246091503928, + "language_loss": 0.75224912, + "learning_rate": 3.584013463499391e-07, + "loss": 0.77432418, + "num_input_tokens_seen": 145824180, + "step": 6755, + "time_per_iteration": 2.716376543045044 + }, + { + "auxiliary_loss_clip": 0.01088404, + "auxiliary_loss_mlp": 0.01001049, + "balance_loss_clip": 1.01070666, + "balance_loss_mlp": 0.99995196, + "epoch": 0.8123609691576985, + "flos": 56425325472000.0, + "grad_norm": 0.7303219970506862, + "language_loss": 0.64430243, + "learning_rate": 3.579565092983521e-07, + "loss": 0.66519696, + "num_input_tokens_seen": 145885300, + "step": 6756, + "time_per_iteration": 3.143256664276123 + }, + { + "auxiliary_loss_clip": 0.01168096, + "auxiliary_loss_mlp": 0.01027691, + "balance_loss_clip": 1.04754972, + "balance_loss_mlp": 1.02035332, + "epoch": 0.8124812120483377, + "flos": 20631973564800.0, + "grad_norm": 2.0199240248579358, + "language_loss": 0.83485627, + "learning_rate": 3.575119213434565e-07, + "loss": 0.85681415, + "num_input_tokens_seen": 145903815, + "step": 6757, + "time_per_iteration": 2.5444040298461914 + }, + { + "auxiliary_loss_clip": 0.0117684, + "auxiliary_loss_mlp": 0.01027642, + "balance_loss_clip": 1.04679644, + "balance_loss_mlp": 1.02050757, + "epoch": 0.8126014549389767, + "flos": 22492397566080.0, + "grad_norm": 1.8435174507660308, + "language_loss": 0.81692988, + "learning_rate": 3.5706758255269765e-07, + "loss": 0.83897471, + "num_input_tokens_seen": 145922270, + "step": 6758, + "time_per_iteration": 2.8257861137390137 + }, + { + "auxiliary_loss_clip": 0.01191578, + "auxiliary_loss_mlp": 0.01023922, + "balance_loss_clip": 1.04604042, + "balance_loss_mlp": 1.01624513, + "epoch": 0.8127216978296158, + "flos": 23287961946240.0, + "grad_norm": 2.0932574179505465, + "language_loss": 0.69706631, + "learning_rate": 3.566234929934795e-07, + "loss": 0.71922135, + "num_input_tokens_seen": 145941470, + "step": 6759, + "time_per_iteration": 2.704735517501831 + }, + { + "auxiliary_loss_clip": 0.01176764, + "auxiliary_loss_mlp": 0.01024437, + "balance_loss_clip": 1.04883528, + "balance_loss_mlp": 1.01714778, + "epoch": 0.812841940720255, + "flos": 25154994049920.0, + "grad_norm": 1.5255887593905724, + "language_loss": 0.71850061, + "learning_rate": 3.561796527331706e-07, + "loss": 0.74051261, + "num_input_tokens_seen": 145963145, + "step": 6760, + "time_per_iteration": 2.6909942626953125 + }, + { + "auxiliary_loss_clip": 0.01196183, + "auxiliary_loss_mlp": 0.01023947, + "balance_loss_clip": 1.04314208, + "balance_loss_mlp": 1.01618624, + "epoch": 0.812962183610894, + "flos": 26648446752000.0, + "grad_norm": 2.9016142271158065, + "language_loss": 0.77685291, + "learning_rate": 3.5573606183910163e-07, + "loss": 0.79905427, + "num_input_tokens_seen": 145983150, + "step": 6761, + "time_per_iteration": 2.727367877960205 + }, + { + "auxiliary_loss_clip": 0.01181791, + "auxiliary_loss_mlp": 0.01027051, + "balance_loss_clip": 1.04434264, + "balance_loss_mlp": 1.01920366, + "epoch": 0.8130824265015331, + "flos": 24966965329920.0, + "grad_norm": 1.77631300415274, + "language_loss": 0.78715312, + "learning_rate": 3.5529272037856493e-07, + "loss": 0.80924153, + "num_input_tokens_seen": 146001365, + "step": 6762, + "time_per_iteration": 2.6960389614105225 + }, + { + "auxiliary_loss_clip": 0.01127621, + "auxiliary_loss_mlp": 0.01000755, + "balance_loss_clip": 1.00930369, + "balance_loss_mlp": 0.99957466, + "epoch": 0.8132026693921722, + "flos": 67622918175360.0, + "grad_norm": 0.7469081443854303, + "language_loss": 0.53842545, + "learning_rate": 3.548496284188149e-07, + "loss": 0.55970919, + "num_input_tokens_seen": 146061570, + "step": 6763, + "time_per_iteration": 4.339489459991455 + }, + { + "auxiliary_loss_clip": 0.01195901, + "auxiliary_loss_mlp": 0.01027539, + "balance_loss_clip": 1.04256058, + "balance_loss_mlp": 1.02023101, + "epoch": 0.8133229122828113, + "flos": 19495149045120.0, + "grad_norm": 1.7913038234015388, + "language_loss": 0.79336548, + "learning_rate": 3.544067860270681e-07, + "loss": 0.81559992, + "num_input_tokens_seen": 146079145, + "step": 6764, + "time_per_iteration": 3.007920742034912 + }, + { + "auxiliary_loss_clip": 0.01198551, + "auxiliary_loss_mlp": 0.01026424, + "balance_loss_clip": 1.04355896, + "balance_loss_mlp": 1.01828778, + "epoch": 0.8134431551734503, + "flos": 20668135582080.0, + "grad_norm": 1.7014625040205797, + "language_loss": 0.71105421, + "learning_rate": 3.539641932705029e-07, + "loss": 0.7333039, + "num_input_tokens_seen": 146097625, + "step": 6765, + "time_per_iteration": 3.764756917953491 + }, + { + "auxiliary_loss_clip": 0.01171721, + "auxiliary_loss_mlp": 0.01024193, + "balance_loss_clip": 1.04746985, + "balance_loss_mlp": 1.01655746, + "epoch": 0.8135633980640895, + "flos": 21507332008320.0, + "grad_norm": 2.4512489610771935, + "language_loss": 0.77417862, + "learning_rate": 3.53521850216262e-07, + "loss": 0.79613769, + "num_input_tokens_seen": 146117195, + "step": 6766, + "time_per_iteration": 2.6159098148345947 + }, + { + "auxiliary_loss_clip": 0.01170602, + "auxiliary_loss_mlp": 0.01023829, + "balance_loss_clip": 1.04821849, + "balance_loss_mlp": 1.01704633, + "epoch": 0.8136836409547286, + "flos": 20554442058240.0, + "grad_norm": 3.518878611624651, + "language_loss": 0.77058065, + "learning_rate": 3.530797569314461e-07, + "loss": 0.79252499, + "num_input_tokens_seen": 146136220, + "step": 6767, + "time_per_iteration": 2.552089214324951 + }, + { + "auxiliary_loss_clip": 0.01167995, + "auxiliary_loss_mlp": 0.01025011, + "balance_loss_clip": 1.04715514, + "balance_loss_mlp": 1.0169462, + "epoch": 0.8138038838453676, + "flos": 20299045380480.0, + "grad_norm": 1.815878760668647, + "language_loss": 0.77562499, + "learning_rate": 3.5263791348312235e-07, + "loss": 0.79755497, + "num_input_tokens_seen": 146155415, + "step": 6768, + "time_per_iteration": 3.4656145572662354 + }, + { + "auxiliary_loss_clip": 0.01185153, + "auxiliary_loss_mlp": 0.01022631, + "balance_loss_clip": 1.04376292, + "balance_loss_mlp": 1.015553, + "epoch": 0.8139241267360068, + "flos": 29789840551680.0, + "grad_norm": 2.0607606499952436, + "language_loss": 0.70859146, + "learning_rate": 3.521963199383171e-07, + "loss": 0.73066932, + "num_input_tokens_seen": 146178370, + "step": 6769, + "time_per_iteration": 2.6893527507781982 + }, + { + "auxiliary_loss_clip": 0.012038, + "auxiliary_loss_mlp": 0.01025856, + "balance_loss_clip": 1.04185867, + "balance_loss_mlp": 1.01802969, + "epoch": 0.8140443696266458, + "flos": 19713270384000.0, + "grad_norm": 2.1556421904190506, + "language_loss": 0.7676686, + "learning_rate": 3.517549763640197e-07, + "loss": 0.78996515, + "num_input_tokens_seen": 146196010, + "step": 6770, + "time_per_iteration": 3.6428956985473633 + }, + { + "auxiliary_loss_clip": 0.01175526, + "auxiliary_loss_mlp": 0.01378319, + "balance_loss_clip": 1.04820275, + "balance_loss_mlp": 1.00022626, + "epoch": 0.8141646125172849, + "flos": 27160568910720.0, + "grad_norm": 2.0095077667702976, + "language_loss": 0.71201205, + "learning_rate": 3.513138828271829e-07, + "loss": 0.7375505, + "num_input_tokens_seen": 146215880, + "step": 6771, + "time_per_iteration": 2.6628313064575195 + }, + { + "auxiliary_loss_clip": 0.01190351, + "auxiliary_loss_mlp": 0.0102619, + "balance_loss_clip": 1.04312921, + "balance_loss_mlp": 1.01895106, + "epoch": 0.8142848554079241, + "flos": 39673102700160.0, + "grad_norm": 1.9824687527967813, + "language_loss": 0.70343554, + "learning_rate": 3.508730393947179e-07, + "loss": 0.72560096, + "num_input_tokens_seen": 146239135, + "step": 6772, + "time_per_iteration": 2.919008731842041 + }, + { + "auxiliary_loss_clip": 0.01193671, + "auxiliary_loss_mlp": 0.01029166, + "balance_loss_clip": 1.0449872, + "balance_loss_mlp": 1.02214193, + "epoch": 0.8144050982985631, + "flos": 22237288197120.0, + "grad_norm": 1.6289832778919884, + "language_loss": 0.71860033, + "learning_rate": 3.504324461335024e-07, + "loss": 0.74082869, + "num_input_tokens_seen": 146259245, + "step": 6773, + "time_per_iteration": 2.807884931564331 + }, + { + "auxiliary_loss_clip": 0.01190919, + "auxiliary_loss_mlp": 0.01033304, + "balance_loss_clip": 1.03938544, + "balance_loss_mlp": 1.02535832, + "epoch": 0.8145253411892022, + "flos": 23038239617280.0, + "grad_norm": 1.591941684886825, + "language_loss": 0.88246727, + "learning_rate": 3.499921031103732e-07, + "loss": 0.90470946, + "num_input_tokens_seen": 146280015, + "step": 6774, + "time_per_iteration": 2.740774393081665 + }, + { + "auxiliary_loss_clip": 0.01202764, + "auxiliary_loss_mlp": 0.01025166, + "balance_loss_clip": 1.04152203, + "balance_loss_mlp": 1.01810849, + "epoch": 0.8146455840798413, + "flos": 24827668387200.0, + "grad_norm": 1.848634634811731, + "language_loss": 0.78263408, + "learning_rate": 3.4955201039212987e-07, + "loss": 0.80491334, + "num_input_tokens_seen": 146300935, + "step": 6775, + "time_per_iteration": 2.7437400817871094 + }, + { + "auxiliary_loss_clip": 0.01183392, + "auxiliary_loss_mlp": 0.01025959, + "balance_loss_clip": 1.04811335, + "balance_loss_mlp": 1.01849008, + "epoch": 0.8147658269704804, + "flos": 19974520978560.0, + "grad_norm": 1.998627274710677, + "language_loss": 0.65875793, + "learning_rate": 3.4911216804553465e-07, + "loss": 0.68085146, + "num_input_tokens_seen": 146319835, + "step": 6776, + "time_per_iteration": 2.576672315597534 + }, + { + "auxiliary_loss_clip": 0.01184343, + "auxiliary_loss_mlp": 0.01025389, + "balance_loss_clip": 1.04332328, + "balance_loss_mlp": 1.01721156, + "epoch": 0.8148860698611194, + "flos": 21178031097600.0, + "grad_norm": 1.841404435448948, + "language_loss": 0.699368, + "learning_rate": 3.4867257613731017e-07, + "loss": 0.72146535, + "num_input_tokens_seen": 146339030, + "step": 6777, + "time_per_iteration": 2.663876533508301 + }, + { + "auxiliary_loss_clip": 0.01187098, + "auxiliary_loss_mlp": 0.01024246, + "balance_loss_clip": 1.04387975, + "balance_loss_mlp": 1.0168848, + "epoch": 0.8150063127517585, + "flos": 19606903234560.0, + "grad_norm": 1.9941495919010455, + "language_loss": 0.85392225, + "learning_rate": 3.4823323473414343e-07, + "loss": 0.87603569, + "num_input_tokens_seen": 146358550, + "step": 6778, + "time_per_iteration": 2.673208236694336 + }, + { + "auxiliary_loss_clip": 0.01198606, + "auxiliary_loss_mlp": 0.0102543, + "balance_loss_clip": 1.04220104, + "balance_loss_mlp": 1.01767504, + "epoch": 0.8151265556423977, + "flos": 22638374438400.0, + "grad_norm": 1.907169601332972, + "language_loss": 0.75990903, + "learning_rate": 3.477941439026812e-07, + "loss": 0.78214943, + "num_input_tokens_seen": 146376770, + "step": 6779, + "time_per_iteration": 2.6518542766571045 + }, + { + "auxiliary_loss_clip": 0.01188179, + "auxiliary_loss_mlp": 0.01024371, + "balance_loss_clip": 1.0462079, + "balance_loss_mlp": 1.01728964, + "epoch": 0.8152467985330367, + "flos": 17968048277760.0, + "grad_norm": 1.7739973685145949, + "language_loss": 0.7307018, + "learning_rate": 3.473553037095349e-07, + "loss": 0.75282729, + "num_input_tokens_seen": 146395795, + "step": 6780, + "time_per_iteration": 2.6152660846710205 + }, + { + "auxiliary_loss_clip": 0.0117661, + "auxiliary_loss_mlp": 0.0102568, + "balance_loss_clip": 1.04231679, + "balance_loss_mlp": 1.01897144, + "epoch": 0.8153670414236758, + "flos": 24969012405120.0, + "grad_norm": 1.8580582276935202, + "language_loss": 0.83349615, + "learning_rate": 3.469167142212743e-07, + "loss": 0.85551906, + "num_input_tokens_seen": 146417640, + "step": 6781, + "time_per_iteration": 2.6838977336883545 + }, + { + "auxiliary_loss_clip": 0.01179914, + "auxiliary_loss_mlp": 0.01028764, + "balance_loss_clip": 1.04737258, + "balance_loss_mlp": 1.02083683, + "epoch": 0.8154872843143149, + "flos": 31066069754880.0, + "grad_norm": 2.7888005074937467, + "language_loss": 0.63328236, + "learning_rate": 3.4647837550443337e-07, + "loss": 0.65536916, + "num_input_tokens_seen": 146436205, + "step": 6782, + "time_per_iteration": 2.6742122173309326 + }, + { + "auxiliary_loss_clip": 0.01198772, + "auxiliary_loss_mlp": 0.01022158, + "balance_loss_clip": 1.04397845, + "balance_loss_mlp": 1.01483893, + "epoch": 0.815607527204954, + "flos": 19391654983680.0, + "grad_norm": 2.0432860583690196, + "language_loss": 0.74703395, + "learning_rate": 3.460402876255086e-07, + "loss": 0.7692433, + "num_input_tokens_seen": 146453595, + "step": 6783, + "time_per_iteration": 2.621577024459839 + }, + { + "auxiliary_loss_clip": 0.01179755, + "auxiliary_loss_mlp": 0.01029362, + "balance_loss_clip": 1.04558039, + "balance_loss_mlp": 1.02195573, + "epoch": 0.815727770095593, + "flos": 26140418743680.0, + "grad_norm": 2.1423414047358778, + "language_loss": 0.71495306, + "learning_rate": 3.456024506509574e-07, + "loss": 0.73704422, + "num_input_tokens_seen": 146474515, + "step": 6784, + "time_per_iteration": 2.738609790802002 + }, + { + "auxiliary_loss_clip": 0.01180917, + "auxiliary_loss_mlp": 0.01378806, + "balance_loss_clip": 1.04946351, + "balance_loss_mlp": 1.00015593, + "epoch": 0.8158480129862322, + "flos": 25337527989120.0, + "grad_norm": 1.6307192198881377, + "language_loss": 0.74061549, + "learning_rate": 3.4516486464719873e-07, + "loss": 0.7662127, + "num_input_tokens_seen": 146493905, + "step": 6785, + "time_per_iteration": 2.6153531074523926 + }, + { + "auxiliary_loss_clip": 0.01198809, + "auxiliary_loss_mlp": 0.01023558, + "balance_loss_clip": 1.04109454, + "balance_loss_mlp": 1.01593411, + "epoch": 0.8159682558768713, + "flos": 34423645559040.0, + "grad_norm": 1.6105150182693326, + "language_loss": 0.62279356, + "learning_rate": 3.4472752968061445e-07, + "loss": 0.64501721, + "num_input_tokens_seen": 146518335, + "step": 6786, + "time_per_iteration": 2.887395143508911 + }, + { + "auxiliary_loss_clip": 0.01179442, + "auxiliary_loss_mlp": 0.01020959, + "balance_loss_clip": 1.04577911, + "balance_loss_mlp": 1.01387787, + "epoch": 0.8160884987675103, + "flos": 18653223185280.0, + "grad_norm": 1.9244716295352724, + "language_loss": 0.73884487, + "learning_rate": 3.442904458175475e-07, + "loss": 0.76084888, + "num_input_tokens_seen": 146535655, + "step": 6787, + "time_per_iteration": 2.5896332263946533 + }, + { + "auxiliary_loss_clip": 0.01174106, + "auxiliary_loss_mlp": 0.01026991, + "balance_loss_clip": 1.04554486, + "balance_loss_mlp": 1.01906335, + "epoch": 0.8162087416581495, + "flos": 31430527102080.0, + "grad_norm": 1.7002594761290957, + "language_loss": 0.76393539, + "learning_rate": 3.438536131243044e-07, + "loss": 0.78594637, + "num_input_tokens_seen": 146556815, + "step": 6788, + "time_per_iteration": 2.7216198444366455 + }, + { + "auxiliary_loss_clip": 0.01191129, + "auxiliary_loss_mlp": 0.01026601, + "balance_loss_clip": 1.04624343, + "balance_loss_mlp": 1.01887584, + "epoch": 0.8163289845487885, + "flos": 37593910915200.0, + "grad_norm": 2.1447591603502913, + "language_loss": 0.61855394, + "learning_rate": 3.434170316671503e-07, + "loss": 0.64073122, + "num_input_tokens_seen": 146581845, + "step": 6789, + "time_per_iteration": 3.749073028564453 + }, + { + "auxiliary_loss_clip": 0.01194857, + "auxiliary_loss_mlp": 0.01030026, + "balance_loss_clip": 1.04951751, + "balance_loss_mlp": 1.0228858, + "epoch": 0.8164492274394276, + "flos": 13953989554560.0, + "grad_norm": 2.4423669321312365, + "language_loss": 0.89387131, + "learning_rate": 3.4298070151231583e-07, + "loss": 0.91612011, + "num_input_tokens_seen": 146597245, + "step": 6790, + "time_per_iteration": 2.6269102096557617 + }, + { + "auxiliary_loss_clip": 0.0119152, + "auxiliary_loss_mlp": 0.01026514, + "balance_loss_clip": 1.04507911, + "balance_loss_mlp": 1.01927829, + "epoch": 0.8165694703300668, + "flos": 28986554747520.0, + "grad_norm": 2.3666922319743926, + "language_loss": 0.60379249, + "learning_rate": 3.425446227259916e-07, + "loss": 0.62597281, + "num_input_tokens_seen": 146618210, + "step": 6791, + "time_per_iteration": 3.6640524864196777 + }, + { + "auxiliary_loss_clip": 0.0118787, + "auxiliary_loss_mlp": 0.01026672, + "balance_loss_clip": 1.04540396, + "balance_loss_mlp": 1.01909661, + "epoch": 0.8166897132207058, + "flos": 25118365155840.0, + "grad_norm": 2.6137601386152802, + "language_loss": 0.82624936, + "learning_rate": 3.421087953743296e-07, + "loss": 0.84839475, + "num_input_tokens_seen": 146637975, + "step": 6792, + "time_per_iteration": 2.652768850326538 + }, + { + "auxiliary_loss_clip": 0.01177151, + "auxiliary_loss_mlp": 0.01021532, + "balance_loss_clip": 1.04482293, + "balance_loss_mlp": 1.01395595, + "epoch": 0.8168099561113449, + "flos": 23148593176320.0, + "grad_norm": 2.3637924385442934, + "language_loss": 0.800883, + "learning_rate": 3.416732195234464e-07, + "loss": 0.8228699, + "num_input_tokens_seen": 146658030, + "step": 6793, + "time_per_iteration": 3.477916955947876 + }, + { + "auxiliary_loss_clip": 0.01180098, + "auxiliary_loss_mlp": 0.01024431, + "balance_loss_clip": 1.04511738, + "balance_loss_mlp": 1.01701021, + "epoch": 0.816930199001984, + "flos": 18407666833920.0, + "grad_norm": 1.5687416712609483, + "language_loss": 0.79627299, + "learning_rate": 3.4123789523941613e-07, + "loss": 0.81831825, + "num_input_tokens_seen": 146677855, + "step": 6794, + "time_per_iteration": 2.596734046936035 + }, + { + "auxiliary_loss_clip": 0.01169598, + "auxiliary_loss_mlp": 0.01025562, + "balance_loss_clip": 1.04340506, + "balance_loss_mlp": 1.01847792, + "epoch": 0.8170504418926231, + "flos": 21251324799360.0, + "grad_norm": 1.435668973465011, + "language_loss": 0.6319375, + "learning_rate": 3.4080282258827884e-07, + "loss": 0.65388918, + "num_input_tokens_seen": 146696230, + "step": 6795, + "time_per_iteration": 2.5916574001312256 + }, + { + "auxiliary_loss_clip": 0.01178422, + "auxiliary_loss_mlp": 0.01024026, + "balance_loss_clip": 1.04518187, + "balance_loss_mlp": 1.0163641, + "epoch": 0.8171706847832622, + "flos": 19099234362240.0, + "grad_norm": 2.0157597058966576, + "language_loss": 0.72412932, + "learning_rate": 3.403680016360342e-07, + "loss": 0.74615377, + "num_input_tokens_seen": 146714835, + "step": 6796, + "time_per_iteration": 3.522017478942871 + }, + { + "auxiliary_loss_clip": 0.01171226, + "auxiliary_loss_mlp": 0.0102611, + "balance_loss_clip": 1.04753947, + "balance_loss_mlp": 1.01839721, + "epoch": 0.8172909276739013, + "flos": 21470128496640.0, + "grad_norm": 4.5091854611458135, + "language_loss": 0.67422599, + "learning_rate": 3.3993343244864403e-07, + "loss": 0.69619936, + "num_input_tokens_seen": 146734425, + "step": 6797, + "time_per_iteration": 2.571536064147949 + }, + { + "auxiliary_loss_clip": 0.01177034, + "auxiliary_loss_mlp": 0.01024028, + "balance_loss_clip": 1.04707193, + "balance_loss_mlp": 1.01710248, + "epoch": 0.8174111705645404, + "flos": 27599792417280.0, + "grad_norm": 1.83921431585598, + "language_loss": 0.72960448, + "learning_rate": 3.394991150920323e-07, + "loss": 0.75161505, + "num_input_tokens_seen": 146757545, + "step": 6798, + "time_per_iteration": 2.6633148193359375 + }, + { + "auxiliary_loss_clip": 0.01204973, + "auxiliary_loss_mlp": 0.01378927, + "balance_loss_clip": 1.04216659, + "balance_loss_mlp": 1.00023341, + "epoch": 0.8175314134551794, + "flos": 14064594508800.0, + "grad_norm": 2.3715153067947203, + "language_loss": 0.74411476, + "learning_rate": 3.3906504963208396e-07, + "loss": 0.76995373, + "num_input_tokens_seen": 146774240, + "step": 6799, + "time_per_iteration": 2.650131940841675 + }, + { + "auxiliary_loss_clip": 0.01199865, + "auxiliary_loss_mlp": 0.01027935, + "balance_loss_clip": 1.04483223, + "balance_loss_mlp": 1.01991236, + "epoch": 0.8176516563458186, + "flos": 22708076780160.0, + "grad_norm": 1.8943169223853389, + "language_loss": 0.66553777, + "learning_rate": 3.3863123613464774e-07, + "loss": 0.68781579, + "num_input_tokens_seen": 146793140, + "step": 6800, + "time_per_iteration": 2.748990058898926 + }, + { + "auxiliary_loss_clip": 0.01185122, + "auxiliary_loss_mlp": 0.01020319, + "balance_loss_clip": 1.04039049, + "balance_loss_mlp": 1.01330984, + "epoch": 0.8177718992364577, + "flos": 21945406279680.0, + "grad_norm": 1.7671091812290007, + "language_loss": 0.74949801, + "learning_rate": 3.381976746655317e-07, + "loss": 0.77155244, + "num_input_tokens_seen": 146812895, + "step": 6801, + "time_per_iteration": 2.648547410964966 + }, + { + "auxiliary_loss_clip": 0.01198796, + "auxiliary_loss_mlp": 0.0102625, + "balance_loss_clip": 1.04466224, + "balance_loss_mlp": 1.01895738, + "epoch": 0.8178921421270967, + "flos": 22017443005440.0, + "grad_norm": 2.3921760493324684, + "language_loss": 0.67310047, + "learning_rate": 3.3776436529050756e-07, + "loss": 0.695351, + "num_input_tokens_seen": 146832445, + "step": 6802, + "time_per_iteration": 2.7594518661499023 + }, + { + "auxiliary_loss_clip": 0.01166323, + "auxiliary_loss_mlp": 0.01023187, + "balance_loss_clip": 1.04605079, + "balance_loss_mlp": 1.01576328, + "epoch": 0.8180123850177359, + "flos": 33183111496320.0, + "grad_norm": 1.601077061215155, + "language_loss": 0.72424001, + "learning_rate": 3.373313080753073e-07, + "loss": 0.74613506, + "num_input_tokens_seen": 146856505, + "step": 6803, + "time_per_iteration": 2.7148542404174805 + }, + { + "auxiliary_loss_clip": 0.0117175, + "auxiliary_loss_mlp": 0.01026805, + "balance_loss_clip": 1.04362893, + "balance_loss_mlp": 1.01946735, + "epoch": 0.8181326279083749, + "flos": 22091167670400.0, + "grad_norm": 4.315331894532294, + "language_loss": 0.77506983, + "learning_rate": 3.3689850308562527e-07, + "loss": 0.79705542, + "num_input_tokens_seen": 146876950, + "step": 6804, + "time_per_iteration": 2.5617611408233643 + }, + { + "auxiliary_loss_clip": 0.01198269, + "auxiliary_loss_mlp": 0.01032538, + "balance_loss_clip": 1.04624653, + "balance_loss_mlp": 1.02456307, + "epoch": 0.818252870799014, + "flos": 15705747936000.0, + "grad_norm": 4.4341297386166705, + "language_loss": 0.77515131, + "learning_rate": 3.364659503871183e-07, + "loss": 0.79745936, + "num_input_tokens_seen": 146894885, + "step": 6805, + "time_per_iteration": 2.6887927055358887 + }, + { + "auxiliary_loss_clip": 0.01189829, + "auxiliary_loss_mlp": 0.01020268, + "balance_loss_clip": 1.03996527, + "balance_loss_mlp": 1.01324606, + "epoch": 0.8183731136896532, + "flos": 18770687637120.0, + "grad_norm": 1.8306359286088558, + "language_loss": 0.83754414, + "learning_rate": 3.3603365004540417e-07, + "loss": 0.85964513, + "num_input_tokens_seen": 146913180, + "step": 6806, + "time_per_iteration": 2.6526472568511963 + }, + { + "auxiliary_loss_clip": 0.01165737, + "auxiliary_loss_mlp": 0.01026317, + "balance_loss_clip": 1.04556715, + "balance_loss_mlp": 1.01905465, + "epoch": 0.8184933565802922, + "flos": 26541792293760.0, + "grad_norm": 1.8526125960837552, + "language_loss": 0.77332366, + "learning_rate": 3.356016021260624e-07, + "loss": 0.79524422, + "num_input_tokens_seen": 146933510, + "step": 6807, + "time_per_iteration": 2.559506893157959 + }, + { + "auxiliary_loss_clip": 0.01177411, + "auxiliary_loss_mlp": 0.01025857, + "balance_loss_clip": 1.04736423, + "balance_loss_mlp": 1.01791739, + "epoch": 0.8186135994709313, + "flos": 17530117660800.0, + "grad_norm": 2.553150768275218, + "language_loss": 0.66085738, + "learning_rate": 3.35169806694634e-07, + "loss": 0.68289006, + "num_input_tokens_seen": 146951760, + "step": 6808, + "time_per_iteration": 2.519566535949707 + }, + { + "auxiliary_loss_clip": 0.01097761, + "auxiliary_loss_mlp": 0.01001757, + "balance_loss_clip": 1.01176858, + "balance_loss_mlp": 1.00049388, + "epoch": 0.8187338423615703, + "flos": 63480300675840.0, + "grad_norm": 0.7153608823735254, + "language_loss": 0.6059984, + "learning_rate": 3.3473826381662186e-07, + "loss": 0.62699354, + "num_input_tokens_seen": 147022900, + "step": 6809, + "time_per_iteration": 3.3827404975891113 + }, + { + "auxiliary_loss_clip": 0.01172023, + "auxiliary_loss_mlp": 0.01025061, + "balance_loss_clip": 1.04618609, + "balance_loss_mlp": 1.01834679, + "epoch": 0.8188540852522095, + "flos": 17529974006400.0, + "grad_norm": 2.701214168638009, + "language_loss": 0.81382346, + "learning_rate": 3.3430697355749216e-07, + "loss": 0.83579433, + "num_input_tokens_seen": 147040590, + "step": 6810, + "time_per_iteration": 2.60941743850708 + }, + { + "auxiliary_loss_clip": 0.01197038, + "auxiliary_loss_mlp": 0.01023336, + "balance_loss_clip": 1.04120553, + "balance_loss_mlp": 1.01618087, + "epoch": 0.8189743281428485, + "flos": 14392530702720.0, + "grad_norm": 2.048142443476825, + "language_loss": 0.75500411, + "learning_rate": 3.3387593598266907e-07, + "loss": 0.77720779, + "num_input_tokens_seen": 147057200, + "step": 6811, + "time_per_iteration": 2.808011293411255 + }, + { + "auxiliary_loss_clip": 0.01183607, + "auxiliary_loss_mlp": 0.01024322, + "balance_loss_clip": 1.04031396, + "balance_loss_mlp": 1.01717877, + "epoch": 0.8190945710334876, + "flos": 25080479285760.0, + "grad_norm": 1.6131038233828658, + "language_loss": 0.78050834, + "learning_rate": 3.3344515115754225e-07, + "loss": 0.80258763, + "num_input_tokens_seen": 147076180, + "step": 6812, + "time_per_iteration": 2.682234764099121 + }, + { + "auxiliary_loss_clip": 0.01197458, + "auxiliary_loss_mlp": 0.01025592, + "balance_loss_clip": 1.04195046, + "balance_loss_mlp": 1.01818347, + "epoch": 0.8192148139241268, + "flos": 21507152440320.0, + "grad_norm": 2.312192809872238, + "language_loss": 0.80039251, + "learning_rate": 3.33014619147461e-07, + "loss": 0.82262295, + "num_input_tokens_seen": 147094205, + "step": 6813, + "time_per_iteration": 2.658773899078369 + }, + { + "auxiliary_loss_clip": 0.01185575, + "auxiliary_loss_mlp": 0.01027747, + "balance_loss_clip": 1.04667735, + "balance_loss_mlp": 1.02009988, + "epoch": 0.8193350568147658, + "flos": 23952166289280.0, + "grad_norm": 1.8164484239401593, + "language_loss": 0.712879, + "learning_rate": 3.325843400177362e-07, + "loss": 0.73501223, + "num_input_tokens_seen": 147115545, + "step": 6814, + "time_per_iteration": 3.5702383518218994 + }, + { + "auxiliary_loss_clip": 0.01178176, + "auxiliary_loss_mlp": 0.0137865, + "balance_loss_clip": 1.04528308, + "balance_loss_mlp": 1.00018501, + "epoch": 0.8194552997054049, + "flos": 20559469962240.0, + "grad_norm": 1.8534045241897683, + "language_loss": 0.73534262, + "learning_rate": 3.32154313833642e-07, + "loss": 0.76091087, + "num_input_tokens_seen": 147135700, + "step": 6815, + "time_per_iteration": 2.669865369796753 + }, + { + "auxiliary_loss_clip": 0.01168752, + "auxiliary_loss_mlp": 0.01026785, + "balance_loss_clip": 1.04649055, + "balance_loss_mlp": 1.01920927, + "epoch": 0.819575542596044, + "flos": 26031753123840.0, + "grad_norm": 1.9552884746915313, + "language_loss": 0.59347576, + "learning_rate": 3.3172454066041164e-07, + "loss": 0.61543113, + "num_input_tokens_seen": 147155205, + "step": 6816, + "time_per_iteration": 2.7069101333618164 + }, + { + "auxiliary_loss_clip": 0.01213726, + "auxiliary_loss_mlp": 0.0137832, + "balance_loss_clip": 1.04429173, + "balance_loss_mlp": 1.00013208, + "epoch": 0.8196957854866831, + "flos": 29096944220160.0, + "grad_norm": 1.8732880594637833, + "language_loss": 0.76021719, + "learning_rate": 3.3129502056324234e-07, + "loss": 0.7861377, + "num_input_tokens_seen": 147176570, + "step": 6817, + "time_per_iteration": 3.745703935623169 + }, + { + "auxiliary_loss_clip": 0.01143025, + "auxiliary_loss_mlp": 0.01002197, + "balance_loss_clip": 1.01614475, + "balance_loss_mlp": 1.00094557, + "epoch": 0.8198160283773221, + "flos": 69033631898880.0, + "grad_norm": 0.7921421323506861, + "language_loss": 0.59680545, + "learning_rate": 3.3086575360729165e-07, + "loss": 0.61825764, + "num_input_tokens_seen": 147234105, + "step": 6818, + "time_per_iteration": 3.3272712230682373 + }, + { + "auxiliary_loss_clip": 0.01184426, + "auxiliary_loss_mlp": 0.01025036, + "balance_loss_clip": 1.04562974, + "balance_loss_mlp": 1.01732278, + "epoch": 0.8199362712679613, + "flos": 16618058496000.0, + "grad_norm": 1.6512305356836317, + "language_loss": 0.71375072, + "learning_rate": 3.3043673985767906e-07, + "loss": 0.73584533, + "num_input_tokens_seen": 147253170, + "step": 6819, + "time_per_iteration": 4.340892791748047 + }, + { + "auxiliary_loss_clip": 0.01180028, + "auxiliary_loss_mlp": 0.01024396, + "balance_loss_clip": 1.03814304, + "balance_loss_mlp": 1.01668286, + "epoch": 0.8200565141586004, + "flos": 21757664868480.0, + "grad_norm": 1.8561295664845325, + "language_loss": 0.77447104, + "learning_rate": 3.3000797937948564e-07, + "loss": 0.79651523, + "num_input_tokens_seen": 147271465, + "step": 6820, + "time_per_iteration": 2.7023589611053467 + }, + { + "auxiliary_loss_clip": 0.01101445, + "auxiliary_loss_mlp": 0.01000524, + "balance_loss_clip": 1.01029229, + "balance_loss_mlp": 0.9994213, + "epoch": 0.8201767570492394, + "flos": 69807112392960.0, + "grad_norm": 0.9352811776441952, + "language_loss": 0.65007555, + "learning_rate": 3.295794722377534e-07, + "loss": 0.67109525, + "num_input_tokens_seen": 147335070, + "step": 6821, + "time_per_iteration": 3.239522933959961 + }, + { + "auxiliary_loss_clip": 0.01167692, + "auxiliary_loss_mlp": 0.01022259, + "balance_loss_clip": 1.04753494, + "balance_loss_mlp": 1.01540399, + "epoch": 0.8202969999398786, + "flos": 23111892455040.0, + "grad_norm": 1.5975145266732045, + "language_loss": 0.79818726, + "learning_rate": 3.291512184974876e-07, + "loss": 0.82008672, + "num_input_tokens_seen": 147355460, + "step": 6822, + "time_per_iteration": 3.844017267227173 + }, + { + "auxiliary_loss_clip": 0.01182599, + "auxiliary_loss_mlp": 0.01026094, + "balance_loss_clip": 1.042377, + "balance_loss_mlp": 1.01805949, + "epoch": 0.8204172428305176, + "flos": 28220616109440.0, + "grad_norm": 5.96703920149104, + "language_loss": 0.66542667, + "learning_rate": 3.2872321822365346e-07, + "loss": 0.68751365, + "num_input_tokens_seen": 147375675, + "step": 6823, + "time_per_iteration": 2.692232370376587 + }, + { + "auxiliary_loss_clip": 0.01176435, + "auxiliary_loss_mlp": 0.01025476, + "balance_loss_clip": 1.04690909, + "balance_loss_mlp": 1.01826632, + "epoch": 0.8205374857211567, + "flos": 20887011106560.0, + "grad_norm": 1.8522631859747873, + "language_loss": 0.7332046, + "learning_rate": 3.282954714811783e-07, + "loss": 0.75522363, + "num_input_tokens_seen": 147394580, + "step": 6824, + "time_per_iteration": 2.593578815460205 + }, + { + "auxiliary_loss_clip": 0.01170356, + "auxiliary_loss_mlp": 0.01023524, + "balance_loss_clip": 1.04146051, + "balance_loss_mlp": 1.01619864, + "epoch": 0.8206577286117959, + "flos": 13152140294400.0, + "grad_norm": 2.3144148322492124, + "language_loss": 0.70373774, + "learning_rate": 3.2786797833495093e-07, + "loss": 0.72567654, + "num_input_tokens_seen": 147409935, + "step": 6825, + "time_per_iteration": 2.605731964111328 + }, + { + "auxiliary_loss_clip": 0.01166473, + "auxiliary_loss_mlp": 0.01024908, + "balance_loss_clip": 1.0469172, + "balance_loss_mlp": 1.01726127, + "epoch": 0.8207779715024349, + "flos": 25265634917760.0, + "grad_norm": 2.005163227958599, + "language_loss": 0.72764945, + "learning_rate": 3.274407388498213e-07, + "loss": 0.74956328, + "num_input_tokens_seen": 147428065, + "step": 6826, + "time_per_iteration": 2.5779502391815186 + }, + { + "auxiliary_loss_clip": 0.01186598, + "auxiliary_loss_mlp": 0.01026034, + "balance_loss_clip": 1.04106593, + "balance_loss_mlp": 1.01867831, + "epoch": 0.820898214393074, + "flos": 19610243199360.0, + "grad_norm": 2.0366048021729273, + "language_loss": 0.74171209, + "learning_rate": 3.270137530906021e-07, + "loss": 0.76383841, + "num_input_tokens_seen": 147447300, + "step": 6827, + "time_per_iteration": 2.7670819759368896 + }, + { + "auxiliary_loss_clip": 0.01194637, + "auxiliary_loss_mlp": 0.01025972, + "balance_loss_clip": 1.04284382, + "balance_loss_mlp": 1.01912379, + "epoch": 0.8210184572837131, + "flos": 15596615439360.0, + "grad_norm": 1.8017601452938392, + "language_loss": 0.83554029, + "learning_rate": 3.265870211220665e-07, + "loss": 0.85774636, + "num_input_tokens_seen": 147465135, + "step": 6828, + "time_per_iteration": 2.6955573558807373 + }, + { + "auxiliary_loss_clip": 0.01189237, + "auxiliary_loss_mlp": 0.01032058, + "balance_loss_clip": 1.04279935, + "balance_loss_mlp": 1.02386844, + "epoch": 0.8211387001743522, + "flos": 20813932886400.0, + "grad_norm": 1.7962935402271936, + "language_loss": 0.81406903, + "learning_rate": 3.2616054300894934e-07, + "loss": 0.83628201, + "num_input_tokens_seen": 147484585, + "step": 6829, + "time_per_iteration": 2.724874496459961 + }, + { + "auxiliary_loss_clip": 0.01172008, + "auxiliary_loss_mlp": 0.01028638, + "balance_loss_clip": 1.0432514, + "balance_loss_mlp": 1.0210799, + "epoch": 0.8212589430649913, + "flos": 27704579368320.0, + "grad_norm": 2.2764507841487975, + "language_loss": 0.84217358, + "learning_rate": 3.2573431881594693e-07, + "loss": 0.86418003, + "num_input_tokens_seen": 147504130, + "step": 6830, + "time_per_iteration": 2.7000339031219482 + }, + { + "auxiliary_loss_clip": 0.01209, + "auxiliary_loss_mlp": 0.01023995, + "balance_loss_clip": 1.03784561, + "balance_loss_mlp": 1.01681828, + "epoch": 0.8213791859556304, + "flos": 22455625017600.0, + "grad_norm": 4.795970438253891, + "language_loss": 0.65948379, + "learning_rate": 3.2530834860771663e-07, + "loss": 0.68181372, + "num_input_tokens_seen": 147523510, + "step": 6831, + "time_per_iteration": 2.7982165813446045 + }, + { + "auxiliary_loss_clip": 0.01175774, + "auxiliary_loss_mlp": 0.01023613, + "balance_loss_clip": 1.04409635, + "balance_loss_mlp": 1.01640117, + "epoch": 0.8214994288462695, + "flos": 16654471908480.0, + "grad_norm": 2.2775949829586146, + "language_loss": 0.74008441, + "learning_rate": 3.248826324488794e-07, + "loss": 0.76207829, + "num_input_tokens_seen": 147540805, + "step": 6832, + "time_per_iteration": 2.6004128456115723 + }, + { + "auxiliary_loss_clip": 0.01171527, + "auxiliary_loss_mlp": 0.01023579, + "balance_loss_clip": 1.05133545, + "balance_loss_mlp": 1.01646173, + "epoch": 0.8216196717369085, + "flos": 25221787390080.0, + "grad_norm": 1.7381784606969222, + "language_loss": 0.87610841, + "learning_rate": 3.244571704040138e-07, + "loss": 0.89805943, + "num_input_tokens_seen": 147560965, + "step": 6833, + "time_per_iteration": 2.576399564743042 + }, + { + "auxiliary_loss_clip": 0.01171188, + "auxiliary_loss_mlp": 0.01025974, + "balance_loss_clip": 1.04318368, + "balance_loss_mlp": 1.01860714, + "epoch": 0.8217399146275477, + "flos": 25371930240000.0, + "grad_norm": 1.8798124087293955, + "language_loss": 0.73354995, + "learning_rate": 3.2403196253766374e-07, + "loss": 0.75552154, + "num_input_tokens_seen": 147580045, + "step": 6834, + "time_per_iteration": 2.6006972789764404 + }, + { + "auxiliary_loss_clip": 0.01172315, + "auxiliary_loss_mlp": 0.01029218, + "balance_loss_clip": 1.04540515, + "balance_loss_mlp": 1.02131426, + "epoch": 0.8218601575181868, + "flos": 25629625388160.0, + "grad_norm": 4.0387588329419675, + "language_loss": 0.7900008, + "learning_rate": 3.2360700891433254e-07, + "loss": 0.81201613, + "num_input_tokens_seen": 147599070, + "step": 6835, + "time_per_iteration": 2.6510047912597656 + }, + { + "auxiliary_loss_clip": 0.01111527, + "auxiliary_loss_mlp": 0.0100133, + "balance_loss_clip": 1.01030517, + "balance_loss_mlp": 1.00023937, + "epoch": 0.8219804004088258, + "flos": 67660229427840.0, + "grad_norm": 0.7887561933563185, + "language_loss": 0.57251251, + "learning_rate": 3.231823095984847e-07, + "loss": 0.59364104, + "num_input_tokens_seen": 147653710, + "step": 6836, + "time_per_iteration": 3.247420310974121 + }, + { + "auxiliary_loss_clip": 0.01185129, + "auxiliary_loss_mlp": 0.0102362, + "balance_loss_clip": 1.04487622, + "balance_loss_mlp": 1.01613665, + "epoch": 0.822100643299465, + "flos": 19464266327040.0, + "grad_norm": 3.192707214188195, + "language_loss": 0.7613194, + "learning_rate": 3.2275786465454814e-07, + "loss": 0.78340685, + "num_input_tokens_seen": 147670360, + "step": 6837, + "time_per_iteration": 2.5911359786987305 + }, + { + "auxiliary_loss_clip": 0.01193319, + "auxiliary_loss_mlp": 0.01025417, + "balance_loss_clip": 1.04275656, + "balance_loss_mlp": 1.01844907, + "epoch": 0.822220886190104, + "flos": 24681368292480.0, + "grad_norm": 2.156621346921314, + "language_loss": 0.75865883, + "learning_rate": 3.2233367414690917e-07, + "loss": 0.78084612, + "num_input_tokens_seen": 147692550, + "step": 6838, + "time_per_iteration": 2.7003393173217773 + }, + { + "auxiliary_loss_clip": 0.01188812, + "auxiliary_loss_mlp": 0.01023407, + "balance_loss_clip": 1.0409081, + "balance_loss_mlp": 1.01654959, + "epoch": 0.8223411290807431, + "flos": 27819062991360.0, + "grad_norm": 2.411845257529495, + "language_loss": 0.84604335, + "learning_rate": 3.219097381399183e-07, + "loss": 0.86816549, + "num_input_tokens_seen": 147709725, + "step": 6839, + "time_per_iteration": 2.6967222690582275 + }, + { + "auxiliary_loss_clip": 0.01194628, + "auxiliary_loss_mlp": 0.01028697, + "balance_loss_clip": 1.04638016, + "balance_loss_mlp": 1.02084088, + "epoch": 0.8224613719713821, + "flos": 23218546913280.0, + "grad_norm": 1.8146422403726723, + "language_loss": 0.80708504, + "learning_rate": 3.2148605669788584e-07, + "loss": 0.82931828, + "num_input_tokens_seen": 147729615, + "step": 6840, + "time_per_iteration": 3.7106094360351562 + }, + { + "auxiliary_loss_clip": 0.01190713, + "auxiliary_loss_mlp": 0.01023431, + "balance_loss_clip": 1.04722309, + "balance_loss_mlp": 1.01628494, + "epoch": 0.8225816148620213, + "flos": 15706250726400.0, + "grad_norm": 2.9995740477423842, + "language_loss": 0.77188623, + "learning_rate": 3.2106262988508405e-07, + "loss": 0.79402769, + "num_input_tokens_seen": 147747665, + "step": 6841, + "time_per_iteration": 2.6343886852264404 + }, + { + "auxiliary_loss_clip": 0.01184456, + "auxiliary_loss_mlp": 0.01023587, + "balance_loss_clip": 1.04304123, + "balance_loss_mlp": 1.0155642, + "epoch": 0.8227018577526604, + "flos": 18515111391360.0, + "grad_norm": 2.3707507652976885, + "language_loss": 0.74553752, + "learning_rate": 3.206394577657465e-07, + "loss": 0.76761794, + "num_input_tokens_seen": 147765445, + "step": 6842, + "time_per_iteration": 3.5090503692626953 + }, + { + "auxiliary_loss_clip": 0.01181017, + "auxiliary_loss_mlp": 0.01024756, + "balance_loss_clip": 1.04712653, + "balance_loss_mlp": 1.01676869, + "epoch": 0.8228221006432994, + "flos": 22236785406720.0, + "grad_norm": 3.146165252630284, + "language_loss": 0.73078507, + "learning_rate": 3.202165404040675e-07, + "loss": 0.75284278, + "num_input_tokens_seen": 147783365, + "step": 6843, + "time_per_iteration": 2.6358165740966797 + }, + { + "auxiliary_loss_clip": 0.01210171, + "auxiliary_loss_mlp": 0.01031166, + "balance_loss_clip": 1.0414108, + "balance_loss_mlp": 1.02378964, + "epoch": 0.8229423435339386, + "flos": 24097532630400.0, + "grad_norm": 1.8659956347301117, + "language_loss": 0.74721718, + "learning_rate": 3.1979387786420396e-07, + "loss": 0.76963055, + "num_input_tokens_seen": 147803605, + "step": 6844, + "time_per_iteration": 3.6882050037384033 + }, + { + "auxiliary_loss_clip": 0.01184213, + "auxiliary_loss_mlp": 0.01027384, + "balance_loss_clip": 1.04237044, + "balance_loss_mlp": 1.01961756, + "epoch": 0.8230625864245776, + "flos": 23878549365120.0, + "grad_norm": 2.37076023758627, + "language_loss": 0.82055211, + "learning_rate": 3.1937147021027346e-07, + "loss": 0.842668, + "num_input_tokens_seen": 147822060, + "step": 6845, + "time_per_iteration": 2.6491644382476807 + }, + { + "auxiliary_loss_clip": 0.01175546, + "auxiliary_loss_mlp": 0.01023904, + "balance_loss_clip": 1.04599571, + "balance_loss_mlp": 1.01710653, + "epoch": 0.8231828293152167, + "flos": 16581106379520.0, + "grad_norm": 2.1082281361922255, + "language_loss": 0.76839095, + "learning_rate": 3.189493175063547e-07, + "loss": 0.79038548, + "num_input_tokens_seen": 147839295, + "step": 6846, + "time_per_iteration": 2.5499484539031982 + }, + { + "auxiliary_loss_clip": 0.0118871, + "auxiliary_loss_mlp": 0.01031362, + "balance_loss_clip": 1.04691458, + "balance_loss_mlp": 1.02390838, + "epoch": 0.8233030722058559, + "flos": 18880071528960.0, + "grad_norm": 2.0479634298537284, + "language_loss": 0.6727941, + "learning_rate": 3.1852741981648776e-07, + "loss": 0.69499487, + "num_input_tokens_seen": 147857945, + "step": 6847, + "time_per_iteration": 3.577348232269287 + }, + { + "auxiliary_loss_clip": 0.01180902, + "auxiliary_loss_mlp": 0.01026244, + "balance_loss_clip": 1.04146791, + "balance_loss_mlp": 1.01861429, + "epoch": 0.8234233150964949, + "flos": 28439024757120.0, + "grad_norm": 2.139389605975689, + "language_loss": 0.69940269, + "learning_rate": 3.1810577720467404e-07, + "loss": 0.72147417, + "num_input_tokens_seen": 147879675, + "step": 6848, + "time_per_iteration": 2.717810869216919 + }, + { + "auxiliary_loss_clip": 0.01191001, + "auxiliary_loss_mlp": 0.01023671, + "balance_loss_clip": 1.04575682, + "balance_loss_mlp": 1.01605678, + "epoch": 0.823543557987134, + "flos": 33765941577600.0, + "grad_norm": 1.5389512136548067, + "language_loss": 0.5654012, + "learning_rate": 3.176843897348769e-07, + "loss": 0.5875479, + "num_input_tokens_seen": 147902870, + "step": 6849, + "time_per_iteration": 2.746870756149292 + }, + { + "auxiliary_loss_clip": 0.01178843, + "auxiliary_loss_mlp": 0.01025055, + "balance_loss_clip": 1.04315364, + "balance_loss_mlp": 1.01722312, + "epoch": 0.8236638008777731, + "flos": 17092366611840.0, + "grad_norm": 2.516299634646279, + "language_loss": 0.7546109, + "learning_rate": 3.1726325747102034e-07, + "loss": 0.77664989, + "num_input_tokens_seen": 147921245, + "step": 6850, + "time_per_iteration": 2.594120502471924 + }, + { + "auxiliary_loss_clip": 0.01192685, + "auxiliary_loss_mlp": 0.01028707, + "balance_loss_clip": 1.03778899, + "balance_loss_mlp": 1.02116656, + "epoch": 0.8237840437684122, + "flos": 61639982334720.0, + "grad_norm": 1.4892238917091112, + "language_loss": 0.64348745, + "learning_rate": 3.1684238047698974e-07, + "loss": 0.66570139, + "num_input_tokens_seen": 147949515, + "step": 6851, + "time_per_iteration": 3.0726592540740967 + }, + { + "auxiliary_loss_clip": 0.01188395, + "auxiliary_loss_mlp": 0.01024303, + "balance_loss_clip": 1.04418099, + "balance_loss_mlp": 1.01609576, + "epoch": 0.8239042866590512, + "flos": 27309023821440.0, + "grad_norm": 1.8891808940419812, + "language_loss": 0.52930236, + "learning_rate": 3.1642175881663155e-07, + "loss": 0.55142933, + "num_input_tokens_seen": 147969245, + "step": 6852, + "time_per_iteration": 2.6852948665618896 + }, + { + "auxiliary_loss_clip": 0.01166836, + "auxiliary_loss_mlp": 0.01023681, + "balance_loss_clip": 1.04617405, + "balance_loss_mlp": 1.01666594, + "epoch": 0.8240245295496904, + "flos": 21726351187200.0, + "grad_norm": 2.539309160039073, + "language_loss": 0.84084386, + "learning_rate": 3.160013925537537e-07, + "loss": 0.86274904, + "num_input_tokens_seen": 147990080, + "step": 6853, + "time_per_iteration": 2.562659740447998 + }, + { + "auxiliary_loss_clip": 0.01197317, + "auxiliary_loss_mlp": 0.01026624, + "balance_loss_clip": 1.0433023, + "balance_loss_mlp": 1.01919687, + "epoch": 0.8241447724403295, + "flos": 20009318279040.0, + "grad_norm": 1.980028039383439, + "language_loss": 0.75664598, + "learning_rate": 3.155812817521266e-07, + "loss": 0.77888536, + "num_input_tokens_seen": 148010455, + "step": 6854, + "time_per_iteration": 2.727858304977417 + }, + { + "auxiliary_loss_clip": 0.01191402, + "auxiliary_loss_mlp": 0.01029056, + "balance_loss_clip": 1.04854488, + "balance_loss_mlp": 1.02099752, + "epoch": 0.8242650153309685, + "flos": 22272983337600.0, + "grad_norm": 2.148909028694383, + "language_loss": 0.7793715, + "learning_rate": 3.151614264754787e-07, + "loss": 0.80157608, + "num_input_tokens_seen": 148028400, + "step": 6855, + "time_per_iteration": 2.649226188659668 + }, + { + "auxiliary_loss_clip": 0.01169103, + "auxiliary_loss_mlp": 0.0102798, + "balance_loss_clip": 1.04619956, + "balance_loss_mlp": 1.02005291, + "epoch": 0.8243852582216077, + "flos": 22309971367680.0, + "grad_norm": 2.443813375902508, + "language_loss": 0.79790807, + "learning_rate": 3.147418267875035e-07, + "loss": 0.81987888, + "num_input_tokens_seen": 148046530, + "step": 6856, + "time_per_iteration": 2.54933500289917 + }, + { + "auxiliary_loss_clip": 0.01205538, + "auxiliary_loss_mlp": 0.01378669, + "balance_loss_clip": 1.03685832, + "balance_loss_mlp": 1.0002079, + "epoch": 0.8245055011122467, + "flos": 24645421756800.0, + "grad_norm": 2.3826956315913286, + "language_loss": 0.65668476, + "learning_rate": 3.1432248275185315e-07, + "loss": 0.68252683, + "num_input_tokens_seen": 148067040, + "step": 6857, + "time_per_iteration": 2.7996926307678223 + }, + { + "auxiliary_loss_clip": 0.01174304, + "auxiliary_loss_mlp": 0.01022796, + "balance_loss_clip": 1.04623222, + "balance_loss_mlp": 1.01533914, + "epoch": 0.8246257440028858, + "flos": 17487275713920.0, + "grad_norm": 2.0402862992023127, + "language_loss": 0.77311176, + "learning_rate": 3.139033944321412e-07, + "loss": 0.79508269, + "num_input_tokens_seen": 148084400, + "step": 6858, + "time_per_iteration": 2.5857789516448975 + }, + { + "auxiliary_loss_clip": 0.01180849, + "auxiliary_loss_mlp": 0.01024647, + "balance_loss_clip": 1.04626131, + "balance_loss_mlp": 1.0170089, + "epoch": 0.824745986893525, + "flos": 25010130499200.0, + "grad_norm": 1.6854980339698884, + "language_loss": 0.78614062, + "learning_rate": 3.1348456189194507e-07, + "loss": 0.80819559, + "num_input_tokens_seen": 148104860, + "step": 6859, + "time_per_iteration": 2.6488406658172607 + }, + { + "auxiliary_loss_clip": 0.01180048, + "auxiliary_loss_mlp": 0.01019812, + "balance_loss_clip": 1.03841984, + "balance_loss_mlp": 1.01235533, + "epoch": 0.824866229784164, + "flos": 18772698798720.0, + "grad_norm": 1.6780405010793704, + "language_loss": 0.82845622, + "learning_rate": 3.1306598519479876e-07, + "loss": 0.85045481, + "num_input_tokens_seen": 148124680, + "step": 6860, + "time_per_iteration": 2.659665107727051 + }, + { + "auxiliary_loss_clip": 0.01182046, + "auxiliary_loss_mlp": 0.01027377, + "balance_loss_clip": 1.04423094, + "balance_loss_mlp": 1.02024555, + "epoch": 0.8249864726748031, + "flos": 23842171866240.0, + "grad_norm": 1.7307309108459399, + "language_loss": 0.78164345, + "learning_rate": 3.1264766440420177e-07, + "loss": 0.80373764, + "num_input_tokens_seen": 148147150, + "step": 6861, + "time_per_iteration": 2.804492235183716 + }, + { + "auxiliary_loss_clip": 0.01175126, + "auxiliary_loss_mlp": 0.01025805, + "balance_loss_clip": 1.04682887, + "balance_loss_mlp": 1.01887918, + "epoch": 0.8251067155654422, + "flos": 20303103617280.0, + "grad_norm": 3.2916627317110625, + "language_loss": 0.68670726, + "learning_rate": 3.122295995836124e-07, + "loss": 0.70871657, + "num_input_tokens_seen": 148167020, + "step": 6862, + "time_per_iteration": 2.617769956588745 + }, + { + "auxiliary_loss_clip": 0.01179157, + "auxiliary_loss_mlp": 0.01027182, + "balance_loss_clip": 1.04305959, + "balance_loss_mlp": 1.01951122, + "epoch": 0.8252269584560813, + "flos": 25009699536000.0, + "grad_norm": 2.02742754705961, + "language_loss": 0.77248549, + "learning_rate": 3.118117907964508e-07, + "loss": 0.79454899, + "num_input_tokens_seen": 148188965, + "step": 6863, + "time_per_iteration": 2.687734603881836 + }, + { + "auxiliary_loss_clip": 0.01199006, + "auxiliary_loss_mlp": 0.01022086, + "balance_loss_clip": 1.04216528, + "balance_loss_mlp": 1.01528525, + "epoch": 0.8253472013467203, + "flos": 17128564542720.0, + "grad_norm": 1.9985237902147104, + "language_loss": 0.80462933, + "learning_rate": 3.1139423810609856e-07, + "loss": 0.82684028, + "num_input_tokens_seen": 148205660, + "step": 6864, + "time_per_iteration": 2.6780688762664795 + }, + { + "auxiliary_loss_clip": 0.01168068, + "auxiliary_loss_mlp": 0.01023252, + "balance_loss_clip": 1.04568231, + "balance_loss_mlp": 1.01631761, + "epoch": 0.8254674442373595, + "flos": 22414794232320.0, + "grad_norm": 1.9604287208189846, + "language_loss": 0.74965405, + "learning_rate": 3.1097694157589714e-07, + "loss": 0.77156734, + "num_input_tokens_seen": 148225545, + "step": 6865, + "time_per_iteration": 2.6366288661956787 + }, + { + "auxiliary_loss_clip": 0.01172748, + "auxiliary_loss_mlp": 0.01022577, + "balance_loss_clip": 1.04536569, + "balance_loss_mlp": 1.01520991, + "epoch": 0.8255876871279986, + "flos": 24786765774720.0, + "grad_norm": 6.573915213335645, + "language_loss": 0.76189834, + "learning_rate": 3.105599012691511e-07, + "loss": 0.78385156, + "num_input_tokens_seen": 148243975, + "step": 6866, + "time_per_iteration": 2.59333872795105 + }, + { + "auxiliary_loss_clip": 0.0117301, + "auxiliary_loss_mlp": 0.01019116, + "balance_loss_clip": 1.04441404, + "balance_loss_mlp": 1.01209509, + "epoch": 0.8257079300186376, + "flos": 27455431656960.0, + "grad_norm": 1.4876219520013207, + "language_loss": 0.82357931, + "learning_rate": 3.101431172491249e-07, + "loss": 0.84550059, + "num_input_tokens_seen": 148265520, + "step": 6867, + "time_per_iteration": 3.5268776416778564 + }, + { + "auxiliary_loss_clip": 0.01194986, + "auxiliary_loss_mlp": 0.01378963, + "balance_loss_clip": 1.04060221, + "balance_loss_mlp": 1.00013447, + "epoch": 0.8258281729092768, + "flos": 16471866142080.0, + "grad_norm": 2.0619550415163346, + "language_loss": 0.71553922, + "learning_rate": 3.097265895790444e-07, + "loss": 0.74127877, + "num_input_tokens_seen": 148283730, + "step": 6868, + "time_per_iteration": 3.585637092590332 + }, + { + "auxiliary_loss_clip": 0.0119313, + "auxiliary_loss_mlp": 0.01027813, + "balance_loss_clip": 1.04341388, + "balance_loss_mlp": 1.02056789, + "epoch": 0.8259484157999158, + "flos": 21433822824960.0, + "grad_norm": 2.297688652447442, + "language_loss": 0.8323518, + "learning_rate": 3.093103183220962e-07, + "loss": 0.85456121, + "num_input_tokens_seen": 148303775, + "step": 6869, + "time_per_iteration": 2.6441543102264404 + }, + { + "auxiliary_loss_clip": 0.01074843, + "auxiliary_loss_mlp": 0.01001395, + "balance_loss_clip": 1.01002765, + "balance_loss_mlp": 1.00029278, + "epoch": 0.8260686586905549, + "flos": 58322342453760.0, + "grad_norm": 0.8225716634762177, + "language_loss": 0.59377176, + "learning_rate": 3.0889430354142796e-07, + "loss": 0.61453414, + "num_input_tokens_seen": 148365285, + "step": 6870, + "time_per_iteration": 3.1616368293762207 + }, + { + "auxiliary_loss_clip": 0.0119553, + "auxiliary_loss_mlp": 0.0102248, + "balance_loss_clip": 1.04000616, + "balance_loss_mlp": 1.01539934, + "epoch": 0.826188901581194, + "flos": 27527288814720.0, + "grad_norm": 2.0873796887568656, + "language_loss": 0.69968671, + "learning_rate": 3.084785453001497e-07, + "loss": 0.72186685, + "num_input_tokens_seen": 148386200, + "step": 6871, + "time_per_iteration": 3.6963188648223877 + }, + { + "auxiliary_loss_clip": 0.01188384, + "auxiliary_loss_mlp": 0.01378685, + "balance_loss_clip": 1.04834378, + "balance_loss_mlp": 1.0001967, + "epoch": 0.8263091444718331, + "flos": 23696051339520.0, + "grad_norm": 2.0987153215769396, + "language_loss": 0.81813705, + "learning_rate": 3.080630436613314e-07, + "loss": 0.8438077, + "num_input_tokens_seen": 148403970, + "step": 6872, + "time_per_iteration": 2.6352624893188477 + }, + { + "auxiliary_loss_clip": 0.01164238, + "auxiliary_loss_mlp": 0.01021291, + "balance_loss_clip": 1.04230654, + "balance_loss_mlp": 1.01436758, + "epoch": 0.8264293873624722, + "flos": 17165157523200.0, + "grad_norm": 1.9893736326093892, + "language_loss": 0.85717529, + "learning_rate": 3.076477986880039e-07, + "loss": 0.87903059, + "num_input_tokens_seen": 148421765, + "step": 6873, + "time_per_iteration": 3.4868412017822266 + }, + { + "auxiliary_loss_clip": 0.01186283, + "auxiliary_loss_mlp": 0.01028374, + "balance_loss_clip": 1.04566669, + "balance_loss_mlp": 1.02066088, + "epoch": 0.8265496302531112, + "flos": 24098645952000.0, + "grad_norm": 2.221685859161567, + "language_loss": 0.69536126, + "learning_rate": 3.0723281044315986e-07, + "loss": 0.71750778, + "num_input_tokens_seen": 148443720, + "step": 6874, + "time_per_iteration": 2.628588914871216 + }, + { + "auxiliary_loss_clip": 0.01163224, + "auxiliary_loss_mlp": 0.01020044, + "balance_loss_clip": 1.04434252, + "balance_loss_mlp": 1.01358235, + "epoch": 0.8266698731437504, + "flos": 14099894599680.0, + "grad_norm": 2.171033904147911, + "language_loss": 0.7641539, + "learning_rate": 3.068180789897521e-07, + "loss": 0.78598654, + "num_input_tokens_seen": 148462130, + "step": 6875, + "time_per_iteration": 2.5293545722961426 + }, + { + "auxiliary_loss_clip": 0.0117889, + "auxiliary_loss_mlp": 0.01022678, + "balance_loss_clip": 1.04464984, + "balance_loss_mlp": 1.01545727, + "epoch": 0.8267901160343895, + "flos": 30777563715840.0, + "grad_norm": 1.4248993276690562, + "language_loss": 0.81495863, + "learning_rate": 3.064036043906966e-07, + "loss": 0.83697432, + "num_input_tokens_seen": 148485570, + "step": 6876, + "time_per_iteration": 2.717259168624878 + }, + { + "auxiliary_loss_clip": 0.01198431, + "auxiliary_loss_mlp": 0.01030171, + "balance_loss_clip": 1.04127276, + "balance_loss_mlp": 1.02224326, + "epoch": 0.8269103589250285, + "flos": 40624915242240.0, + "grad_norm": 2.2299543382624547, + "language_loss": 0.6807729, + "learning_rate": 3.059893867088668e-07, + "loss": 0.7030589, + "num_input_tokens_seen": 148509715, + "step": 6877, + "time_per_iteration": 2.8231875896453857 + }, + { + "auxiliary_loss_clip": 0.01174594, + "auxiliary_loss_mlp": 0.0102286, + "balance_loss_clip": 1.04571652, + "balance_loss_mlp": 1.01530194, + "epoch": 0.8270306018156677, + "flos": 30263645877120.0, + "grad_norm": 1.9863622464100361, + "language_loss": 0.66931945, + "learning_rate": 3.055754260071004e-07, + "loss": 0.69129395, + "num_input_tokens_seen": 148532010, + "step": 6878, + "time_per_iteration": 2.676929473876953 + }, + { + "auxiliary_loss_clip": 0.01177318, + "auxiliary_loss_mlp": 0.01026682, + "balance_loss_clip": 1.04557025, + "balance_loss_mlp": 1.01971769, + "epoch": 0.8271508447063067, + "flos": 25226599812480.0, + "grad_norm": 3.990933754117971, + "language_loss": 0.73590887, + "learning_rate": 3.051617223481948e-07, + "loss": 0.75794888, + "num_input_tokens_seen": 148553330, + "step": 6879, + "time_per_iteration": 2.6198081970214844 + }, + { + "auxiliary_loss_clip": 0.01203703, + "auxiliary_loss_mlp": 0.01028706, + "balance_loss_clip": 1.04308391, + "balance_loss_mlp": 1.02098083, + "epoch": 0.8272710875969458, + "flos": 17566602900480.0, + "grad_norm": 2.059877419558502, + "language_loss": 0.75097513, + "learning_rate": 3.047482757949078e-07, + "loss": 0.77329928, + "num_input_tokens_seen": 148570960, + "step": 6880, + "time_per_iteration": 2.6559531688690186 + }, + { + "auxiliary_loss_clip": 0.01191096, + "auxiliary_loss_mlp": 0.01378239, + "balance_loss_clip": 1.04201531, + "balance_loss_mlp": 1.00019252, + "epoch": 0.827391330487585, + "flos": 19755465886080.0, + "grad_norm": 1.869827403505638, + "language_loss": 0.85788113, + "learning_rate": 3.043350864099605e-07, + "loss": 0.88357449, + "num_input_tokens_seen": 148589520, + "step": 6881, + "time_per_iteration": 2.6387150287628174 + }, + { + "auxiliary_loss_clip": 0.01177965, + "auxiliary_loss_mlp": 0.0102594, + "balance_loss_clip": 1.04458499, + "balance_loss_mlp": 1.01890326, + "epoch": 0.827511573378224, + "flos": 16835174254080.0, + "grad_norm": 2.2065371541973353, + "language_loss": 0.80711174, + "learning_rate": 3.039221542560315e-07, + "loss": 0.8291508, + "num_input_tokens_seen": 148606085, + "step": 6882, + "time_per_iteration": 2.616635322570801 + }, + { + "auxiliary_loss_clip": 0.01175992, + "auxiliary_loss_mlp": 0.01028795, + "balance_loss_clip": 1.04577684, + "balance_loss_mlp": 1.02144265, + "epoch": 0.8276318162688631, + "flos": 18369242259840.0, + "grad_norm": 2.4326930401326026, + "language_loss": 0.7415213, + "learning_rate": 3.0350947939576356e-07, + "loss": 0.76356912, + "num_input_tokens_seen": 148625240, + "step": 6883, + "time_per_iteration": 2.636768341064453 + }, + { + "auxiliary_loss_clip": 0.01181879, + "auxiliary_loss_mlp": 0.01026488, + "balance_loss_clip": 1.046206, + "balance_loss_mlp": 1.01907277, + "epoch": 0.8277520591595022, + "flos": 19352691705600.0, + "grad_norm": 1.6735190109618676, + "language_loss": 0.72023982, + "learning_rate": 3.0309706189175876e-07, + "loss": 0.74232352, + "num_input_tokens_seen": 148645075, + "step": 6884, + "time_per_iteration": 2.636002540588379 + }, + { + "auxiliary_loss_clip": 0.0108681, + "auxiliary_loss_mlp": 0.01003184, + "balance_loss_clip": 1.00992346, + "balance_loss_mlp": 1.0021708, + "epoch": 0.8278723020501413, + "flos": 67918858329600.0, + "grad_norm": 0.7749153162803333, + "language_loss": 0.574202, + "learning_rate": 3.0268490180658045e-07, + "loss": 0.59510195, + "num_input_tokens_seen": 148707855, + "step": 6885, + "time_per_iteration": 3.211604118347168 + }, + { + "auxiliary_loss_clip": 0.01171166, + "auxiliary_loss_mlp": 0.01024949, + "balance_loss_clip": 1.04908824, + "balance_loss_mlp": 1.01737928, + "epoch": 0.8279925449407803, + "flos": 18185738653440.0, + "grad_norm": 2.146882675967367, + "language_loss": 0.7915715, + "learning_rate": 3.0227299920275305e-07, + "loss": 0.81353265, + "num_input_tokens_seen": 148724170, + "step": 6886, + "time_per_iteration": 2.579195737838745 + }, + { + "auxiliary_loss_clip": 0.01201009, + "auxiliary_loss_mlp": 0.01025877, + "balance_loss_clip": 1.04588354, + "balance_loss_mlp": 1.01773489, + "epoch": 0.8281127878314195, + "flos": 20631434860800.0, + "grad_norm": 2.485592463509054, + "language_loss": 0.85565448, + "learning_rate": 3.018613541427613e-07, + "loss": 0.87792325, + "num_input_tokens_seen": 148743690, + "step": 6887, + "time_per_iteration": 2.6960692405700684 + }, + { + "auxiliary_loss_clip": 0.01166099, + "auxiliary_loss_mlp": 0.01027834, + "balance_loss_clip": 1.04460025, + "balance_loss_mlp": 1.02075315, + "epoch": 0.8282330307220586, + "flos": 18004282122240.0, + "grad_norm": 1.6152483072945694, + "language_loss": 0.73451054, + "learning_rate": 3.0144996668905243e-07, + "loss": 0.75644994, + "num_input_tokens_seen": 148761070, + "step": 6888, + "time_per_iteration": 2.6534008979797363 + }, + { + "auxiliary_loss_clip": 0.01210648, + "auxiliary_loss_mlp": 0.01378733, + "balance_loss_clip": 1.03777802, + "balance_loss_mlp": 1.00020266, + "epoch": 0.8283532736126976, + "flos": 20084120352000.0, + "grad_norm": 1.8850587926143185, + "language_loss": 0.82475424, + "learning_rate": 3.010388369040331e-07, + "loss": 0.85064805, + "num_input_tokens_seen": 148779730, + "step": 6889, + "time_per_iteration": 2.7285306453704834 + }, + { + "auxiliary_loss_clip": 0.01178949, + "auxiliary_loss_mlp": 0.01027155, + "balance_loss_clip": 1.04746735, + "balance_loss_mlp": 1.01982951, + "epoch": 0.8284735165033368, + "flos": 31868421805440.0, + "grad_norm": 1.665385625881463, + "language_loss": 0.83032316, + "learning_rate": 3.0062796485007156e-07, + "loss": 0.85238421, + "num_input_tokens_seen": 148800670, + "step": 6890, + "time_per_iteration": 2.7322092056274414 + }, + { + "auxiliary_loss_clip": 0.01168026, + "auxiliary_loss_mlp": 0.01378927, + "balance_loss_clip": 1.04568696, + "balance_loss_mlp": 1.00019884, + "epoch": 0.8285937593939758, + "flos": 26651319840000.0, + "grad_norm": 2.3779970494854625, + "language_loss": 0.65590066, + "learning_rate": 3.002173505894965e-07, + "loss": 0.68137014, + "num_input_tokens_seen": 148819820, + "step": 6891, + "time_per_iteration": 2.622018337249756 + }, + { + "auxiliary_loss_clip": 0.01178822, + "auxiliary_loss_mlp": 0.01028726, + "balance_loss_clip": 1.04424787, + "balance_loss_mlp": 1.02056658, + "epoch": 0.8287140022846149, + "flos": 20193683811840.0, + "grad_norm": 2.9148508421697414, + "language_loss": 0.62712812, + "learning_rate": 2.998069941845973e-07, + "loss": 0.6492036, + "num_input_tokens_seen": 148838890, + "step": 6892, + "time_per_iteration": 3.5314855575561523 + }, + { + "auxiliary_loss_clip": 0.01060801, + "auxiliary_loss_mlp": 0.01002905, + "balance_loss_clip": 1.00942934, + "balance_loss_mlp": 1.00183773, + "epoch": 0.8288342451752541, + "flos": 70755980019840.0, + "grad_norm": 0.7061893796176267, + "language_loss": 0.57438266, + "learning_rate": 2.993968956976258e-07, + "loss": 0.5950197, + "num_input_tokens_seen": 148906635, + "step": 6893, + "time_per_iteration": 3.2609429359436035 + }, + { + "auxiliary_loss_clip": 0.01174282, + "auxiliary_loss_mlp": 0.01028164, + "balance_loss_clip": 1.04833913, + "balance_loss_mlp": 1.0200932, + "epoch": 0.8289544880658931, + "flos": 24572235795840.0, + "grad_norm": 4.488401555038412, + "language_loss": 0.70076096, + "learning_rate": 2.9898705519079313e-07, + "loss": 0.72278535, + "num_input_tokens_seen": 148925740, + "step": 6894, + "time_per_iteration": 3.5120251178741455 + }, + { + "auxiliary_loss_clip": 0.01178362, + "auxiliary_loss_mlp": 0.01024047, + "balance_loss_clip": 1.04308939, + "balance_loss_mlp": 1.01671851, + "epoch": 0.8290747309565322, + "flos": 22273378387200.0, + "grad_norm": 1.665227310638405, + "language_loss": 0.74652553, + "learning_rate": 2.985774727262715e-07, + "loss": 0.76854962, + "num_input_tokens_seen": 148944585, + "step": 6895, + "time_per_iteration": 2.6334421634674072 + }, + { + "auxiliary_loss_clip": 0.01166323, + "auxiliary_loss_mlp": 0.01025662, + "balance_loss_clip": 1.04552794, + "balance_loss_mlp": 1.01869178, + "epoch": 0.8291949738471713, + "flos": 23255570856960.0, + "grad_norm": 1.646394548842829, + "language_loss": 0.81449842, + "learning_rate": 2.981681483661949e-07, + "loss": 0.83641827, + "num_input_tokens_seen": 148964170, + "step": 6896, + "time_per_iteration": 2.5668222904205322 + }, + { + "auxiliary_loss_clip": 0.01182881, + "auxiliary_loss_mlp": 0.01023252, + "balance_loss_clip": 1.05015409, + "balance_loss_mlp": 1.01568222, + "epoch": 0.8293152167378104, + "flos": 52555768185600.0, + "grad_norm": 1.5721382283084278, + "language_loss": 0.7097851, + "learning_rate": 2.9775908217265633e-07, + "loss": 0.73184639, + "num_input_tokens_seen": 148989405, + "step": 6897, + "time_per_iteration": 3.745338201522827 + }, + { + "auxiliary_loss_clip": 0.01122253, + "auxiliary_loss_mlp": 0.01001086, + "balance_loss_clip": 1.0093987, + "balance_loss_mlp": 0.99988824, + "epoch": 0.8294354596284494, + "flos": 63356156294400.0, + "grad_norm": 0.8542182553235683, + "language_loss": 0.5041697, + "learning_rate": 2.9735027420771253e-07, + "loss": 0.52540308, + "num_input_tokens_seen": 149049740, + "step": 6898, + "time_per_iteration": 3.3843648433685303 + }, + { + "auxiliary_loss_clip": 0.01183021, + "auxiliary_loss_mlp": 0.01022134, + "balance_loss_clip": 1.04868591, + "balance_loss_mlp": 1.01540422, + "epoch": 0.8295557025190886, + "flos": 24827021942400.0, + "grad_norm": 1.8671913769681294, + "language_loss": 0.71548712, + "learning_rate": 2.969417245333774e-07, + "loss": 0.73753858, + "num_input_tokens_seen": 149069120, + "step": 6899, + "time_per_iteration": 4.178526163101196 + }, + { + "auxiliary_loss_clip": 0.01189574, + "auxiliary_loss_mlp": 0.01021095, + "balance_loss_clip": 1.04370165, + "balance_loss_mlp": 1.01408851, + "epoch": 0.8296759454097277, + "flos": 25118580637440.0, + "grad_norm": 3.2085127581738724, + "language_loss": 0.78339672, + "learning_rate": 2.9653343321162915e-07, + "loss": 0.80550343, + "num_input_tokens_seen": 149088630, + "step": 6900, + "time_per_iteration": 2.691709041595459 + }, + { + "auxiliary_loss_clip": 0.01195264, + "auxiliary_loss_mlp": 0.01024225, + "balance_loss_clip": 1.04490435, + "balance_loss_mlp": 1.01611853, + "epoch": 0.8297961883003667, + "flos": 24132581326080.0, + "grad_norm": 1.962970047190695, + "language_loss": 0.64980531, + "learning_rate": 2.9612540030440446e-07, + "loss": 0.67200017, + "num_input_tokens_seen": 149109175, + "step": 6901, + "time_per_iteration": 2.7056055068969727 + }, + { + "auxiliary_loss_clip": 0.01086577, + "auxiliary_loss_mlp": 0.01000702, + "balance_loss_clip": 1.00934124, + "balance_loss_mlp": 0.99965906, + "epoch": 0.8299164311910058, + "flos": 67446561375360.0, + "grad_norm": 0.8671253988904755, + "language_loss": 0.64080286, + "learning_rate": 2.9571762587360206e-07, + "loss": 0.66167563, + "num_input_tokens_seen": 149165560, + "step": 6902, + "time_per_iteration": 3.133528470993042 + }, + { + "auxiliary_loss_clip": 0.01195597, + "auxiliary_loss_mlp": 0.01026237, + "balance_loss_clip": 1.03604984, + "balance_loss_mlp": 1.0188694, + "epoch": 0.8300366740816449, + "flos": 25228682801280.0, + "grad_norm": 1.6604532657425135, + "language_loss": 0.73935705, + "learning_rate": 2.953101099810806e-07, + "loss": 0.76157546, + "num_input_tokens_seen": 149185165, + "step": 6903, + "time_per_iteration": 2.7080941200256348 + }, + { + "auxiliary_loss_clip": 0.01171231, + "auxiliary_loss_mlp": 0.01021991, + "balance_loss_clip": 1.04622614, + "balance_loss_mlp": 1.01473403, + "epoch": 0.830156916972284, + "flos": 18041018757120.0, + "grad_norm": 2.053573435014479, + "language_loss": 0.82728809, + "learning_rate": 2.9490285268865965e-07, + "loss": 0.84922028, + "num_input_tokens_seen": 149202655, + "step": 6904, + "time_per_iteration": 2.6784980297088623 + }, + { + "auxiliary_loss_clip": 0.011837, + "auxiliary_loss_mlp": 0.01024707, + "balance_loss_clip": 1.04897082, + "balance_loss_mlp": 1.01667845, + "epoch": 0.830277159862923, + "flos": 26322485806080.0, + "grad_norm": 2.339964613138459, + "language_loss": 0.7954824, + "learning_rate": 2.9449585405812085e-07, + "loss": 0.81756651, + "num_input_tokens_seen": 149220035, + "step": 6905, + "time_per_iteration": 2.587489128112793 + }, + { + "auxiliary_loss_clip": 0.01200067, + "auxiliary_loss_mlp": 0.01026092, + "balance_loss_clip": 1.04489994, + "balance_loss_mlp": 1.01862359, + "epoch": 0.8303974027535622, + "flos": 19938861751680.0, + "grad_norm": 1.9000166248093133, + "language_loss": 0.73895699, + "learning_rate": 2.940891141512043e-07, + "loss": 0.76121855, + "num_input_tokens_seen": 149238055, + "step": 6906, + "time_per_iteration": 2.6284894943237305 + }, + { + "auxiliary_loss_clip": 0.01182614, + "auxiliary_loss_mlp": 0.01026424, + "balance_loss_clip": 1.04277408, + "balance_loss_mlp": 1.01893759, + "epoch": 0.8305176456442013, + "flos": 17165552572800.0, + "grad_norm": 2.082016004276685, + "language_loss": 0.72049117, + "learning_rate": 2.9368263302961385e-07, + "loss": 0.74258143, + "num_input_tokens_seen": 149256755, + "step": 6907, + "time_per_iteration": 2.586228370666504 + }, + { + "auxiliary_loss_clip": 0.01208241, + "auxiliary_loss_mlp": 0.01027678, + "balance_loss_clip": 1.03718591, + "balance_loss_mlp": 1.02050757, + "epoch": 0.8306378885348403, + "flos": 25627614226560.0, + "grad_norm": 1.7103817384316966, + "language_loss": 0.79866982, + "learning_rate": 2.9327641075501075e-07, + "loss": 0.82102901, + "num_input_tokens_seen": 149275745, + "step": 6908, + "time_per_iteration": 2.7999424934387207 + }, + { + "auxiliary_loss_clip": 0.01174362, + "auxiliary_loss_mlp": 0.01027243, + "balance_loss_clip": 1.04032052, + "balance_loss_mlp": 1.01931512, + "epoch": 0.8307581314254795, + "flos": 33947864985600.0, + "grad_norm": 2.325125728898319, + "language_loss": 0.66747761, + "learning_rate": 2.9287044738901866e-07, + "loss": 0.68949366, + "num_input_tokens_seen": 149293730, + "step": 6909, + "time_per_iteration": 2.7434558868408203 + }, + { + "auxiliary_loss_clip": 0.0117916, + "auxiliary_loss_mlp": 0.01378667, + "balance_loss_clip": 1.04614592, + "balance_loss_mlp": 1.0001936, + "epoch": 0.8308783743161186, + "flos": 17562724231680.0, + "grad_norm": 2.9768974665843477, + "language_loss": 0.90834129, + "learning_rate": 2.9246474299322274e-07, + "loss": 0.93391955, + "num_input_tokens_seen": 149309290, + "step": 6910, + "time_per_iteration": 2.6441333293914795 + }, + { + "auxiliary_loss_clip": 0.0109716, + "auxiliary_loss_mlp": 0.01003542, + "balance_loss_clip": 1.01009357, + "balance_loss_mlp": 1.00238609, + "epoch": 0.8309986172067576, + "flos": 69412885649280.0, + "grad_norm": 0.8882694239954934, + "language_loss": 0.63216597, + "learning_rate": 2.920592976291678e-07, + "loss": 0.65317297, + "num_input_tokens_seen": 149366620, + "step": 6911, + "time_per_iteration": 3.18229079246521 + }, + { + "auxiliary_loss_clip": 0.01173989, + "auxiliary_loss_mlp": 0.01025921, + "balance_loss_clip": 1.04591537, + "balance_loss_mlp": 1.01853275, + "epoch": 0.8311188600973968, + "flos": 22309755886080.0, + "grad_norm": 1.8658193903517848, + "language_loss": 0.80568045, + "learning_rate": 2.916541113583595e-07, + "loss": 0.82767951, + "num_input_tokens_seen": 149385120, + "step": 6912, + "time_per_iteration": 2.60840106010437 + }, + { + "auxiliary_loss_clip": 0.01202023, + "auxiliary_loss_mlp": 0.01022809, + "balance_loss_clip": 1.04623079, + "balance_loss_mlp": 1.01565063, + "epoch": 0.8312391029880358, + "flos": 18770077105920.0, + "grad_norm": 2.6049761612753644, + "language_loss": 0.6614269, + "learning_rate": 2.912491842422642e-07, + "loss": 0.68367517, + "num_input_tokens_seen": 149402825, + "step": 6913, + "time_per_iteration": 2.6721935272216797 + }, + { + "auxiliary_loss_clip": 0.01177136, + "auxiliary_loss_mlp": 0.01024203, + "balance_loss_clip": 1.04646122, + "balance_loss_mlp": 1.01748586, + "epoch": 0.8313593458786749, + "flos": 20376648714240.0, + "grad_norm": 1.6320299300842072, + "language_loss": 0.71043628, + "learning_rate": 2.9084451634230857e-07, + "loss": 0.73244965, + "num_input_tokens_seen": 149422125, + "step": 6914, + "time_per_iteration": 2.6295714378356934 + }, + { + "auxiliary_loss_clip": 0.01191533, + "auxiliary_loss_mlp": 0.01023917, + "balance_loss_clip": 1.04265344, + "balance_loss_mlp": 1.01617122, + "epoch": 0.831479588769314, + "flos": 32124069878400.0, + "grad_norm": 3.7209750276400744, + "language_loss": 0.71331704, + "learning_rate": 2.9044010771988125e-07, + "loss": 0.73547155, + "num_input_tokens_seen": 149441940, + "step": 6915, + "time_per_iteration": 2.910135507583618 + }, + { + "auxiliary_loss_clip": 0.01175727, + "auxiliary_loss_mlp": 0.01027081, + "balance_loss_clip": 1.04221463, + "balance_loss_mlp": 1.01945758, + "epoch": 0.8315998316599531, + "flos": 45185929338240.0, + "grad_norm": 1.7373500314275234, + "language_loss": 0.71829396, + "learning_rate": 2.900359584363303e-07, + "loss": 0.74032199, + "num_input_tokens_seen": 149465045, + "step": 6916, + "time_per_iteration": 2.81333065032959 + }, + { + "auxiliary_loss_clip": 0.01200741, + "auxiliary_loss_mlp": 0.01034379, + "balance_loss_clip": 1.0449208, + "balance_loss_mlp": 1.02649379, + "epoch": 0.8317200745505922, + "flos": 18363747479040.0, + "grad_norm": 2.6438972378565344, + "language_loss": 0.8437103, + "learning_rate": 2.8963206855296494e-07, + "loss": 0.86606157, + "num_input_tokens_seen": 149481285, + "step": 6917, + "time_per_iteration": 2.701179265975952 + }, + { + "auxiliary_loss_clip": 0.01178918, + "auxiliary_loss_mlp": 0.01023054, + "balance_loss_clip": 1.0455029, + "balance_loss_mlp": 1.01595187, + "epoch": 0.8318403174412313, + "flos": 24206557386240.0, + "grad_norm": 1.7516744923082073, + "language_loss": 0.7677753, + "learning_rate": 2.892284381310548e-07, + "loss": 0.78979498, + "num_input_tokens_seen": 149502700, + "step": 6918, + "time_per_iteration": 3.5262908935546875 + }, + { + "auxiliary_loss_clip": 0.01181085, + "auxiliary_loss_mlp": 0.01019554, + "balance_loss_clip": 1.04310739, + "balance_loss_mlp": 1.01187706, + "epoch": 0.8319605603318704, + "flos": 22418780641920.0, + "grad_norm": 3.598014432396304, + "language_loss": 0.72304595, + "learning_rate": 2.888250672318302e-07, + "loss": 0.74505234, + "num_input_tokens_seen": 149520100, + "step": 6919, + "time_per_iteration": 3.5482192039489746 + }, + { + "auxiliary_loss_clip": 0.01169248, + "auxiliary_loss_mlp": 0.01028934, + "balance_loss_clip": 1.04819012, + "balance_loss_mlp": 1.02099419, + "epoch": 0.8320808032225094, + "flos": 37414501459200.0, + "grad_norm": 1.7197235509215016, + "language_loss": 0.68553543, + "learning_rate": 2.884219559164831e-07, + "loss": 0.70751721, + "num_input_tokens_seen": 149543245, + "step": 6920, + "time_per_iteration": 2.6813721656799316 + }, + { + "auxiliary_loss_clip": 0.01176092, + "auxiliary_loss_mlp": 0.01025483, + "balance_loss_clip": 1.04600286, + "balance_loss_mlp": 1.01789188, + "epoch": 0.8322010461131486, + "flos": 12787395638400.0, + "grad_norm": 2.699605975305065, + "language_loss": 0.81376749, + "learning_rate": 2.880191042461635e-07, + "loss": 0.83578324, + "num_input_tokens_seen": 149559185, + "step": 6921, + "time_per_iteration": 2.7025792598724365 + }, + { + "auxiliary_loss_clip": 0.01208592, + "auxiliary_loss_mlp": 0.01023843, + "balance_loss_clip": 1.04134595, + "balance_loss_mlp": 1.01696455, + "epoch": 0.8323212890037877, + "flos": 15815455050240.0, + "grad_norm": 1.7641058143551054, + "language_loss": 0.80547482, + "learning_rate": 2.876165122819849e-07, + "loss": 0.82779914, + "num_input_tokens_seen": 149577165, + "step": 6922, + "time_per_iteration": 2.8098697662353516 + }, + { + "auxiliary_loss_clip": 0.01168295, + "auxiliary_loss_mlp": 0.01023918, + "balance_loss_clip": 1.04878104, + "balance_loss_mlp": 1.01650012, + "epoch": 0.8324415318944267, + "flos": 21719276208000.0, + "grad_norm": 1.7484702794609057, + "language_loss": 0.79294133, + "learning_rate": 2.872141800850201e-07, + "loss": 0.81486344, + "num_input_tokens_seen": 149594340, + "step": 6923, + "time_per_iteration": 3.4586594104766846 + }, + { + "auxiliary_loss_clip": 0.01168202, + "auxiliary_loss_mlp": 0.01024549, + "balance_loss_clip": 1.04741907, + "balance_loss_mlp": 1.01662731, + "epoch": 0.8325617747850659, + "flos": 34198700636160.0, + "grad_norm": 1.9188281940012035, + "language_loss": 0.73473376, + "learning_rate": 2.868121077163024e-07, + "loss": 0.7566613, + "num_input_tokens_seen": 149613895, + "step": 6924, + "time_per_iteration": 2.7562918663024902 + }, + { + "auxiliary_loss_clip": 0.01177707, + "auxiliary_loss_mlp": 0.01022498, + "balance_loss_clip": 1.04434872, + "balance_loss_mlp": 1.01491642, + "epoch": 0.8326820176757049, + "flos": 18369457741440.0, + "grad_norm": 1.9966778690574103, + "language_loss": 0.72695768, + "learning_rate": 2.864102952368257e-07, + "loss": 0.74895978, + "num_input_tokens_seen": 149631820, + "step": 6925, + "time_per_iteration": 3.5453805923461914 + }, + { + "auxiliary_loss_clip": 0.01188214, + "auxiliary_loss_mlp": 0.01025892, + "balance_loss_clip": 1.03720784, + "balance_loss_mlp": 1.01817632, + "epoch": 0.832802260566344, + "flos": 35991325716480.0, + "grad_norm": 1.242888384609189, + "language_loss": 0.59206218, + "learning_rate": 2.860087427075444e-07, + "loss": 0.61420321, + "num_input_tokens_seen": 149656070, + "step": 6926, + "time_per_iteration": 2.8745954036712646 + }, + { + "auxiliary_loss_clip": 0.01179622, + "auxiliary_loss_mlp": 0.01027789, + "balance_loss_clip": 1.04466796, + "balance_loss_mlp": 1.02070498, + "epoch": 0.8329225034569832, + "flos": 14244434928000.0, + "grad_norm": 3.392930872311364, + "language_loss": 0.86458695, + "learning_rate": 2.856074501893744e-07, + "loss": 0.88666105, + "num_input_tokens_seen": 149671270, + "step": 6927, + "time_per_iteration": 2.703157663345337 + }, + { + "auxiliary_loss_clip": 0.01180241, + "auxiliary_loss_mlp": 0.01026343, + "balance_loss_clip": 1.0474509, + "balance_loss_mlp": 1.01904702, + "epoch": 0.8330427463476222, + "flos": 18077468083200.0, + "grad_norm": 1.6891502401705625, + "language_loss": 0.81495512, + "learning_rate": 2.8520641774319054e-07, + "loss": 0.83702099, + "num_input_tokens_seen": 149689360, + "step": 6928, + "time_per_iteration": 2.5668718814849854 + }, + { + "auxiliary_loss_clip": 0.01186207, + "auxiliary_loss_mlp": 0.01021828, + "balance_loss_clip": 1.04209399, + "balance_loss_mlp": 1.0138377, + "epoch": 0.8331629892382613, + "flos": 18040839189120.0, + "grad_norm": 2.6161127026247297, + "language_loss": 0.75957161, + "learning_rate": 2.848056454298309e-07, + "loss": 0.78165197, + "num_input_tokens_seen": 149706685, + "step": 6929, + "time_per_iteration": 2.6248111724853516 + }, + { + "auxiliary_loss_clip": 0.011853, + "auxiliary_loss_mlp": 0.01022682, + "balance_loss_clip": 1.04534483, + "balance_loss_mlp": 1.01528561, + "epoch": 0.8332832321289004, + "flos": 17457398576640.0, + "grad_norm": 2.728468348929177, + "language_loss": 0.65264511, + "learning_rate": 2.844051333100905e-07, + "loss": 0.67472494, + "num_input_tokens_seen": 149724230, + "step": 6930, + "time_per_iteration": 2.6229724884033203 + }, + { + "auxiliary_loss_clip": 0.0118836, + "auxiliary_loss_mlp": 0.01021644, + "balance_loss_clip": 1.04912901, + "balance_loss_mlp": 1.0151, + "epoch": 0.8334034750195395, + "flos": 15084852416640.0, + "grad_norm": 1.8126382894537756, + "language_loss": 0.8389045, + "learning_rate": 2.840048814447269e-07, + "loss": 0.86100459, + "num_input_tokens_seen": 149742395, + "step": 6931, + "time_per_iteration": 2.6134331226348877 + }, + { + "auxiliary_loss_clip": 0.01174425, + "auxiliary_loss_mlp": 0.01027877, + "balance_loss_clip": 1.04158664, + "balance_loss_mlp": 1.02005935, + "epoch": 0.8335237179101785, + "flos": 19427170556160.0, + "grad_norm": 2.2798944923051927, + "language_loss": 0.73728633, + "learning_rate": 2.836048898944587e-07, + "loss": 0.75930935, + "num_input_tokens_seen": 149760820, + "step": 6932, + "time_per_iteration": 2.6459667682647705 + }, + { + "auxiliary_loss_clip": 0.01180794, + "auxiliary_loss_mlp": 0.01024731, + "balance_loss_clip": 1.04153454, + "balance_loss_mlp": 1.01777458, + "epoch": 0.8336439608008177, + "flos": 21762046327680.0, + "grad_norm": 2.392058994420174, + "language_loss": 0.72290075, + "learning_rate": 2.832051587199642e-07, + "loss": 0.74495596, + "num_input_tokens_seen": 149778075, + "step": 6933, + "time_per_iteration": 2.6325974464416504 + }, + { + "auxiliary_loss_clip": 0.01074041, + "auxiliary_loss_mlp": 0.01000867, + "balance_loss_clip": 1.00880241, + "balance_loss_mlp": 0.99976426, + "epoch": 0.8337642036914568, + "flos": 59702783990400.0, + "grad_norm": 0.8045411684192689, + "language_loss": 0.57788742, + "learning_rate": 2.828056879818821e-07, + "loss": 0.59863651, + "num_input_tokens_seen": 149837150, + "step": 6934, + "time_per_iteration": 3.1450114250183105 + }, + { + "auxiliary_loss_clip": 0.01187799, + "auxiliary_loss_mlp": 0.01021922, + "balance_loss_clip": 1.03754961, + "balance_loss_mlp": 1.01528811, + "epoch": 0.8338844465820958, + "flos": 27162185022720.0, + "grad_norm": 1.8175949265227225, + "language_loss": 0.83343679, + "learning_rate": 2.824064777408117e-07, + "loss": 0.85553396, + "num_input_tokens_seen": 149856940, + "step": 6935, + "time_per_iteration": 2.7079203128814697 + }, + { + "auxiliary_loss_clip": 0.0117684, + "auxiliary_loss_mlp": 0.01025599, + "balance_loss_clip": 1.04756296, + "balance_loss_mlp": 1.01707554, + "epoch": 0.8340046894727349, + "flos": 30481264425600.0, + "grad_norm": 1.6578624644738784, + "language_loss": 0.75796103, + "learning_rate": 2.8200752805731263e-07, + "loss": 0.77998543, + "num_input_tokens_seen": 149879930, + "step": 6936, + "time_per_iteration": 2.693403720855713 + }, + { + "auxiliary_loss_clip": 0.0117737, + "auxiliary_loss_mlp": 0.01024806, + "balance_loss_clip": 1.04693079, + "balance_loss_mlp": 1.01716137, + "epoch": 0.834124932363374, + "flos": 27126166659840.0, + "grad_norm": 1.920936241952421, + "language_loss": 0.81221068, + "learning_rate": 2.8160883899190625e-07, + "loss": 0.83423245, + "num_input_tokens_seen": 149903200, + "step": 6937, + "time_per_iteration": 2.7562575340270996 + }, + { + "auxiliary_loss_clip": 0.01184809, + "auxiliary_loss_mlp": 0.01026225, + "balance_loss_clip": 1.0445267, + "balance_loss_mlp": 1.01828229, + "epoch": 0.8342451752540131, + "flos": 24569865498240.0, + "grad_norm": 2.0123997330603336, + "language_loss": 0.73423672, + "learning_rate": 2.8121041060507234e-07, + "loss": 0.75634712, + "num_input_tokens_seen": 149922230, + "step": 6938, + "time_per_iteration": 2.7542836666107178 + }, + { + "auxiliary_loss_clip": 0.01182228, + "auxiliary_loss_mlp": 0.01028272, + "balance_loss_clip": 1.04641867, + "balance_loss_mlp": 1.02058911, + "epoch": 0.8343654181446521, + "flos": 26615085995520.0, + "grad_norm": 1.7898351705806401, + "language_loss": 0.71323895, + "learning_rate": 2.808122429572528e-07, + "loss": 0.73534393, + "num_input_tokens_seen": 149942435, + "step": 6939, + "time_per_iteration": 2.669241189956665 + }, + { + "auxiliary_loss_clip": 0.01200877, + "auxiliary_loss_mlp": 0.0102638, + "balance_loss_clip": 1.04320443, + "balance_loss_mlp": 1.01924825, + "epoch": 0.8344856610352913, + "flos": 20777268078720.0, + "grad_norm": 2.7639710366952674, + "language_loss": 0.75945878, + "learning_rate": 2.804143361088489e-07, + "loss": 0.78173143, + "num_input_tokens_seen": 149961615, + "step": 6940, + "time_per_iteration": 2.693439483642578 + }, + { + "auxiliary_loss_clip": 0.01177777, + "auxiliary_loss_mlp": 0.01028144, + "balance_loss_clip": 1.04286289, + "balance_loss_mlp": 1.02060997, + "epoch": 0.8346059039259304, + "flos": 26095960684800.0, + "grad_norm": 2.4647365935489165, + "language_loss": 0.78062391, + "learning_rate": 2.8001669012022277e-07, + "loss": 0.80268312, + "num_input_tokens_seen": 149979585, + "step": 6941, + "time_per_iteration": 2.69897723197937 + }, + { + "auxiliary_loss_clip": 0.01176567, + "auxiliary_loss_mlp": 0.01025656, + "balance_loss_clip": 1.04896152, + "balance_loss_mlp": 1.01836085, + "epoch": 0.8347261468165694, + "flos": 29027708755200.0, + "grad_norm": 1.7043869043047628, + "language_loss": 0.68999445, + "learning_rate": 2.7961930505169795e-07, + "loss": 0.7120167, + "num_input_tokens_seen": 150003830, + "step": 6942, + "time_per_iteration": 2.7214033603668213 + }, + { + "auxiliary_loss_clip": 0.011808, + "auxiliary_loss_mlp": 0.01378979, + "balance_loss_clip": 1.04843402, + "balance_loss_mlp": 1.00018311, + "epoch": 0.8348463897072086, + "flos": 26396461866240.0, + "grad_norm": 1.846709155984776, + "language_loss": 0.7668519, + "learning_rate": 2.792221809635558e-07, + "loss": 0.79244965, + "num_input_tokens_seen": 150024460, + "step": 6943, + "time_per_iteration": 2.6753835678100586 + }, + { + "auxiliary_loss_clip": 0.01227744, + "auxiliary_loss_mlp": 0.0102097, + "balance_loss_clip": 1.04187727, + "balance_loss_mlp": 1.0137136, + "epoch": 0.8349666325978476, + "flos": 23367720096000.0, + "grad_norm": 2.4912278864250816, + "language_loss": 0.75163722, + "learning_rate": 2.788253179160411e-07, + "loss": 0.77412438, + "num_input_tokens_seen": 150045620, + "step": 6944, + "time_per_iteration": 3.7965590953826904 + }, + { + "auxiliary_loss_clip": 0.01186374, + "auxiliary_loss_mlp": 0.01028358, + "balance_loss_clip": 1.04538369, + "balance_loss_mlp": 1.02169752, + "epoch": 0.8350868754884867, + "flos": 12896528135040.0, + "grad_norm": 2.150626993081801, + "language_loss": 0.65044737, + "learning_rate": 2.7842871596935725e-07, + "loss": 0.67259473, + "num_input_tokens_seen": 150064135, + "step": 6945, + "time_per_iteration": 3.8992578983306885 + }, + { + "auxiliary_loss_clip": 0.011823, + "auxiliary_loss_mlp": 0.01027739, + "balance_loss_clip": 1.04607642, + "balance_loss_mlp": 1.01974654, + "epoch": 0.8352071183791259, + "flos": 26505522535680.0, + "grad_norm": 1.751833517084248, + "language_loss": 0.69189537, + "learning_rate": 2.780323751836682e-07, + "loss": 0.71399575, + "num_input_tokens_seen": 150085350, + "step": 6946, + "time_per_iteration": 2.653759002685547 + }, + { + "auxiliary_loss_clip": 0.01181799, + "auxiliary_loss_mlp": 0.01378487, + "balance_loss_clip": 1.04115725, + "balance_loss_mlp": 1.00015652, + "epoch": 0.8353273612697649, + "flos": 20668063754880.0, + "grad_norm": 1.540783380888081, + "language_loss": 0.78556043, + "learning_rate": 2.7763629561909876e-07, + "loss": 0.81116325, + "num_input_tokens_seen": 150106180, + "step": 6947, + "time_per_iteration": 2.6653549671173096 + }, + { + "auxiliary_loss_clip": 0.0116592, + "auxiliary_loss_mlp": 0.01026035, + "balance_loss_clip": 1.04629076, + "balance_loss_mlp": 1.01883435, + "epoch": 0.835447604160404, + "flos": 19754137082880.0, + "grad_norm": 10.335253144073139, + "language_loss": 0.7681464, + "learning_rate": 2.772404773357335e-07, + "loss": 0.79006588, + "num_input_tokens_seen": 150125585, + "step": 6948, + "time_per_iteration": 2.648625612258911 + }, + { + "auxiliary_loss_clip": 0.01184185, + "auxiliary_loss_mlp": 0.01023544, + "balance_loss_clip": 1.04186893, + "balance_loss_mlp": 1.01634407, + "epoch": 0.8355678470510431, + "flos": 23435842239360.0, + "grad_norm": 1.8456759721420875, + "language_loss": 0.78267348, + "learning_rate": 2.7684492039361853e-07, + "loss": 0.80475074, + "num_input_tokens_seen": 150144810, + "step": 6949, + "time_per_iteration": 3.6135144233703613 + }, + { + "auxiliary_loss_clip": 0.01172282, + "auxiliary_loss_mlp": 0.01025724, + "balance_loss_clip": 1.05133772, + "balance_loss_mlp": 1.01877105, + "epoch": 0.8356880899416822, + "flos": 21214588164480.0, + "grad_norm": 3.1979963673121614, + "language_loss": 0.83776706, + "learning_rate": 2.764496248527586e-07, + "loss": 0.85974711, + "num_input_tokens_seen": 150163785, + "step": 6950, + "time_per_iteration": 2.671732187271118 + }, + { + "auxiliary_loss_clip": 0.01202687, + "auxiliary_loss_mlp": 0.01029374, + "balance_loss_clip": 1.04300117, + "balance_loss_mlp": 1.02201879, + "epoch": 0.8358083328323213, + "flos": 28037543466240.0, + "grad_norm": 2.092637242396246, + "language_loss": 0.78612751, + "learning_rate": 2.760545907731211e-07, + "loss": 0.80844808, + "num_input_tokens_seen": 150184360, + "step": 6951, + "time_per_iteration": 2.789229154586792 + }, + { + "auxiliary_loss_clip": 0.01175773, + "auxiliary_loss_mlp": 0.01021672, + "balance_loss_clip": 1.04330587, + "balance_loss_mlp": 1.01424575, + "epoch": 0.8359285757229604, + "flos": 27783655159680.0, + "grad_norm": 1.7924023581003437, + "language_loss": 0.67957759, + "learning_rate": 2.75659818214631e-07, + "loss": 0.70155203, + "num_input_tokens_seen": 150205465, + "step": 6952, + "time_per_iteration": 3.5284221172332764 + }, + { + "auxiliary_loss_clip": 0.01188696, + "auxiliary_loss_mlp": 0.01028496, + "balance_loss_clip": 1.04498613, + "balance_loss_mlp": 1.02162385, + "epoch": 0.8360488186135995, + "flos": 21435115714560.0, + "grad_norm": 2.3057421280712966, + "language_loss": 0.7821368, + "learning_rate": 2.752653072371749e-07, + "loss": 0.80430871, + "num_input_tokens_seen": 150224900, + "step": 6953, + "time_per_iteration": 2.8793702125549316 + }, + { + "auxiliary_loss_clip": 0.01189508, + "auxiliary_loss_mlp": 0.01024581, + "balance_loss_clip": 1.04441619, + "balance_loss_mlp": 1.01801825, + "epoch": 0.8361690615042385, + "flos": 27632327160960.0, + "grad_norm": 1.7112220122146444, + "language_loss": 0.74869716, + "learning_rate": 2.7487105790060105e-07, + "loss": 0.77083802, + "num_input_tokens_seen": 150244310, + "step": 6954, + "time_per_iteration": 2.7228808403015137 + }, + { + "auxiliary_loss_clip": 0.01177955, + "auxiliary_loss_mlp": 0.01024264, + "balance_loss_clip": 1.04462743, + "balance_loss_mlp": 1.01741552, + "epoch": 0.8362893043948777, + "flos": 39202529598720.0, + "grad_norm": 1.854674708725108, + "language_loss": 0.69218922, + "learning_rate": 2.7447707026471587e-07, + "loss": 0.7142114, + "num_input_tokens_seen": 150267285, + "step": 6955, + "time_per_iteration": 2.7869839668273926 + }, + { + "auxiliary_loss_clip": 0.01192566, + "auxiliary_loss_mlp": 0.01021126, + "balance_loss_clip": 1.04120803, + "balance_loss_mlp": 1.01420259, + "epoch": 0.8364095472855168, + "flos": 24785329230720.0, + "grad_norm": 2.2302564010965407, + "language_loss": 0.80135339, + "learning_rate": 2.740833443892874e-07, + "loss": 0.82349032, + "num_input_tokens_seen": 150285455, + "step": 6956, + "time_per_iteration": 2.7111668586730957 + }, + { + "auxiliary_loss_clip": 0.01184855, + "auxiliary_loss_mlp": 0.01020681, + "balance_loss_clip": 1.04360604, + "balance_loss_mlp": 1.01334345, + "epoch": 0.8365297901761558, + "flos": 22743412784640.0, + "grad_norm": 1.8866187415628786, + "language_loss": 0.80004025, + "learning_rate": 2.7368988033404327e-07, + "loss": 0.82209563, + "num_input_tokens_seen": 150302970, + "step": 6957, + "time_per_iteration": 2.66463303565979 + }, + { + "auxiliary_loss_clip": 0.01194415, + "auxiliary_loss_mlp": 0.01027644, + "balance_loss_clip": 1.04248977, + "balance_loss_mlp": 1.02064693, + "epoch": 0.836650033066795, + "flos": 28396003242240.0, + "grad_norm": 1.4572565963441617, + "language_loss": 0.84420705, + "learning_rate": 2.732966781586712e-07, + "loss": 0.86642766, + "num_input_tokens_seen": 150322715, + "step": 6958, + "time_per_iteration": 2.751164436340332 + }, + { + "auxiliary_loss_clip": 0.01169423, + "auxiliary_loss_mlp": 0.01025573, + "balance_loss_clip": 1.04355812, + "balance_loss_mlp": 1.01804233, + "epoch": 0.836770275957434, + "flos": 22236857233920.0, + "grad_norm": 1.5755474781326202, + "language_loss": 0.6647433, + "learning_rate": 2.729037379228205e-07, + "loss": 0.68669331, + "num_input_tokens_seen": 150342900, + "step": 6959, + "time_per_iteration": 2.630934715270996 + }, + { + "auxiliary_loss_clip": 0.01186368, + "auxiliary_loss_mlp": 0.01023946, + "balance_loss_clip": 1.04725039, + "balance_loss_mlp": 1.01619697, + "epoch": 0.8368905188480731, + "flos": 22491930689280.0, + "grad_norm": 1.5869722625305738, + "language_loss": 0.80651116, + "learning_rate": 2.725110596860998e-07, + "loss": 0.82861435, + "num_input_tokens_seen": 150363580, + "step": 6960, + "time_per_iteration": 2.7457423210144043 + }, + { + "auxiliary_loss_clip": 0.01202573, + "auxiliary_loss_mlp": 0.01023513, + "balance_loss_clip": 1.04434729, + "balance_loss_mlp": 1.01626801, + "epoch": 0.8370107617387123, + "flos": 13370405287680.0, + "grad_norm": 2.1679447408548556, + "language_loss": 0.70072842, + "learning_rate": 2.7211864350807776e-07, + "loss": 0.7229892, + "num_input_tokens_seen": 150381780, + "step": 6961, + "time_per_iteration": 2.665884017944336 + }, + { + "auxiliary_loss_clip": 0.01168371, + "auxiliary_loss_mlp": 0.01026067, + "balance_loss_clip": 1.04792929, + "balance_loss_mlp": 1.01782084, + "epoch": 0.8371310046293513, + "flos": 25261289372160.0, + "grad_norm": 1.8416720619075055, + "language_loss": 0.73583251, + "learning_rate": 2.717264894482836e-07, + "loss": 0.75777686, + "num_input_tokens_seen": 150402120, + "step": 6962, + "time_per_iteration": 2.639270782470703 + }, + { + "auxiliary_loss_clip": 0.01184348, + "auxiliary_loss_mlp": 0.01028566, + "balance_loss_clip": 1.0498879, + "balance_loss_mlp": 1.02065063, + "epoch": 0.8372512475199904, + "flos": 19792705311360.0, + "grad_norm": 3.4990941033461267, + "language_loss": 0.81014043, + "learning_rate": 2.7133459756620646e-07, + "loss": 0.83226955, + "num_input_tokens_seen": 150419315, + "step": 6963, + "time_per_iteration": 2.5663492679595947 + }, + { + "auxiliary_loss_clip": 0.01170291, + "auxiliary_loss_mlp": 0.01023362, + "balance_loss_clip": 1.04448414, + "balance_loss_mlp": 1.01573849, + "epoch": 0.8373714904106295, + "flos": 19391224020480.0, + "grad_norm": 1.613283915612607, + "language_loss": 0.73854864, + "learning_rate": 2.7094296792129733e-07, + "loss": 0.76048511, + "num_input_tokens_seen": 150438915, + "step": 6964, + "time_per_iteration": 2.6290717124938965 + }, + { + "auxiliary_loss_clip": 0.01175888, + "auxiliary_loss_mlp": 0.01023438, + "balance_loss_clip": 1.04435587, + "balance_loss_mlp": 1.01640785, + "epoch": 0.8374917333012686, + "flos": 14975935401600.0, + "grad_norm": 2.054177559905409, + "language_loss": 0.75458717, + "learning_rate": 2.7055160057296424e-07, + "loss": 0.77658045, + "num_input_tokens_seen": 150456155, + "step": 6965, + "time_per_iteration": 2.574399709701538 + }, + { + "auxiliary_loss_clip": 0.0119366, + "auxiliary_loss_mlp": 0.01027526, + "balance_loss_clip": 1.04280722, + "balance_loss_mlp": 1.02047753, + "epoch": 0.8376119761919076, + "flos": 30331839847680.0, + "grad_norm": 1.61114141117464, + "language_loss": 0.72292036, + "learning_rate": 2.7016049558057896e-07, + "loss": 0.74513221, + "num_input_tokens_seen": 150478115, + "step": 6966, + "time_per_iteration": 2.7754411697387695 + }, + { + "auxiliary_loss_clip": 0.01176294, + "auxiliary_loss_mlp": 0.01024533, + "balance_loss_clip": 1.0463388, + "balance_loss_mlp": 1.01684976, + "epoch": 0.8377322190825467, + "flos": 29423336129280.0, + "grad_norm": 1.6325340480384816, + "language_loss": 0.70443785, + "learning_rate": 2.6976965300347074e-07, + "loss": 0.72644609, + "num_input_tokens_seen": 150500725, + "step": 6967, + "time_per_iteration": 2.714664936065674 + }, + { + "auxiliary_loss_clip": 0.01178334, + "auxiliary_loss_mlp": 0.01022391, + "balance_loss_clip": 1.04173589, + "balance_loss_mlp": 1.01569772, + "epoch": 0.8378524619731859, + "flos": 26687086807680.0, + "grad_norm": 2.409486607852809, + "language_loss": 0.69888949, + "learning_rate": 2.693790729009309e-07, + "loss": 0.72089672, + "num_input_tokens_seen": 150522335, + "step": 6968, + "time_per_iteration": 2.7437684535980225 + }, + { + "auxiliary_loss_clip": 0.01187391, + "auxiliary_loss_mlp": 0.01024138, + "balance_loss_clip": 1.04496634, + "balance_loss_mlp": 1.01728976, + "epoch": 0.8379727048638249, + "flos": 20703866636160.0, + "grad_norm": 2.425576174782614, + "language_loss": 0.88589156, + "learning_rate": 2.6898875533220946e-07, + "loss": 0.90800685, + "num_input_tokens_seen": 150541640, + "step": 6969, + "time_per_iteration": 2.630267858505249 + }, + { + "auxiliary_loss_clip": 0.0116282, + "auxiliary_loss_mlp": 0.01026061, + "balance_loss_clip": 1.04637408, + "balance_loss_mlp": 1.01947498, + "epoch": 0.838092947754464, + "flos": 20084084438400.0, + "grad_norm": 1.6812327260311877, + "language_loss": 0.82068825, + "learning_rate": 2.685987003565171e-07, + "loss": 0.84257704, + "num_input_tokens_seen": 150559680, + "step": 6970, + "time_per_iteration": 3.468018054962158 + }, + { + "auxiliary_loss_clip": 0.0118863, + "auxiliary_loss_mlp": 0.01024739, + "balance_loss_clip": 1.04642236, + "balance_loss_mlp": 1.01749134, + "epoch": 0.8382131906451031, + "flos": 18113270964480.0, + "grad_norm": 2.3115752878090694, + "language_loss": 0.75606525, + "learning_rate": 2.6820890803302566e-07, + "loss": 0.77819896, + "num_input_tokens_seen": 150575205, + "step": 6971, + "time_per_iteration": 3.686236619949341 + }, + { + "auxiliary_loss_clip": 0.0118658, + "auxiliary_loss_mlp": 0.0102792, + "balance_loss_clip": 1.04680645, + "balance_loss_mlp": 1.02094376, + "epoch": 0.8383334335357422, + "flos": 17092653920640.0, + "grad_norm": 2.4411000176743807, + "language_loss": 0.81571114, + "learning_rate": 2.6781937842086557e-07, + "loss": 0.83785617, + "num_input_tokens_seen": 150593995, + "step": 6972, + "time_per_iteration": 2.6264572143554688 + }, + { + "auxiliary_loss_clip": 0.01179649, + "auxiliary_loss_mlp": 0.01026645, + "balance_loss_clip": 1.04545569, + "balance_loss_mlp": 1.01879478, + "epoch": 0.8384536764263812, + "flos": 20704728562560.0, + "grad_norm": 1.8695046955317511, + "language_loss": 0.6721983, + "learning_rate": 2.6743011157912933e-07, + "loss": 0.69426125, + "num_input_tokens_seen": 150613715, + "step": 6973, + "time_per_iteration": 2.6335339546203613 + }, + { + "auxiliary_loss_clip": 0.011966, + "auxiliary_loss_mlp": 0.01023806, + "balance_loss_clip": 1.03790641, + "balance_loss_mlp": 1.01583362, + "epoch": 0.8385739193170204, + "flos": 28986842056320.0, + "grad_norm": 2.0009519018099113, + "language_loss": 0.65089262, + "learning_rate": 2.6704110756686725e-07, + "loss": 0.67309672, + "num_input_tokens_seen": 150634540, + "step": 6974, + "time_per_iteration": 2.7746825218200684 + }, + { + "auxiliary_loss_clip": 0.0117915, + "auxiliary_loss_mlp": 0.01378695, + "balance_loss_clip": 1.04223013, + "balance_loss_mlp": 1.00018632, + "epoch": 0.8386941622076595, + "flos": 23438068882560.0, + "grad_norm": 1.6658354247429985, + "language_loss": 0.83861649, + "learning_rate": 2.6665236644309085e-07, + "loss": 0.86419493, + "num_input_tokens_seen": 150654850, + "step": 6975, + "time_per_iteration": 3.6557085514068604 + }, + { + "auxiliary_loss_clip": 0.01174451, + "auxiliary_loss_mlp": 0.01027699, + "balance_loss_clip": 1.04379189, + "balance_loss_mlp": 1.02062702, + "epoch": 0.8388144050982985, + "flos": 23002724044800.0, + "grad_norm": 1.9990634652157493, + "language_loss": 0.79849297, + "learning_rate": 2.662638882667727e-07, + "loss": 0.8205145, + "num_input_tokens_seen": 150673790, + "step": 6976, + "time_per_iteration": 2.6167569160461426 + }, + { + "auxiliary_loss_clip": 0.01171024, + "auxiliary_loss_mlp": 0.01030317, + "balance_loss_clip": 1.0471344, + "balance_loss_mlp": 1.02228189, + "epoch": 0.8389346479889377, + "flos": 24280353878400.0, + "grad_norm": 2.0288264831879075, + "language_loss": 0.72937876, + "learning_rate": 2.658756730968443e-07, + "loss": 0.75139225, + "num_input_tokens_seen": 150692255, + "step": 6977, + "time_per_iteration": 2.5466606616973877 + }, + { + "auxiliary_loss_clip": 0.0119231, + "auxiliary_loss_mlp": 0.01024139, + "balance_loss_clip": 1.0482223, + "balance_loss_mlp": 1.01726377, + "epoch": 0.8390548908795767, + "flos": 21215019127680.0, + "grad_norm": 2.019652835103244, + "language_loss": 0.88360941, + "learning_rate": 2.654877209921975e-07, + "loss": 0.90577388, + "num_input_tokens_seen": 150709790, + "step": 6978, + "time_per_iteration": 3.621208667755127 + }, + { + "auxiliary_loss_clip": 0.01206296, + "auxiliary_loss_mlp": 0.0102532, + "balance_loss_clip": 1.04050362, + "balance_loss_mlp": 1.01721978, + "epoch": 0.8391751337702158, + "flos": 35627299332480.0, + "grad_norm": 3.1036861284554162, + "language_loss": 0.62693119, + "learning_rate": 2.651000320116843e-07, + "loss": 0.64924729, + "num_input_tokens_seen": 150730675, + "step": 6979, + "time_per_iteration": 2.9585678577423096 + }, + { + "auxiliary_loss_clip": 0.0118863, + "auxiliary_loss_mlp": 0.01378715, + "balance_loss_clip": 1.04101419, + "balance_loss_mlp": 1.00020885, + "epoch": 0.839295376660855, + "flos": 21325229032320.0, + "grad_norm": 1.7557445085467398, + "language_loss": 0.76271689, + "learning_rate": 2.647126062141163e-07, + "loss": 0.78839028, + "num_input_tokens_seen": 150749750, + "step": 6980, + "time_per_iteration": 2.676874876022339 + }, + { + "auxiliary_loss_clip": 0.01186144, + "auxiliary_loss_mlp": 0.01024143, + "balance_loss_clip": 1.04248321, + "balance_loss_mlp": 1.01635313, + "epoch": 0.839415619551494, + "flos": 18442535961600.0, + "grad_norm": 1.7460414539694302, + "language_loss": 0.84013152, + "learning_rate": 2.643254436582669e-07, + "loss": 0.86223441, + "num_input_tokens_seen": 150769240, + "step": 6981, + "time_per_iteration": 2.674426794052124 + }, + { + "auxiliary_loss_clip": 0.01206662, + "auxiliary_loss_mlp": 0.01025166, + "balance_loss_clip": 1.04320478, + "balance_loss_mlp": 1.01757824, + "epoch": 0.8395358624421331, + "flos": 23221958705280.0, + "grad_norm": 2.012333913919314, + "language_loss": 0.82279205, + "learning_rate": 2.6393854440286743e-07, + "loss": 0.8451103, + "num_input_tokens_seen": 150788410, + "step": 6982, + "time_per_iteration": 2.7669150829315186 + }, + { + "auxiliary_loss_clip": 0.01166822, + "auxiliary_loss_mlp": 0.01028022, + "balance_loss_clip": 1.04801106, + "balance_loss_mlp": 1.02095246, + "epoch": 0.8396561053327722, + "flos": 24381657210240.0, + "grad_norm": 2.3321812869227188, + "language_loss": 0.7072103, + "learning_rate": 2.6355190850661045e-07, + "loss": 0.72915876, + "num_input_tokens_seen": 150805245, + "step": 6983, + "time_per_iteration": 2.586242198944092 + }, + { + "auxiliary_loss_clip": 0.01184866, + "auxiliary_loss_mlp": 0.01021782, + "balance_loss_clip": 1.04645157, + "balance_loss_mlp": 1.01439738, + "epoch": 0.8397763482234113, + "flos": 22237755073920.0, + "grad_norm": 1.4521223980280331, + "language_loss": 0.86480236, + "learning_rate": 2.631655360281486e-07, + "loss": 0.88686883, + "num_input_tokens_seen": 150824920, + "step": 6984, + "time_per_iteration": 2.6822683811187744 + }, + { + "auxiliary_loss_clip": 0.01184379, + "auxiliary_loss_mlp": 0.01378935, + "balance_loss_clip": 1.04786038, + "balance_loss_mlp": 1.00016165, + "epoch": 0.8398965911140504, + "flos": 22163743100160.0, + "grad_norm": 2.1330161738345343, + "language_loss": 0.65934324, + "learning_rate": 2.6277942702609323e-07, + "loss": 0.68497634, + "num_input_tokens_seen": 150844400, + "step": 6985, + "time_per_iteration": 2.713224172592163 + }, + { + "auxiliary_loss_clip": 0.01197208, + "auxiliary_loss_mlp": 0.01026561, + "balance_loss_clip": 1.04482913, + "balance_loss_mlp": 1.01887846, + "epoch": 0.8400168340046895, + "flos": 21542775753600.0, + "grad_norm": 2.0242558863854487, + "language_loss": 0.88117278, + "learning_rate": 2.623935815590186e-07, + "loss": 0.90341043, + "num_input_tokens_seen": 150862780, + "step": 6986, + "time_per_iteration": 2.6612610816955566 + }, + { + "auxiliary_loss_clip": 0.01189947, + "auxiliary_loss_mlp": 0.01029345, + "balance_loss_clip": 1.04860544, + "balance_loss_mlp": 1.02182937, + "epoch": 0.8401370768953286, + "flos": 22491966602880.0, + "grad_norm": 2.4121009850914295, + "language_loss": 0.8099094, + "learning_rate": 2.6200799968545516e-07, + "loss": 0.8321023, + "num_input_tokens_seen": 150883075, + "step": 6987, + "time_per_iteration": 2.6701202392578125 + }, + { + "auxiliary_loss_clip": 0.010843, + "auxiliary_loss_mlp": 0.01001857, + "balance_loss_clip": 1.01178789, + "balance_loss_mlp": 1.00068915, + "epoch": 0.8402573197859676, + "flos": 59238890818560.0, + "grad_norm": 0.7971750444381227, + "language_loss": 0.56450117, + "learning_rate": 2.616226814638969e-07, + "loss": 0.58536267, + "num_input_tokens_seen": 150948180, + "step": 6988, + "time_per_iteration": 3.235410451889038 + }, + { + "auxiliary_loss_clip": 0.0118774, + "auxiliary_loss_mlp": 0.01026194, + "balance_loss_clip": 1.04569018, + "balance_loss_mlp": 1.01894021, + "epoch": 0.8403775626766068, + "flos": 22674608282880.0, + "grad_norm": 1.967577196576237, + "language_loss": 0.77185524, + "learning_rate": 2.612376269527954e-07, + "loss": 0.79399455, + "num_input_tokens_seen": 150967885, + "step": 6989, + "time_per_iteration": 2.701538562774658 + }, + { + "auxiliary_loss_clip": 0.01183218, + "auxiliary_loss_mlp": 0.01027234, + "balance_loss_clip": 1.04576874, + "balance_loss_mlp": 1.01952744, + "epoch": 0.8404978055672458, + "flos": 19609704495360.0, + "grad_norm": 1.9018177541859729, + "language_loss": 0.67712903, + "learning_rate": 2.608528362105635e-07, + "loss": 0.69923353, + "num_input_tokens_seen": 150987255, + "step": 6990, + "time_per_iteration": 2.669032573699951 + }, + { + "auxiliary_loss_clip": 0.011958, + "auxiliary_loss_mlp": 0.01026583, + "balance_loss_clip": 1.04111004, + "balance_loss_mlp": 1.01991606, + "epoch": 0.8406180484578849, + "flos": 27526929678720.0, + "grad_norm": 1.8177741234862173, + "language_loss": 0.7356869, + "learning_rate": 2.6046830929557374e-07, + "loss": 0.75791073, + "num_input_tokens_seen": 151006905, + "step": 6991, + "time_per_iteration": 2.6937270164489746 + }, + { + "auxiliary_loss_clip": 0.01189802, + "auxiliary_loss_mlp": 0.0102379, + "balance_loss_clip": 1.04436624, + "balance_loss_mlp": 1.01662207, + "epoch": 0.8407382913485241, + "flos": 22127473342080.0, + "grad_norm": 1.8936703574562286, + "language_loss": 0.84746027, + "learning_rate": 2.6008404626615776e-07, + "loss": 0.86959624, + "num_input_tokens_seen": 151025405, + "step": 6992, + "time_per_iteration": 2.6795594692230225 + }, + { + "auxiliary_loss_clip": 0.01180799, + "auxiliary_loss_mlp": 0.01025493, + "balance_loss_clip": 1.04785895, + "balance_loss_mlp": 1.01864719, + "epoch": 0.8408585342391631, + "flos": 13918473982080.0, + "grad_norm": 3.5515954838108508, + "language_loss": 0.74308479, + "learning_rate": 2.597000471806092e-07, + "loss": 0.76514769, + "num_input_tokens_seen": 151041970, + "step": 6993, + "time_per_iteration": 2.6334187984466553 + }, + { + "auxiliary_loss_clip": 0.01183402, + "auxiliary_loss_mlp": 0.01030954, + "balance_loss_clip": 1.04828918, + "balance_loss_mlp": 1.02286005, + "epoch": 0.8409787771298022, + "flos": 20187865808640.0, + "grad_norm": 2.2783776627196275, + "language_loss": 0.72728813, + "learning_rate": 2.593163120971793e-07, + "loss": 0.74943167, + "num_input_tokens_seen": 151060835, + "step": 6994, + "time_per_iteration": 2.663815498352051 + }, + { + "auxiliary_loss_clip": 0.01191555, + "auxiliary_loss_mlp": 0.0102437, + "balance_loss_clip": 1.03786361, + "balance_loss_mlp": 1.01736045, + "epoch": 0.8410990200204413, + "flos": 23142523777920.0, + "grad_norm": 1.8927665101861464, + "language_loss": 0.68959022, + "learning_rate": 2.5893284107408165e-07, + "loss": 0.71174943, + "num_input_tokens_seen": 151078205, + "step": 6995, + "time_per_iteration": 2.7393176555633545 + }, + { + "auxiliary_loss_clip": 0.01202207, + "auxiliary_loss_mlp": 0.01023422, + "balance_loss_clip": 1.0416714, + "balance_loss_mlp": 1.01551259, + "epoch": 0.8412192629110804, + "flos": 24027219757440.0, + "grad_norm": 1.9450423127361867, + "language_loss": 0.78318048, + "learning_rate": 2.5854963416948726e-07, + "loss": 0.80543679, + "num_input_tokens_seen": 151100470, + "step": 6996, + "time_per_iteration": 3.8072009086608887 + }, + { + "auxiliary_loss_clip": 0.01196966, + "auxiliary_loss_mlp": 0.01023748, + "balance_loss_clip": 1.03607821, + "balance_loss_mlp": 1.01632726, + "epoch": 0.8413395058017195, + "flos": 25591703604480.0, + "grad_norm": 1.7911617534155846, + "language_loss": 0.69751954, + "learning_rate": 2.5816669144152816e-07, + "loss": 0.71972668, + "num_input_tokens_seen": 151121650, + "step": 6997, + "time_per_iteration": 2.772547721862793 + }, + { + "auxiliary_loss_clip": 0.01060707, + "auxiliary_loss_mlp": 0.01002299, + "balance_loss_clip": 1.00942314, + "balance_loss_mlp": 1.00126827, + "epoch": 0.8414597486923585, + "flos": 63635396624640.0, + "grad_norm": 0.85243210809405, + "language_loss": 0.66335237, + "learning_rate": 2.5778401294829777e-07, + "loss": 0.68398243, + "num_input_tokens_seen": 151180390, + "step": 6998, + "time_per_iteration": 3.2006399631500244 + }, + { + "auxiliary_loss_clip": 0.01172585, + "auxiliary_loss_mlp": 0.01378484, + "balance_loss_clip": 1.04494667, + "balance_loss_mlp": 1.0001725, + "epoch": 0.8415799915829977, + "flos": 19098731571840.0, + "grad_norm": 1.8922417952484634, + "language_loss": 0.65173328, + "learning_rate": 2.574015987478473e-07, + "loss": 0.67724395, + "num_input_tokens_seen": 151198520, + "step": 6999, + "time_per_iteration": 2.5915369987487793 + }, + { + "auxiliary_loss_clip": 0.01193045, + "auxiliary_loss_mlp": 0.01023573, + "balance_loss_clip": 1.04543817, + "balance_loss_mlp": 1.01569307, + "epoch": 0.8417002344736367, + "flos": 19821612781440.0, + "grad_norm": 2.212099102069244, + "language_loss": 0.86835837, + "learning_rate": 2.570194488981887e-07, + "loss": 0.89052463, + "num_input_tokens_seen": 151215065, + "step": 7000, + "time_per_iteration": 2.5784637928009033 + }, + { + "auxiliary_loss_clip": 0.01060754, + "auxiliary_loss_mlp": 0.01002591, + "balance_loss_clip": 1.00946546, + "balance_loss_mlp": 1.00153565, + "epoch": 0.8418204773642758, + "flos": 62161516834560.0, + "grad_norm": 0.8442122769135412, + "language_loss": 0.60325325, + "learning_rate": 2.566375634572939e-07, + "loss": 0.6238867, + "num_input_tokens_seen": 151275705, + "step": 7001, + "time_per_iteration": 3.932056188583374 + }, + { + "auxiliary_loss_clip": 0.01197614, + "auxiliary_loss_mlp": 0.01028583, + "balance_loss_clip": 1.04157829, + "balance_loss_mlp": 1.02137113, + "epoch": 0.841940720254915, + "flos": 17092905315840.0, + "grad_norm": 6.110566904974112, + "language_loss": 0.76275545, + "learning_rate": 2.562559424830943e-07, + "loss": 0.78501737, + "num_input_tokens_seen": 151293665, + "step": 7002, + "time_per_iteration": 2.6920642852783203 + }, + { + "auxiliary_loss_clip": 0.01179444, + "auxiliary_loss_mlp": 0.01020941, + "balance_loss_clip": 1.04272044, + "balance_loss_mlp": 1.01321936, + "epoch": 0.842060963145554, + "flos": 16283586026880.0, + "grad_norm": 2.0935913870723692, + "language_loss": 0.70482564, + "learning_rate": 2.5587458603348256e-07, + "loss": 0.72682953, + "num_input_tokens_seen": 151310955, + "step": 7003, + "time_per_iteration": 2.6165051460266113 + }, + { + "auxiliary_loss_clip": 0.01187318, + "auxiliary_loss_mlp": 0.01028547, + "balance_loss_clip": 1.04161441, + "balance_loss_mlp": 1.02083373, + "epoch": 0.8421812060361931, + "flos": 21908238681600.0, + "grad_norm": 3.384536650548104, + "language_loss": 0.84200585, + "learning_rate": 2.554934941663085e-07, + "loss": 0.86416459, + "num_input_tokens_seen": 151328490, + "step": 7004, + "time_per_iteration": 3.635490655899048 + }, + { + "auxiliary_loss_clip": 0.01193247, + "auxiliary_loss_mlp": 0.01023452, + "balance_loss_clip": 1.04267895, + "balance_loss_mlp": 1.01558435, + "epoch": 0.8423014489268322, + "flos": 27777693502080.0, + "grad_norm": 2.2245329135729897, + "language_loss": 0.73287827, + "learning_rate": 2.5511266693938484e-07, + "loss": 0.75504529, + "num_input_tokens_seen": 151346950, + "step": 7005, + "time_per_iteration": 2.7007925510406494 + }, + { + "auxiliary_loss_clip": 0.01184233, + "auxiliary_loss_mlp": 0.01024659, + "balance_loss_clip": 1.04656386, + "balance_loss_mlp": 1.01688039, + "epoch": 0.8424216918174713, + "flos": 25117610970240.0, + "grad_norm": 1.6088726759794558, + "language_loss": 0.77761871, + "learning_rate": 2.547321044104822e-07, + "loss": 0.79970765, + "num_input_tokens_seen": 151368445, + "step": 7006, + "time_per_iteration": 2.6963112354278564 + }, + { + "auxiliary_loss_clip": 0.01173801, + "auxiliary_loss_mlp": 0.01027953, + "balance_loss_clip": 1.05033565, + "balance_loss_mlp": 1.02053201, + "epoch": 0.8425419347081103, + "flos": 24748448941440.0, + "grad_norm": 1.716323902495106, + "language_loss": 0.76865226, + "learning_rate": 2.5435180663733113e-07, + "loss": 0.7906698, + "num_input_tokens_seen": 151388745, + "step": 7007, + "time_per_iteration": 2.673840284347534 + }, + { + "auxiliary_loss_clip": 0.01206823, + "auxiliary_loss_mlp": 0.01024135, + "balance_loss_clip": 1.03930199, + "balance_loss_mlp": 1.01644337, + "epoch": 0.8426621775987495, + "flos": 24820916630400.0, + "grad_norm": 2.611717955304855, + "language_loss": 0.72042537, + "learning_rate": 2.539717736776241e-07, + "loss": 0.74273491, + "num_input_tokens_seen": 151404970, + "step": 7008, + "time_per_iteration": 2.869875192642212 + }, + { + "auxiliary_loss_clip": 0.01174639, + "auxiliary_loss_mlp": 0.01023158, + "balance_loss_clip": 1.04713511, + "balance_loss_mlp": 1.01583004, + "epoch": 0.8427824204893886, + "flos": 23550074467200.0, + "grad_norm": 1.7727618647494383, + "language_loss": 0.76364243, + "learning_rate": 2.535920055890097e-07, + "loss": 0.78562045, + "num_input_tokens_seen": 151426265, + "step": 7009, + "time_per_iteration": 2.6795923709869385 + }, + { + "auxiliary_loss_clip": 0.01196384, + "auxiliary_loss_mlp": 0.0102611, + "balance_loss_clip": 1.04075694, + "balance_loss_mlp": 1.01790833, + "epoch": 0.8429026633800276, + "flos": 16143858120960.0, + "grad_norm": 2.142241107328737, + "language_loss": 0.64548635, + "learning_rate": 2.5321250242910006e-07, + "loss": 0.66771126, + "num_input_tokens_seen": 151444180, + "step": 7010, + "time_per_iteration": 2.720486879348755 + }, + { + "auxiliary_loss_clip": 0.01168905, + "auxiliary_loss_mlp": 0.01032877, + "balance_loss_clip": 1.04939663, + "balance_loss_mlp": 1.02519989, + "epoch": 0.8430229062706668, + "flos": 22198540400640.0, + "grad_norm": 1.8202671695956036, + "language_loss": 0.86573207, + "learning_rate": 2.5283326425546493e-07, + "loss": 0.88774991, + "num_input_tokens_seen": 151463290, + "step": 7011, + "time_per_iteration": 2.578160285949707 + }, + { + "auxiliary_loss_clip": 0.01190956, + "auxiliary_loss_mlp": 0.01024311, + "balance_loss_clip": 1.04635024, + "balance_loss_mlp": 1.01748967, + "epoch": 0.8431431491613058, + "flos": 35330317683840.0, + "grad_norm": 2.40232543104635, + "language_loss": 0.69079232, + "learning_rate": 2.5245429112563443e-07, + "loss": 0.71294498, + "num_input_tokens_seen": 151483965, + "step": 7012, + "time_per_iteration": 2.7652063369750977 + }, + { + "auxiliary_loss_clip": 0.01176613, + "auxiliary_loss_mlp": 0.01027471, + "balance_loss_clip": 1.04708421, + "balance_loss_mlp": 1.01999664, + "epoch": 0.8432633920519449, + "flos": 25812374808960.0, + "grad_norm": 1.906066378071046, + "language_loss": 0.81856263, + "learning_rate": 2.5207558309709865e-07, + "loss": 0.84060347, + "num_input_tokens_seen": 151503700, + "step": 7013, + "time_per_iteration": 2.662374973297119 + }, + { + "auxiliary_loss_clip": 0.01101795, + "auxiliary_loss_mlp": 0.01373712, + "balance_loss_clip": 1.00891972, + "balance_loss_mlp": 0.99977535, + "epoch": 0.8433836349425841, + "flos": 64959531592320.0, + "grad_norm": 0.6579082955463176, + "language_loss": 0.56316084, + "learning_rate": 2.516971402273065e-07, + "loss": 0.5879159, + "num_input_tokens_seen": 151569765, + "step": 7014, + "time_per_iteration": 3.2437400817871094 + }, + { + "auxiliary_loss_clip": 0.01184177, + "auxiliary_loss_mlp": 0.01023026, + "balance_loss_clip": 1.04217625, + "balance_loss_mlp": 1.01614141, + "epoch": 0.8435038778332231, + "flos": 20229989483520.0, + "grad_norm": 2.19634932737461, + "language_loss": 0.67718327, + "learning_rate": 2.513189625736687e-07, + "loss": 0.69925523, + "num_input_tokens_seen": 151586660, + "step": 7015, + "time_per_iteration": 2.7147512435913086 + }, + { + "auxiliary_loss_clip": 0.01197885, + "auxiliary_loss_mlp": 0.01027766, + "balance_loss_clip": 1.04295921, + "balance_loss_mlp": 1.02019596, + "epoch": 0.8436241207238622, + "flos": 20992229020800.0, + "grad_norm": 2.414486238599894, + "language_loss": 0.71518111, + "learning_rate": 2.509410501935534e-07, + "loss": 0.73743761, + "num_input_tokens_seen": 151602295, + "step": 7016, + "time_per_iteration": 2.643181085586548 + }, + { + "auxiliary_loss_clip": 0.01187872, + "auxiliary_loss_mlp": 0.01022471, + "balance_loss_clip": 1.04434967, + "balance_loss_mlp": 1.01488626, + "epoch": 0.8437443636145013, + "flos": 14682257804160.0, + "grad_norm": 2.9385297940623, + "language_loss": 0.7554307, + "learning_rate": 2.5056340314429116e-07, + "loss": 0.77753407, + "num_input_tokens_seen": 151619760, + "step": 7017, + "time_per_iteration": 2.6350185871124268 + }, + { + "auxiliary_loss_clip": 0.01203509, + "auxiliary_loss_mlp": 0.01024656, + "balance_loss_clip": 1.03872466, + "balance_loss_mlp": 1.016922, + "epoch": 0.8438646065051404, + "flos": 21608814908160.0, + "grad_norm": 2.2797483251783377, + "language_loss": 0.80652249, + "learning_rate": 2.5018602148316904e-07, + "loss": 0.82880414, + "num_input_tokens_seen": 151635795, + "step": 7018, + "time_per_iteration": 2.6828248500823975 + }, + { + "auxiliary_loss_clip": 0.01169612, + "auxiliary_loss_mlp": 0.01026144, + "balance_loss_clip": 1.05164289, + "balance_loss_mlp": 1.0193615, + "epoch": 0.8439848493957794, + "flos": 23289937194240.0, + "grad_norm": 2.1389433763513717, + "language_loss": 0.804277, + "learning_rate": 2.498089052674359e-07, + "loss": 0.82623452, + "num_input_tokens_seen": 151653770, + "step": 7019, + "time_per_iteration": 2.6180505752563477 + }, + { + "auxiliary_loss_clip": 0.01176941, + "auxiliary_loss_mlp": 0.01029318, + "balance_loss_clip": 1.04709101, + "balance_loss_mlp": 1.02177525, + "epoch": 0.8441050922864186, + "flos": 19719339782400.0, + "grad_norm": 1.847444414038663, + "language_loss": 0.75464427, + "learning_rate": 2.494320545543007e-07, + "loss": 0.77670693, + "num_input_tokens_seen": 151673340, + "step": 7020, + "time_per_iteration": 2.5861854553222656 + }, + { + "auxiliary_loss_clip": 0.01170721, + "auxiliary_loss_mlp": 0.01026943, + "balance_loss_clip": 1.04662585, + "balance_loss_mlp": 1.01878309, + "epoch": 0.8442253351770577, + "flos": 21835268202240.0, + "grad_norm": 4.53619579488249, + "language_loss": 0.66860384, + "learning_rate": 2.490554694009308e-07, + "loss": 0.69058049, + "num_input_tokens_seen": 151694205, + "step": 7021, + "time_per_iteration": 2.624830722808838 + }, + { + "auxiliary_loss_clip": 0.01179296, + "auxiliary_loss_mlp": 0.01029506, + "balance_loss_clip": 1.04495192, + "balance_loss_mlp": 1.02267551, + "epoch": 0.8443455780676967, + "flos": 34346365447680.0, + "grad_norm": 1.535426423146779, + "language_loss": 0.78192568, + "learning_rate": 2.4867914986445426e-07, + "loss": 0.80401373, + "num_input_tokens_seen": 151716595, + "step": 7022, + "time_per_iteration": 4.535080194473267 + }, + { + "auxiliary_loss_clip": 0.01186148, + "auxiliary_loss_mlp": 0.01020693, + "balance_loss_clip": 1.04210329, + "balance_loss_mlp": 1.01382685, + "epoch": 0.8444658209583359, + "flos": 48214599281280.0, + "grad_norm": 2.115380131273959, + "language_loss": 0.71067649, + "learning_rate": 2.483030960019581e-07, + "loss": 0.73274493, + "num_input_tokens_seen": 151740525, + "step": 7023, + "time_per_iteration": 2.8734636306762695 + }, + { + "auxiliary_loss_clip": 0.01107752, + "auxiliary_loss_mlp": 0.01001032, + "balance_loss_clip": 1.00986409, + "balance_loss_mlp": 0.99995309, + "epoch": 0.8445860638489749, + "flos": 68484773105280.0, + "grad_norm": 0.7412955507372794, + "language_loss": 0.5547471, + "learning_rate": 2.479273078704891e-07, + "loss": 0.57583487, + "num_input_tokens_seen": 151793890, + "step": 7024, + "time_per_iteration": 3.3374950885772705 + }, + { + "auxiliary_loss_clip": 0.0110346, + "auxiliary_loss_mlp": 0.01002058, + "balance_loss_clip": 1.01057291, + "balance_loss_mlp": 1.00075293, + "epoch": 0.844706306739614, + "flos": 62833331882880.0, + "grad_norm": 0.7843080718754644, + "language_loss": 0.64684463, + "learning_rate": 2.475517855270552e-07, + "loss": 0.66789973, + "num_input_tokens_seen": 151853970, + "step": 7025, + "time_per_iteration": 3.80672025680542 + }, + { + "auxiliary_loss_clip": 0.01166472, + "auxiliary_loss_mlp": 0.01028777, + "balance_loss_clip": 1.04725385, + "balance_loss_mlp": 1.0215348, + "epoch": 0.8448265496302532, + "flos": 14976114969600.0, + "grad_norm": 1.822197727638569, + "language_loss": 0.73031437, + "learning_rate": 2.4717652902862143e-07, + "loss": 0.75226688, + "num_input_tokens_seen": 151872945, + "step": 7026, + "time_per_iteration": 2.5323221683502197 + }, + { + "auxiliary_loss_clip": 0.01190512, + "auxiliary_loss_mlp": 0.01024954, + "balance_loss_clip": 1.04506254, + "balance_loss_mlp": 1.01713431, + "epoch": 0.8449467925208922, + "flos": 23441265192960.0, + "grad_norm": 1.9401042881517747, + "language_loss": 0.81221133, + "learning_rate": 2.4680153843211495e-07, + "loss": 0.83436596, + "num_input_tokens_seen": 151892875, + "step": 7027, + "time_per_iteration": 3.7263236045837402 + }, + { + "auxiliary_loss_clip": 0.01185397, + "auxiliary_loss_mlp": 0.0102622, + "balance_loss_clip": 1.04679918, + "balance_loss_mlp": 1.01862335, + "epoch": 0.8450670354115313, + "flos": 22748045639040.0, + "grad_norm": 1.6788906808750963, + "language_loss": 0.72230399, + "learning_rate": 2.464268137944212e-07, + "loss": 0.74442011, + "num_input_tokens_seen": 151914170, + "step": 7028, + "time_per_iteration": 2.735118865966797 + }, + { + "auxiliary_loss_clip": 0.01187634, + "auxiliary_loss_mlp": 0.0103007, + "balance_loss_clip": 1.03886771, + "balance_loss_mlp": 1.02219605, + "epoch": 0.8451872783021703, + "flos": 29825571605760.0, + "grad_norm": 1.9762510134569795, + "language_loss": 0.78206348, + "learning_rate": 2.46052355172385e-07, + "loss": 0.80424052, + "num_input_tokens_seen": 151932210, + "step": 7029, + "time_per_iteration": 2.733365774154663 + }, + { + "auxiliary_loss_clip": 0.01169439, + "auxiliary_loss_mlp": 0.0102706, + "balance_loss_clip": 1.04865313, + "balance_loss_mlp": 1.01958597, + "epoch": 0.8453075211928095, + "flos": 21870029589120.0, + "grad_norm": 2.40243499115462, + "language_loss": 0.74648333, + "learning_rate": 2.456781626228128e-07, + "loss": 0.76844835, + "num_input_tokens_seen": 151951715, + "step": 7030, + "time_per_iteration": 3.507667303085327 + }, + { + "auxiliary_loss_clip": 0.01110807, + "auxiliary_loss_mlp": 0.01373649, + "balance_loss_clip": 1.00883567, + "balance_loss_mlp": 0.99973357, + "epoch": 0.8454277640834486, + "flos": 58751869288320.0, + "grad_norm": 0.9155304949602574, + "language_loss": 0.66331899, + "learning_rate": 2.453042362024675e-07, + "loss": 0.68816352, + "num_input_tokens_seen": 152004960, + "step": 7031, + "time_per_iteration": 3.289062976837158 + }, + { + "auxiliary_loss_clip": 0.01165687, + "auxiliary_loss_mlp": 0.01029476, + "balance_loss_clip": 1.04512024, + "balance_loss_mlp": 1.02217448, + "epoch": 0.8455480069740876, + "flos": 27090076469760.0, + "grad_norm": 1.4940787279424292, + "language_loss": 0.73364139, + "learning_rate": 2.449305759680751e-07, + "loss": 0.75559294, + "num_input_tokens_seen": 152026285, + "step": 7032, + "time_per_iteration": 2.6056556701660156 + }, + { + "auxiliary_loss_clip": 0.01191513, + "auxiliary_loss_mlp": 0.01022658, + "balance_loss_clip": 1.04426122, + "balance_loss_mlp": 1.01549983, + "epoch": 0.8456682498647268, + "flos": 27198670262400.0, + "grad_norm": 1.5921569446964896, + "language_loss": 0.75170839, + "learning_rate": 2.445571819763188e-07, + "loss": 0.77385008, + "num_input_tokens_seen": 152048585, + "step": 7033, + "time_per_iteration": 2.716874837875366 + }, + { + "auxiliary_loss_clip": 0.0116642, + "auxiliary_loss_mlp": 0.01023404, + "balance_loss_clip": 1.04703331, + "balance_loss_mlp": 1.01607561, + "epoch": 0.8457884927553658, + "flos": 20631901737600.0, + "grad_norm": 1.8886722913489704, + "language_loss": 0.58599174, + "learning_rate": 2.4418405428384227e-07, + "loss": 0.60788989, + "num_input_tokens_seen": 152068795, + "step": 7034, + "time_per_iteration": 2.544949769973755 + }, + { + "auxiliary_loss_clip": 0.01166794, + "auxiliary_loss_mlp": 0.01379052, + "balance_loss_clip": 1.04668248, + "balance_loss_mlp": 1.00014555, + "epoch": 0.8459087356460049, + "flos": 15299023259520.0, + "grad_norm": 1.6825371153150337, + "language_loss": 0.71696746, + "learning_rate": 2.4381119294724864e-07, + "loss": 0.74242598, + "num_input_tokens_seen": 152086240, + "step": 7035, + "time_per_iteration": 2.639983892440796 + }, + { + "auxiliary_loss_clip": 0.01167493, + "auxiliary_loss_mlp": 0.01026806, + "balance_loss_clip": 1.04655218, + "balance_loss_mlp": 1.01967716, + "epoch": 0.846028978536644, + "flos": 18843155326080.0, + "grad_norm": 2.1650477928426373, + "language_loss": 0.5355382, + "learning_rate": 2.434385980231004e-07, + "loss": 0.55748117, + "num_input_tokens_seen": 152105080, + "step": 7036, + "time_per_iteration": 2.5691113471984863 + }, + { + "auxiliary_loss_clip": 0.0117714, + "auxiliary_loss_mlp": 0.01024401, + "balance_loss_clip": 1.04664493, + "balance_loss_mlp": 1.01730561, + "epoch": 0.8461492214272831, + "flos": 52661740285440.0, + "grad_norm": 1.5960164636352512, + "language_loss": 0.65628129, + "learning_rate": 2.4306626956792043e-07, + "loss": 0.67829669, + "num_input_tokens_seen": 152130025, + "step": 7037, + "time_per_iteration": 2.949506998062134 + }, + { + "auxiliary_loss_clip": 0.01177111, + "auxiliary_loss_mlp": 0.01030613, + "balance_loss_clip": 1.04475355, + "balance_loss_mlp": 1.02323973, + "epoch": 0.8462694643179222, + "flos": 18588405093120.0, + "grad_norm": 1.6825936085052944, + "language_loss": 0.75582492, + "learning_rate": 2.4269420763819017e-07, + "loss": 0.77790213, + "num_input_tokens_seen": 152148070, + "step": 7038, + "time_per_iteration": 2.5992705821990967 + }, + { + "auxiliary_loss_clip": 0.01172882, + "auxiliary_loss_mlp": 0.01026547, + "balance_loss_clip": 1.04438663, + "balance_loss_mlp": 1.01974583, + "epoch": 0.8463897072085613, + "flos": 24387080163840.0, + "grad_norm": 3.031235098650313, + "language_loss": 0.83713293, + "learning_rate": 2.4232241229035223e-07, + "loss": 0.85912716, + "num_input_tokens_seen": 152165825, + "step": 7039, + "time_per_iteration": 2.7175228595733643 + }, + { + "auxiliary_loss_clip": 0.0107465, + "auxiliary_loss_mlp": 0.01000634, + "balance_loss_clip": 1.0098269, + "balance_loss_mlp": 0.99949521, + "epoch": 0.8465099500992004, + "flos": 68702140258560.0, + "grad_norm": 0.7904345853028937, + "language_loss": 0.56737083, + "learning_rate": 2.419508835808064e-07, + "loss": 0.58812362, + "num_input_tokens_seen": 152222380, + "step": 7040, + "time_per_iteration": 3.208669424057007 + }, + { + "auxiliary_loss_clip": 0.01185023, + "auxiliary_loss_mlp": 0.01024519, + "balance_loss_clip": 1.04470158, + "balance_loss_mlp": 1.01693988, + "epoch": 0.8466301929898394, + "flos": 13735724561280.0, + "grad_norm": 2.6306803456346195, + "language_loss": 0.63012147, + "learning_rate": 2.415796215659134e-07, + "loss": 0.65221685, + "num_input_tokens_seen": 152239085, + "step": 7041, + "time_per_iteration": 2.572829246520996 + }, + { + "auxiliary_loss_clip": 0.01192764, + "auxiliary_loss_mlp": 0.01028283, + "balance_loss_clip": 1.04018879, + "balance_loss_mlp": 1.02093649, + "epoch": 0.8467504358804786, + "flos": 19241260738560.0, + "grad_norm": 1.9753615759914238, + "language_loss": 0.77116573, + "learning_rate": 2.412086263019939e-07, + "loss": 0.79337621, + "num_input_tokens_seen": 152257110, + "step": 7042, + "time_per_iteration": 2.6696815490722656 + }, + { + "auxiliary_loss_clip": 0.01163469, + "auxiliary_loss_mlp": 0.01022519, + "balance_loss_clip": 1.04704309, + "balance_loss_mlp": 1.01564956, + "epoch": 0.8468706787711177, + "flos": 21324115710720.0, + "grad_norm": 1.6225801289086463, + "language_loss": 0.80061442, + "learning_rate": 2.408378978453276e-07, + "loss": 0.82247436, + "num_input_tokens_seen": 152277230, + "step": 7043, + "time_per_iteration": 2.5780715942382812 + }, + { + "auxiliary_loss_clip": 0.01074487, + "auxiliary_loss_mlp": 0.01002352, + "balance_loss_clip": 1.00942576, + "balance_loss_mlp": 1.00123739, + "epoch": 0.8469909216617567, + "flos": 64877439058560.0, + "grad_norm": 0.8111084117805586, + "language_loss": 0.63983548, + "learning_rate": 2.404674362521533e-07, + "loss": 0.66060388, + "num_input_tokens_seen": 152335725, + "step": 7044, + "time_per_iteration": 3.0881388187408447 + }, + { + "auxiliary_loss_clip": 0.0117335, + "auxiliary_loss_mlp": 0.01021624, + "balance_loss_clip": 1.04638147, + "balance_loss_mlp": 1.01455212, + "epoch": 0.8471111645523959, + "flos": 19280583152640.0, + "grad_norm": 2.575126342702655, + "language_loss": 0.74955684, + "learning_rate": 2.4009724157866997e-07, + "loss": 0.77150655, + "num_input_tokens_seen": 152352785, + "step": 7045, + "time_per_iteration": 2.573145627975464 + }, + { + "auxiliary_loss_clip": 0.01165788, + "auxiliary_loss_mlp": 0.01026153, + "balance_loss_clip": 1.04683423, + "balance_loss_mlp": 1.01945353, + "epoch": 0.8472314074430349, + "flos": 22015826893440.0, + "grad_norm": 2.656479974821828, + "language_loss": 0.76327819, + "learning_rate": 2.3972731388103564e-07, + "loss": 0.78519756, + "num_input_tokens_seen": 152371265, + "step": 7046, + "time_per_iteration": 2.575348377227783 + }, + { + "auxiliary_loss_clip": 0.01118753, + "auxiliary_loss_mlp": 0.01001544, + "balance_loss_clip": 1.00854731, + "balance_loss_mlp": 1.00038743, + "epoch": 0.847351650333674, + "flos": 57882580243200.0, + "grad_norm": 0.8049783265480821, + "language_loss": 0.62395555, + "learning_rate": 2.393576532153687e-07, + "loss": 0.64515853, + "num_input_tokens_seen": 152435050, + "step": 7047, + "time_per_iteration": 3.6716067790985107 + }, + { + "auxiliary_loss_clip": 0.01072419, + "auxiliary_loss_mlp": 0.0100167, + "balance_loss_clip": 1.00947165, + "balance_loss_mlp": 1.00056124, + "epoch": 0.8474718932243132, + "flos": 41284238313600.0, + "grad_norm": 0.9324146578953956, + "language_loss": 0.57804579, + "learning_rate": 2.389882596377453e-07, + "loss": 0.59878665, + "num_input_tokens_seen": 152489315, + "step": 7048, + "time_per_iteration": 5.4601476192474365 + }, + { + "auxiliary_loss_clip": 0.01165971, + "auxiliary_loss_mlp": 0.0102637, + "balance_loss_clip": 1.04590058, + "balance_loss_mlp": 1.01880598, + "epoch": 0.8475921361149522, + "flos": 38180906974080.0, + "grad_norm": 1.811981254549643, + "language_loss": 0.76199883, + "learning_rate": 2.386191332042031e-07, + "loss": 0.7839222, + "num_input_tokens_seen": 152511210, + "step": 7049, + "time_per_iteration": 2.706230640411377 + }, + { + "auxiliary_loss_clip": 0.01172107, + "auxiliary_loss_mlp": 0.01029403, + "balance_loss_clip": 1.0486021, + "balance_loss_mlp": 1.02156472, + "epoch": 0.8477123790055913, + "flos": 25375054723200.0, + "grad_norm": 1.8029175273874884, + "language_loss": 0.72635901, + "learning_rate": 2.3825027397073794e-07, + "loss": 0.7483741, + "num_input_tokens_seen": 152531685, + "step": 7050, + "time_per_iteration": 2.6450352668762207 + }, + { + "auxiliary_loss_clip": 0.01178347, + "auxiliary_loss_mlp": 0.01027712, + "balance_loss_clip": 1.05033302, + "balance_loss_mlp": 1.01954031, + "epoch": 0.8478326218962304, + "flos": 30225185389440.0, + "grad_norm": 1.9408064025115461, + "language_loss": 0.66906434, + "learning_rate": 2.3788168199330515e-07, + "loss": 0.69112492, + "num_input_tokens_seen": 152553245, + "step": 7051, + "time_per_iteration": 2.6512303352355957 + }, + { + "auxiliary_loss_clip": 0.01167082, + "auxiliary_loss_mlp": 0.01023565, + "balance_loss_clip": 1.03913939, + "balance_loss_mlp": 1.01677573, + "epoch": 0.8479528647868695, + "flos": 38213800853760.0, + "grad_norm": 1.8568332546004354, + "language_loss": 0.72674227, + "learning_rate": 2.3751335732782074e-07, + "loss": 0.7486487, + "num_input_tokens_seen": 152574505, + "step": 7052, + "time_per_iteration": 2.841672897338867 + }, + { + "auxiliary_loss_clip": 0.01178566, + "auxiliary_loss_mlp": 0.01027413, + "balance_loss_clip": 1.04980016, + "balance_loss_mlp": 1.0200932, + "epoch": 0.8480731076775085, + "flos": 20957790856320.0, + "grad_norm": 1.7588110952475406, + "language_loss": 0.79451036, + "learning_rate": 2.371453000301582e-07, + "loss": 0.81657016, + "num_input_tokens_seen": 152593190, + "step": 7053, + "time_per_iteration": 3.517390251159668 + }, + { + "auxiliary_loss_clip": 0.01192557, + "auxiliary_loss_mlp": 0.01028568, + "balance_loss_clip": 1.04348969, + "balance_loss_mlp": 1.02085495, + "epoch": 0.8481933505681477, + "flos": 32596510487040.0, + "grad_norm": 1.985953538720378, + "language_loss": 0.74191558, + "learning_rate": 2.3677751015615222e-07, + "loss": 0.76412678, + "num_input_tokens_seen": 152615265, + "step": 7054, + "time_per_iteration": 2.7640230655670166 + }, + { + "auxiliary_loss_clip": 0.01170513, + "auxiliary_loss_mlp": 0.01027522, + "balance_loss_clip": 1.04102421, + "balance_loss_mlp": 1.01986027, + "epoch": 0.8483135934587868, + "flos": 20741177888640.0, + "grad_norm": 2.6189318063887277, + "language_loss": 0.85435343, + "learning_rate": 2.3640998776159593e-07, + "loss": 0.87633383, + "num_input_tokens_seen": 152632770, + "step": 7055, + "time_per_iteration": 2.7030186653137207 + }, + { + "auxiliary_loss_clip": 0.01187901, + "auxiliary_loss_mlp": 0.01025886, + "balance_loss_clip": 1.04577065, + "balance_loss_mlp": 1.01896858, + "epoch": 0.8484338363494258, + "flos": 21653057485440.0, + "grad_norm": 1.725428733302246, + "language_loss": 0.81433225, + "learning_rate": 2.3604273290224253e-07, + "loss": 0.83647013, + "num_input_tokens_seen": 152653485, + "step": 7056, + "time_per_iteration": 3.6215198040008545 + }, + { + "auxiliary_loss_clip": 0.01189871, + "auxiliary_loss_mlp": 0.01026294, + "balance_loss_clip": 1.04715133, + "balance_loss_mlp": 1.01839399, + "epoch": 0.848554079240065, + "flos": 15013964926080.0, + "grad_norm": 1.8205863402186915, + "language_loss": 0.74659342, + "learning_rate": 2.356757456338039e-07, + "loss": 0.76875508, + "num_input_tokens_seen": 152670970, + "step": 7057, + "time_per_iteration": 2.7038474082946777 + }, + { + "auxiliary_loss_clip": 0.01089566, + "auxiliary_loss_mlp": 0.01002255, + "balance_loss_clip": 1.01700473, + "balance_loss_mlp": 1.00119388, + "epoch": 0.848674322130704, + "flos": 68060453742720.0, + "grad_norm": 0.7523407753838783, + "language_loss": 0.59070301, + "learning_rate": 2.3530902601195147e-07, + "loss": 0.61162126, + "num_input_tokens_seen": 152739460, + "step": 7058, + "time_per_iteration": 3.3138270378112793 + }, + { + "auxiliary_loss_clip": 0.01176932, + "auxiliary_loss_mlp": 0.01024255, + "balance_loss_clip": 1.04739475, + "balance_loss_mlp": 1.01597595, + "epoch": 0.8487945650213431, + "flos": 18475788977280.0, + "grad_norm": 3.9438700579114583, + "language_loss": 0.78546405, + "learning_rate": 2.34942574092317e-07, + "loss": 0.80747592, + "num_input_tokens_seen": 152754710, + "step": 7059, + "time_per_iteration": 2.6098203659057617 + }, + { + "auxiliary_loss_clip": 0.01181159, + "auxiliary_loss_mlp": 0.01023606, + "balance_loss_clip": 1.04642344, + "balance_loss_mlp": 1.01686192, + "epoch": 0.8489148079119821, + "flos": 23473189405440.0, + "grad_norm": 5.545137229990216, + "language_loss": 0.76794356, + "learning_rate": 2.3457638993049045e-07, + "loss": 0.78999114, + "num_input_tokens_seen": 152772700, + "step": 7060, + "time_per_iteration": 2.6561036109924316 + }, + { + "auxiliary_loss_clip": 0.01216711, + "auxiliary_loss_mlp": 0.01029565, + "balance_loss_clip": 1.04524899, + "balance_loss_mlp": 1.02123809, + "epoch": 0.8490350508026213, + "flos": 19937604775680.0, + "grad_norm": 2.0114488233966274, + "language_loss": 0.64232588, + "learning_rate": 2.3421047358202252e-07, + "loss": 0.66478872, + "num_input_tokens_seen": 152791550, + "step": 7061, + "time_per_iteration": 2.7233998775482178 + }, + { + "auxiliary_loss_clip": 0.01180275, + "auxiliary_loss_mlp": 0.0102818, + "balance_loss_clip": 1.04668057, + "balance_loss_mlp": 1.02108669, + "epoch": 0.8491552936932604, + "flos": 24279958828800.0, + "grad_norm": 2.3183562453720965, + "language_loss": 0.8328613, + "learning_rate": 2.3384482510242144e-07, + "loss": 0.85494578, + "num_input_tokens_seen": 152809410, + "step": 7062, + "time_per_iteration": 2.647160291671753 + }, + { + "auxiliary_loss_clip": 0.01168994, + "auxiliary_loss_mlp": 0.01026949, + "balance_loss_clip": 1.04568863, + "balance_loss_mlp": 1.01936102, + "epoch": 0.8492755365838994, + "flos": 22522526098560.0, + "grad_norm": 2.1090199878984555, + "language_loss": 0.775374, + "learning_rate": 2.3347944454715575e-07, + "loss": 0.79733342, + "num_input_tokens_seen": 152825800, + "step": 7063, + "time_per_iteration": 2.5842573642730713 + }, + { + "auxiliary_loss_clip": 0.0117038, + "auxiliary_loss_mlp": 0.01026517, + "balance_loss_clip": 1.04754972, + "balance_loss_mlp": 1.01903677, + "epoch": 0.8493957794745386, + "flos": 26980441182720.0, + "grad_norm": 2.067267506876012, + "language_loss": 0.67494655, + "learning_rate": 2.331143319716542e-07, + "loss": 0.69691557, + "num_input_tokens_seen": 152845330, + "step": 7064, + "time_per_iteration": 2.7201452255249023 + }, + { + "auxiliary_loss_clip": 0.01200444, + "auxiliary_loss_mlp": 0.01022677, + "balance_loss_clip": 1.04447901, + "balance_loss_mlp": 1.01500583, + "epoch": 0.8495160223651776, + "flos": 29861985018240.0, + "grad_norm": 2.1583679380545546, + "language_loss": 0.66170037, + "learning_rate": 2.3274948743130363e-07, + "loss": 0.68393159, + "num_input_tokens_seen": 152865165, + "step": 7065, + "time_per_iteration": 2.7227060794830322 + }, + { + "auxiliary_loss_clip": 0.01167062, + "auxiliary_loss_mlp": 0.01022448, + "balance_loss_clip": 1.04578972, + "balance_loss_mlp": 1.01534581, + "epoch": 0.8496362652558167, + "flos": 23075443128960.0, + "grad_norm": 1.6091396653482641, + "language_loss": 0.79238951, + "learning_rate": 2.3238491098145085e-07, + "loss": 0.81428468, + "num_input_tokens_seen": 152884695, + "step": 7066, + "time_per_iteration": 2.689392566680908 + }, + { + "auxiliary_loss_clip": 0.01176749, + "auxiliary_loss_mlp": 0.01024491, + "balance_loss_clip": 1.04586124, + "balance_loss_mlp": 1.01655734, + "epoch": 0.8497565081464559, + "flos": 14609107756800.0, + "grad_norm": 2.375074856686803, + "language_loss": 0.73014903, + "learning_rate": 2.3202060267740141e-07, + "loss": 0.75216144, + "num_input_tokens_seen": 152902220, + "step": 7067, + "time_per_iteration": 2.6238157749176025 + }, + { + "auxiliary_loss_clip": 0.01195817, + "auxiliary_loss_mlp": 0.01023153, + "balance_loss_clip": 1.03834879, + "balance_loss_mlp": 1.01591063, + "epoch": 0.8498767510370949, + "flos": 21136446126720.0, + "grad_norm": 2.1124079276169083, + "language_loss": 0.7711193, + "learning_rate": 2.3165656257442044e-07, + "loss": 0.79330897, + "num_input_tokens_seen": 152920740, + "step": 7068, + "time_per_iteration": 2.753187656402588 + }, + { + "auxiliary_loss_clip": 0.01173473, + "auxiliary_loss_mlp": 0.01022336, + "balance_loss_clip": 1.0461545, + "balance_loss_mlp": 1.01563692, + "epoch": 0.849996993927734, + "flos": 23654538195840.0, + "grad_norm": 1.794731198712494, + "language_loss": 0.90245414, + "learning_rate": 2.31292790727734e-07, + "loss": 0.92441225, + "num_input_tokens_seen": 152938305, + "step": 7069, + "time_per_iteration": 2.625326156616211 + }, + { + "auxiliary_loss_clip": 0.01165426, + "auxiliary_loss_mlp": 0.01026738, + "balance_loss_clip": 1.04609251, + "balance_loss_mlp": 1.01968408, + "epoch": 0.8501172368183731, + "flos": 20558069331840.0, + "grad_norm": 2.258666903410028, + "language_loss": 0.80129361, + "learning_rate": 2.3092928719252392e-07, + "loss": 0.82321525, + "num_input_tokens_seen": 152956705, + "step": 7070, + "time_per_iteration": 2.689514636993408 + }, + { + "auxiliary_loss_clip": 0.01172815, + "auxiliary_loss_mlp": 0.01022874, + "balance_loss_clip": 1.04466796, + "balance_loss_mlp": 1.01587987, + "epoch": 0.8502374797090122, + "flos": 22272624201600.0, + "grad_norm": 2.2916350938927077, + "language_loss": 0.78391802, + "learning_rate": 2.3056605202393475e-07, + "loss": 0.80587494, + "num_input_tokens_seen": 152974265, + "step": 7071, + "time_per_iteration": 2.60034441947937 + }, + { + "auxiliary_loss_clip": 0.01169274, + "auxiliary_loss_mlp": 0.01378905, + "balance_loss_clip": 1.04214752, + "balance_loss_mlp": 1.00013614, + "epoch": 0.8503577225996513, + "flos": 23659817495040.0, + "grad_norm": 1.9805254950018771, + "language_loss": 0.66783911, + "learning_rate": 2.3020308527706888e-07, + "loss": 0.69332093, + "num_input_tokens_seen": 152993680, + "step": 7072, + "time_per_iteration": 2.67438006401062 + }, + { + "auxiliary_loss_clip": 0.01189954, + "auxiliary_loss_mlp": 0.01024917, + "balance_loss_clip": 1.04427373, + "balance_loss_mlp": 1.01774979, + "epoch": 0.8504779654902904, + "flos": 26758513002240.0, + "grad_norm": 4.388238043720553, + "language_loss": 0.88753629, + "learning_rate": 2.2984038700698715e-07, + "loss": 0.90968496, + "num_input_tokens_seen": 153012990, + "step": 7073, + "time_per_iteration": 2.663034200668335 + }, + { + "auxiliary_loss_clip": 0.01173286, + "auxiliary_loss_mlp": 0.01025755, + "balance_loss_clip": 1.04590249, + "balance_loss_mlp": 1.01870418, + "epoch": 0.8505982083809295, + "flos": 26468247196800.0, + "grad_norm": 1.6420025619604774, + "language_loss": 0.78789985, + "learning_rate": 2.2947795726871222e-07, + "loss": 0.80989027, + "num_input_tokens_seen": 153034015, + "step": 7074, + "time_per_iteration": 4.565369606018066 + }, + { + "auxiliary_loss_clip": 0.01178137, + "auxiliary_loss_mlp": 0.01378663, + "balance_loss_clip": 1.05079341, + "balance_loss_mlp": 1.0001471, + "epoch": 0.8507184512715685, + "flos": 20303390926080.0, + "grad_norm": 1.727833459516524, + "language_loss": 0.85765696, + "learning_rate": 2.2911579611722253e-07, + "loss": 0.88322496, + "num_input_tokens_seen": 153053160, + "step": 7075, + "time_per_iteration": 2.7492356300354004 + }, + { + "auxiliary_loss_clip": 0.01181775, + "auxiliary_loss_mlp": 0.01027679, + "balance_loss_clip": 1.04383969, + "balance_loss_mlp": 1.02102697, + "epoch": 0.8508386941622077, + "flos": 19025186474880.0, + "grad_norm": 1.7596662087668948, + "language_loss": 0.87428045, + "learning_rate": 2.2875390360745905e-07, + "loss": 0.89637506, + "num_input_tokens_seen": 153072565, + "step": 7076, + "time_per_iteration": 2.888784885406494 + }, + { + "auxiliary_loss_clip": 0.01200617, + "auxiliary_loss_mlp": 0.01034547, + "balance_loss_clip": 1.04436266, + "balance_loss_mlp": 1.02678657, + "epoch": 0.8509589370528468, + "flos": 16433405654400.0, + "grad_norm": 1.9305850566477587, + "language_loss": 0.7759949, + "learning_rate": 2.2839227979432008e-07, + "loss": 0.79834652, + "num_input_tokens_seen": 153090215, + "step": 7077, + "time_per_iteration": 2.6786420345306396 + }, + { + "auxiliary_loss_clip": 0.01186577, + "auxiliary_loss_mlp": 0.01025114, + "balance_loss_clip": 1.04473054, + "balance_loss_mlp": 1.01755941, + "epoch": 0.8510791799434858, + "flos": 18259714713600.0, + "grad_norm": 2.0870518915557685, + "language_loss": 0.85083985, + "learning_rate": 2.2803092473266373e-07, + "loss": 0.87295681, + "num_input_tokens_seen": 153107740, + "step": 7078, + "time_per_iteration": 2.7659378051757812 + }, + { + "auxiliary_loss_clip": 0.01171503, + "auxiliary_loss_mlp": 0.01028722, + "balance_loss_clip": 1.04990077, + "balance_loss_mlp": 1.02072322, + "epoch": 0.851199422834125, + "flos": 23441372933760.0, + "grad_norm": 2.1957613963214153, + "language_loss": 0.86583543, + "learning_rate": 2.2766983847730724e-07, + "loss": 0.88783765, + "num_input_tokens_seen": 153127410, + "step": 7079, + "time_per_iteration": 3.5401265621185303 + }, + { + "auxiliary_loss_clip": 0.01203513, + "auxiliary_loss_mlp": 0.0102331, + "balance_loss_clip": 1.04273808, + "balance_loss_mlp": 1.01531076, + "epoch": 0.851319665724764, + "flos": 16289404030080.0, + "grad_norm": 3.726224109329892, + "language_loss": 0.66530502, + "learning_rate": 2.2730902108302663e-07, + "loss": 0.68757319, + "num_input_tokens_seen": 153144325, + "step": 7080, + "time_per_iteration": 2.6827566623687744 + }, + { + "auxiliary_loss_clip": 0.01175063, + "auxiliary_loss_mlp": 0.01025427, + "balance_loss_clip": 1.04238474, + "balance_loss_mlp": 1.01779127, + "epoch": 0.8514399086154031, + "flos": 18989347680000.0, + "grad_norm": 1.7377097480134704, + "language_loss": 0.68725473, + "learning_rate": 2.269484726045583e-07, + "loss": 0.70925963, + "num_input_tokens_seen": 153163240, + "step": 7081, + "time_per_iteration": 2.6621015071868896 + }, + { + "auxiliary_loss_clip": 0.01201481, + "auxiliary_loss_mlp": 0.01021241, + "balance_loss_clip": 1.04636741, + "balance_loss_mlp": 1.01438904, + "epoch": 0.8515601515060423, + "flos": 24571194301440.0, + "grad_norm": 1.5893197439241176, + "language_loss": 0.79264724, + "learning_rate": 2.2658819309659672e-07, + "loss": 0.81487447, + "num_input_tokens_seen": 153183440, + "step": 7082, + "time_per_iteration": 3.6198153495788574 + }, + { + "auxiliary_loss_clip": 0.01183071, + "auxiliary_loss_mlp": 0.01018352, + "balance_loss_clip": 1.04656315, + "balance_loss_mlp": 1.01193309, + "epoch": 0.8516803943966813, + "flos": 19529443555200.0, + "grad_norm": 2.7434865664545116, + "language_loss": 0.84952915, + "learning_rate": 2.2622818261379706e-07, + "loss": 0.87154341, + "num_input_tokens_seen": 153200460, + "step": 7083, + "time_per_iteration": 2.663966178894043 + }, + { + "auxiliary_loss_clip": 0.01180865, + "auxiliary_loss_mlp": 0.01025491, + "balance_loss_clip": 1.04283428, + "balance_loss_mlp": 1.01837087, + "epoch": 0.8518006372873204, + "flos": 20265792364800.0, + "grad_norm": 1.718393879087239, + "language_loss": 0.75073969, + "learning_rate": 2.2586844121077142e-07, + "loss": 0.77280319, + "num_input_tokens_seen": 153218970, + "step": 7084, + "time_per_iteration": 2.613914966583252 + }, + { + "auxiliary_loss_clip": 0.0120178, + "auxiliary_loss_mlp": 0.01022575, + "balance_loss_clip": 1.04008782, + "balance_loss_mlp": 1.01528549, + "epoch": 0.8519208801779595, + "flos": 24133227770880.0, + "grad_norm": 1.639957564534775, + "language_loss": 0.71877486, + "learning_rate": 2.2550896894209215e-07, + "loss": 0.74101847, + "num_input_tokens_seen": 153238485, + "step": 7085, + "time_per_iteration": 2.7337658405303955 + }, + { + "auxiliary_loss_clip": 0.0112331, + "auxiliary_loss_mlp": 0.01000628, + "balance_loss_clip": 1.0097847, + "balance_loss_mlp": 0.99937612, + "epoch": 0.8520411230685986, + "flos": 63035223252480.0, + "grad_norm": 0.6807112239517664, + "language_loss": 0.56609982, + "learning_rate": 2.2514976586229184e-07, + "loss": 0.58733916, + "num_input_tokens_seen": 153306430, + "step": 7086, + "time_per_iteration": 3.562818765640259 + }, + { + "auxiliary_loss_clip": 0.01074046, + "auxiliary_loss_mlp": 0.01001339, + "balance_loss_clip": 1.00939775, + "balance_loss_mlp": 1.00021291, + "epoch": 0.8521613659592376, + "flos": 65836865283840.0, + "grad_norm": 0.7492460442448982, + "language_loss": 0.54642314, + "learning_rate": 2.247908320258609e-07, + "loss": 0.567177, + "num_input_tokens_seen": 153366520, + "step": 7087, + "time_per_iteration": 3.3080649375915527 + }, + { + "auxiliary_loss_clip": 0.01197629, + "auxiliary_loss_mlp": 0.01028855, + "balance_loss_clip": 1.04286599, + "balance_loss_mlp": 1.02094591, + "epoch": 0.8522816088498768, + "flos": 23112323418240.0, + "grad_norm": 2.0991664194267226, + "language_loss": 0.79556811, + "learning_rate": 2.2443216748724914e-07, + "loss": 0.81783295, + "num_input_tokens_seen": 153387230, + "step": 7088, + "time_per_iteration": 2.7592086791992188 + }, + { + "auxiliary_loss_clip": 0.01182134, + "auxiliary_loss_mlp": 0.01378907, + "balance_loss_clip": 1.04784417, + "balance_loss_mlp": 1.00010121, + "epoch": 0.8524018517405159, + "flos": 31758140073600.0, + "grad_norm": 2.341376807450982, + "language_loss": 0.74246937, + "learning_rate": 2.2407377230086588e-07, + "loss": 0.76807976, + "num_input_tokens_seen": 153409585, + "step": 7089, + "time_per_iteration": 2.697312593460083 + }, + { + "auxiliary_loss_clip": 0.01190882, + "auxiliary_loss_mlp": 0.01032683, + "balance_loss_clip": 1.04440618, + "balance_loss_mlp": 1.02520216, + "epoch": 0.8525220946311549, + "flos": 18690318956160.0, + "grad_norm": 1.8686242113160094, + "language_loss": 0.83472699, + "learning_rate": 2.23715646521079e-07, + "loss": 0.85696268, + "num_input_tokens_seen": 153427105, + "step": 7090, + "time_per_iteration": 2.7150912284851074 + }, + { + "auxiliary_loss_clip": 0.01180315, + "auxiliary_loss_mlp": 0.0137912, + "balance_loss_clip": 1.04540646, + "balance_loss_mlp": 1.00010991, + "epoch": 0.852642337521794, + "flos": 21793216354560.0, + "grad_norm": 1.8401347052734673, + "language_loss": 0.84508562, + "learning_rate": 2.2335779020221724e-07, + "loss": 0.87068003, + "num_input_tokens_seen": 153443725, + "step": 7091, + "time_per_iteration": 2.5974068641662598 + }, + { + "auxiliary_loss_clip": 0.01074342, + "auxiliary_loss_mlp": 0.01001066, + "balance_loss_clip": 1.01496887, + "balance_loss_mlp": 1.00010061, + "epoch": 0.8527625804124331, + "flos": 69040132260480.0, + "grad_norm": 1.1125827964916415, + "language_loss": 0.56480068, + "learning_rate": 2.2300020339856497e-07, + "loss": 0.58555478, + "num_input_tokens_seen": 153506410, + "step": 7092, + "time_per_iteration": 3.3915176391601562 + }, + { + "auxiliary_loss_clip": 0.01182742, + "auxiliary_loss_mlp": 0.01022352, + "balance_loss_clip": 1.04394197, + "balance_loss_mlp": 1.01502609, + "epoch": 0.8528828233030722, + "flos": 26979399688320.0, + "grad_norm": 2.2823620829318996, + "language_loss": 0.77892751, + "learning_rate": 2.2264288616436966e-07, + "loss": 0.80097842, + "num_input_tokens_seen": 153526665, + "step": 7093, + "time_per_iteration": 2.7462124824523926 + }, + { + "auxiliary_loss_clip": 0.0118057, + "auxiliary_loss_mlp": 0.01027872, + "balance_loss_clip": 1.04388785, + "balance_loss_mlp": 1.01993322, + "epoch": 0.8530030661937112, + "flos": 17487598936320.0, + "grad_norm": 2.0084559703681926, + "language_loss": 0.72747439, + "learning_rate": 2.222858385538351e-07, + "loss": 0.74955881, + "num_input_tokens_seen": 153543465, + "step": 7094, + "time_per_iteration": 2.6940994262695312 + }, + { + "auxiliary_loss_clip": 0.01171988, + "auxiliary_loss_mlp": 0.01030775, + "balance_loss_clip": 1.04506254, + "balance_loss_mlp": 1.0239408, + "epoch": 0.8531233090843504, + "flos": 22160798184960.0, + "grad_norm": 2.424829084211818, + "language_loss": 0.68159425, + "learning_rate": 2.2192906062112527e-07, + "loss": 0.70362186, + "num_input_tokens_seen": 153563340, + "step": 7095, + "time_per_iteration": 2.6850638389587402 + }, + { + "auxiliary_loss_clip": 0.01168206, + "auxiliary_loss_mlp": 0.01026071, + "balance_loss_clip": 1.04675841, + "balance_loss_mlp": 1.01865649, + "epoch": 0.8532435519749895, + "flos": 37635388145280.0, + "grad_norm": 2.2966292894433997, + "language_loss": 0.70525765, + "learning_rate": 2.2157255242036377e-07, + "loss": 0.72720039, + "num_input_tokens_seen": 153587005, + "step": 7096, + "time_per_iteration": 2.7396697998046875 + }, + { + "auxiliary_loss_clip": 0.01189568, + "auxiliary_loss_mlp": 0.0102653, + "balance_loss_clip": 1.04377878, + "balance_loss_mlp": 1.01928532, + "epoch": 0.8533637948656285, + "flos": 21398163598080.0, + "grad_norm": 1.6252669384473406, + "language_loss": 0.74612355, + "learning_rate": 2.2121631400563135e-07, + "loss": 0.7682845, + "num_input_tokens_seen": 153606835, + "step": 7097, + "time_per_iteration": 2.701582193374634 + }, + { + "auxiliary_loss_clip": 0.01076118, + "auxiliary_loss_mlp": 0.0100136, + "balance_loss_clip": 1.01673913, + "balance_loss_mlp": 1.00029874, + "epoch": 0.8534840377562677, + "flos": 53345122490880.0, + "grad_norm": 0.7720794952733137, + "language_loss": 0.52947617, + "learning_rate": 2.208603454309701e-07, + "loss": 0.55025089, + "num_input_tokens_seen": 153664925, + "step": 7098, + "time_per_iteration": 3.1659090518951416 + }, + { + "auxiliary_loss_clip": 0.01201766, + "auxiliary_loss_mlp": 0.01028293, + "balance_loss_clip": 1.04187083, + "balance_loss_mlp": 1.02053249, + "epoch": 0.8536042806469067, + "flos": 20814148368000.0, + "grad_norm": 1.8047696033765863, + "language_loss": 0.71316433, + "learning_rate": 2.2050464675037994e-07, + "loss": 0.73546493, + "num_input_tokens_seen": 153683550, + "step": 7099, + "time_per_iteration": 2.763566493988037 + }, + { + "auxiliary_loss_clip": 0.01184578, + "auxiliary_loss_mlp": 0.01023058, + "balance_loss_clip": 1.04442489, + "balance_loss_mlp": 1.01556587, + "epoch": 0.8537245235375458, + "flos": 24681368292480.0, + "grad_norm": 2.426271440888175, + "language_loss": 0.7310735, + "learning_rate": 2.2014921801782016e-07, + "loss": 0.75314987, + "num_input_tokens_seen": 153703040, + "step": 7100, + "time_per_iteration": 4.774994850158691 + }, + { + "auxiliary_loss_clip": 0.0118417, + "auxiliary_loss_mlp": 0.01023301, + "balance_loss_clip": 1.04127574, + "balance_loss_mlp": 1.015818, + "epoch": 0.853844766428185, + "flos": 24384817607040.0, + "grad_norm": 2.5384511849347096, + "language_loss": 0.74126333, + "learning_rate": 2.1979405928720872e-07, + "loss": 0.76333803, + "num_input_tokens_seen": 153722695, + "step": 7101, + "time_per_iteration": 2.667436122894287 + }, + { + "auxiliary_loss_clip": 0.01189564, + "auxiliary_loss_mlp": 0.01031864, + "balance_loss_clip": 1.04502678, + "balance_loss_mlp": 1.02494955, + "epoch": 0.853965009318824, + "flos": 20955707867520.0, + "grad_norm": 1.6060212920153274, + "language_loss": 0.79238802, + "learning_rate": 2.1943917061242257e-07, + "loss": 0.81460232, + "num_input_tokens_seen": 153742550, + "step": 7102, + "time_per_iteration": 2.715902805328369 + }, + { + "auxiliary_loss_clip": 0.01183556, + "auxiliary_loss_mlp": 0.01378882, + "balance_loss_clip": 1.0463382, + "balance_loss_mlp": 1.00015438, + "epoch": 0.8540852522094631, + "flos": 24201816791040.0, + "grad_norm": 1.6555908481381723, + "language_loss": 0.66610694, + "learning_rate": 2.1908455204729903e-07, + "loss": 0.69173133, + "num_input_tokens_seen": 153761700, + "step": 7103, + "time_per_iteration": 2.772773027420044 + }, + { + "auxiliary_loss_clip": 0.01182904, + "auxiliary_loss_mlp": 0.01027352, + "balance_loss_clip": 1.04287601, + "balance_loss_mlp": 1.02021766, + "epoch": 0.8542054951001022, + "flos": 25082921410560.0, + "grad_norm": 2.7739083188641573, + "language_loss": 0.78420484, + "learning_rate": 2.1873020364563265e-07, + "loss": 0.80630744, + "num_input_tokens_seen": 153780765, + "step": 7104, + "time_per_iteration": 3.6080472469329834 + }, + { + "auxiliary_loss_clip": 0.01174043, + "auxiliary_loss_mlp": 0.01022089, + "balance_loss_clip": 1.04668641, + "balance_loss_mlp": 1.01448345, + "epoch": 0.8543257379907413, + "flos": 24316551809280.0, + "grad_norm": 2.3730281019705632, + "language_loss": 0.76110065, + "learning_rate": 2.183761254611789e-07, + "loss": 0.78306198, + "num_input_tokens_seen": 153801090, + "step": 7105, + "time_per_iteration": 2.6630208492279053 + }, + { + "auxiliary_loss_clip": 0.01175291, + "auxiliary_loss_mlp": 0.01024651, + "balance_loss_clip": 1.04695415, + "balance_loss_mlp": 1.01745105, + "epoch": 0.8544459808813804, + "flos": 55286630467200.0, + "grad_norm": 2.023728812119525, + "language_loss": 0.70162594, + "learning_rate": 2.1802231754764987e-07, + "loss": 0.72362536, + "num_input_tokens_seen": 153826530, + "step": 7106, + "time_per_iteration": 2.8877880573272705 + }, + { + "auxiliary_loss_clip": 0.01184589, + "auxiliary_loss_mlp": 0.01019594, + "balance_loss_clip": 1.04301643, + "balance_loss_mlp": 1.01233387, + "epoch": 0.8545662237720195, + "flos": 25776248705280.0, + "grad_norm": 2.1654229787203656, + "language_loss": 0.76279342, + "learning_rate": 2.17668779958718e-07, + "loss": 0.78483534, + "num_input_tokens_seen": 153849110, + "step": 7107, + "time_per_iteration": 2.661689043045044 + }, + { + "auxiliary_loss_clip": 0.01169238, + "auxiliary_loss_mlp": 0.01024006, + "balance_loss_clip": 1.04869211, + "balance_loss_mlp": 1.01700532, + "epoch": 0.8546864666626586, + "flos": 11108320427520.0, + "grad_norm": 2.4143600506090874, + "language_loss": 0.80456841, + "learning_rate": 2.1731551274801553e-07, + "loss": 0.82650089, + "num_input_tokens_seen": 153865550, + "step": 7108, + "time_per_iteration": 3.445105791091919 + }, + { + "auxiliary_loss_clip": 0.01187221, + "auxiliary_loss_mlp": 0.01022053, + "balance_loss_clip": 1.04571104, + "balance_loss_mlp": 1.01470995, + "epoch": 0.8548067095532976, + "flos": 25520169669120.0, + "grad_norm": 2.357848665113171, + "language_loss": 0.61585689, + "learning_rate": 2.169625159691324e-07, + "loss": 0.63794965, + "num_input_tokens_seen": 153885425, + "step": 7109, + "time_per_iteration": 2.6490166187286377 + }, + { + "auxiliary_loss_clip": 0.0120932, + "auxiliary_loss_mlp": 0.01024592, + "balance_loss_clip": 1.04194355, + "balance_loss_mlp": 1.01732075, + "epoch": 0.8549269524439368, + "flos": 24717853532160.0, + "grad_norm": 2.167988412900692, + "language_loss": 0.74224436, + "learning_rate": 2.1660978967561784e-07, + "loss": 0.76458347, + "num_input_tokens_seen": 153904760, + "step": 7110, + "time_per_iteration": 2.7770512104034424 + }, + { + "auxiliary_loss_clip": 0.01166339, + "auxiliary_loss_mlp": 0.01023961, + "balance_loss_clip": 1.04501653, + "balance_loss_mlp": 1.01645696, + "epoch": 0.8550471953345758, + "flos": 19825599191040.0, + "grad_norm": 2.287886138313883, + "language_loss": 0.7878359, + "learning_rate": 2.1625733392098035e-07, + "loss": 0.80973887, + "num_input_tokens_seen": 153920370, + "step": 7111, + "time_per_iteration": 2.58793568611145 + }, + { + "auxiliary_loss_clip": 0.01165456, + "auxiliary_loss_mlp": 0.01021976, + "balance_loss_clip": 1.04554486, + "balance_loss_mlp": 1.01502645, + "epoch": 0.8551674382252149, + "flos": 22820441500800.0, + "grad_norm": 1.9460917477766035, + "language_loss": 0.79608339, + "learning_rate": 2.159051487586867e-07, + "loss": 0.8179577, + "num_input_tokens_seen": 153940500, + "step": 7112, + "time_per_iteration": 2.6385529041290283 + }, + { + "auxiliary_loss_clip": 0.01187247, + "auxiliary_loss_mlp": 0.01025356, + "balance_loss_clip": 1.04531646, + "balance_loss_mlp": 1.0176549, + "epoch": 0.8552876811158541, + "flos": 20631255292800.0, + "grad_norm": 2.1751915099090073, + "language_loss": 0.72333956, + "learning_rate": 2.155532342421642e-07, + "loss": 0.74546564, + "num_input_tokens_seen": 153958500, + "step": 7113, + "time_per_iteration": 2.7068850994110107 + }, + { + "auxiliary_loss_clip": 0.01178668, + "auxiliary_loss_mlp": 0.01029901, + "balance_loss_clip": 1.04656255, + "balance_loss_mlp": 1.02240884, + "epoch": 0.8554079240064931, + "flos": 23112359331840.0, + "grad_norm": 1.8403504092206282, + "language_loss": 0.78155851, + "learning_rate": 2.1520159042479636e-07, + "loss": 0.80364418, + "num_input_tokens_seen": 153976790, + "step": 7114, + "time_per_iteration": 2.6477653980255127 + }, + { + "auxiliary_loss_clip": 0.01176483, + "auxiliary_loss_mlp": 0.01027737, + "balance_loss_clip": 1.0473932, + "balance_loss_mlp": 1.02044189, + "epoch": 0.8555281668971322, + "flos": 22128047959680.0, + "grad_norm": 2.0796658694340975, + "language_loss": 0.71083623, + "learning_rate": 2.148502173599287e-07, + "loss": 0.73287839, + "num_input_tokens_seen": 153994930, + "step": 7115, + "time_per_iteration": 2.6386325359344482 + }, + { + "auxiliary_loss_clip": 0.01178604, + "auxiliary_loss_mlp": 0.01020882, + "balance_loss_clip": 1.04380798, + "balance_loss_mlp": 1.0137651, + "epoch": 0.8556484097877713, + "flos": 31139040234240.0, + "grad_norm": 1.7333638127360975, + "language_loss": 0.6574055, + "learning_rate": 2.1449911510086372e-07, + "loss": 0.67940032, + "num_input_tokens_seen": 154014400, + "step": 7116, + "time_per_iteration": 2.750997304916382 + }, + { + "auxiliary_loss_clip": 0.01173334, + "auxiliary_loss_mlp": 0.0102483, + "balance_loss_clip": 1.04471838, + "balance_loss_mlp": 1.01779974, + "epoch": 0.8557686526784104, + "flos": 24316551809280.0, + "grad_norm": 2.092015430555334, + "language_loss": 0.76868153, + "learning_rate": 2.141482837008628e-07, + "loss": 0.79066318, + "num_input_tokens_seen": 154034940, + "step": 7117, + "time_per_iteration": 2.6195414066314697 + }, + { + "auxiliary_loss_clip": 0.01166512, + "auxiliary_loss_mlp": 0.01024115, + "balance_loss_clip": 1.04398799, + "balance_loss_mlp": 1.01704597, + "epoch": 0.8558888955690495, + "flos": 17712723427200.0, + "grad_norm": 2.3186169408934862, + "language_loss": 0.71990991, + "learning_rate": 2.1379772321314826e-07, + "loss": 0.74181616, + "num_input_tokens_seen": 154052985, + "step": 7118, + "time_per_iteration": 2.623929738998413 + }, + { + "auxiliary_loss_clip": 0.0120338, + "auxiliary_loss_mlp": 0.01025813, + "balance_loss_clip": 1.04180849, + "balance_loss_mlp": 1.01824045, + "epoch": 0.8560091384596886, + "flos": 19171702051200.0, + "grad_norm": 1.9566379634145779, + "language_loss": 0.81631351, + "learning_rate": 2.1344743369089802e-07, + "loss": 0.8386054, + "num_input_tokens_seen": 154068765, + "step": 7119, + "time_per_iteration": 2.7446062564849854 + }, + { + "auxiliary_loss_clip": 0.01183498, + "auxiliary_loss_mlp": 0.01026577, + "balance_loss_clip": 1.04363251, + "balance_loss_mlp": 1.01955247, + "epoch": 0.8561293813503277, + "flos": 23914855036800.0, + "grad_norm": 1.743847628748894, + "language_loss": 0.81959343, + "learning_rate": 2.130974151872522e-07, + "loss": 0.84169412, + "num_input_tokens_seen": 154089100, + "step": 7120, + "time_per_iteration": 2.7340519428253174 + }, + { + "auxiliary_loss_clip": 0.01194371, + "auxiliary_loss_mlp": 0.01026367, + "balance_loss_clip": 1.04333174, + "balance_loss_mlp": 1.01871967, + "epoch": 0.8562496242409667, + "flos": 22529206028160.0, + "grad_norm": 1.8074213474736256, + "language_loss": 0.78776777, + "learning_rate": 2.1274766775530773e-07, + "loss": 0.80997521, + "num_input_tokens_seen": 154108965, + "step": 7121, + "time_per_iteration": 2.7567012310028076 + }, + { + "auxiliary_loss_clip": 0.01168434, + "auxiliary_loss_mlp": 0.01024885, + "balance_loss_clip": 1.04515862, + "balance_loss_mlp": 1.01736343, + "epoch": 0.8563698671316058, + "flos": 14712745472640.0, + "grad_norm": 2.1437907905707436, + "language_loss": 0.7955097, + "learning_rate": 2.1239819144812077e-07, + "loss": 0.81744283, + "num_input_tokens_seen": 154123425, + "step": 7122, + "time_per_iteration": 2.590430736541748 + }, + { + "auxiliary_loss_clip": 0.0118518, + "auxiliary_loss_mlp": 0.01025757, + "balance_loss_clip": 1.04139137, + "balance_loss_mlp": 1.01799607, + "epoch": 0.856490110022245, + "flos": 39167768211840.0, + "grad_norm": 6.27711536254312, + "language_loss": 0.69982916, + "learning_rate": 2.1204898631870716e-07, + "loss": 0.72193855, + "num_input_tokens_seen": 154148315, + "step": 7123, + "time_per_iteration": 2.8663082122802734 + }, + { + "auxiliary_loss_clip": 0.01186085, + "auxiliary_loss_mlp": 0.01025983, + "balance_loss_clip": 1.04616845, + "balance_loss_mlp": 1.01877999, + "epoch": 0.856610352912884, + "flos": 29059345658880.0, + "grad_norm": 1.8487635647934075, + "language_loss": 0.75970584, + "learning_rate": 2.1170005242004006e-07, + "loss": 0.7818265, + "num_input_tokens_seen": 154169665, + "step": 7124, + "time_per_iteration": 2.783012866973877 + }, + { + "auxiliary_loss_clip": 0.01189421, + "auxiliary_loss_mlp": 0.0102175, + "balance_loss_clip": 1.04393971, + "balance_loss_mlp": 1.01495469, + "epoch": 0.8567305958035231, + "flos": 23878333883520.0, + "grad_norm": 2.259000435116814, + "language_loss": 0.78072196, + "learning_rate": 2.1135138980505384e-07, + "loss": 0.80283368, + "num_input_tokens_seen": 154190335, + "step": 7125, + "time_per_iteration": 2.715985059738159 + }, + { + "auxiliary_loss_clip": 0.01183917, + "auxiliary_loss_mlp": 0.01022785, + "balance_loss_clip": 1.04728568, + "balance_loss_mlp": 1.0150125, + "epoch": 0.8568508386941622, + "flos": 22200120599040.0, + "grad_norm": 1.7527196863849852, + "language_loss": 0.72734606, + "learning_rate": 2.110029985266395e-07, + "loss": 0.74941307, + "num_input_tokens_seen": 154210040, + "step": 7126, + "time_per_iteration": 4.434968948364258 + }, + { + "auxiliary_loss_clip": 0.01187447, + "auxiliary_loss_mlp": 0.01028666, + "balance_loss_clip": 1.04270172, + "balance_loss_mlp": 1.02134669, + "epoch": 0.8569710815848013, + "flos": 17307507121920.0, + "grad_norm": 1.8350389605661714, + "language_loss": 0.73721749, + "learning_rate": 2.1065487863764787e-07, + "loss": 0.75937867, + "num_input_tokens_seen": 154228385, + "step": 7127, + "time_per_iteration": 2.665342330932617 + }, + { + "auxiliary_loss_clip": 0.01189051, + "auxiliary_loss_mlp": 0.01028152, + "balance_loss_clip": 1.03654242, + "balance_loss_mlp": 1.02084112, + "epoch": 0.8570913244754403, + "flos": 23732285184000.0, + "grad_norm": 4.419608680470103, + "language_loss": 0.85767823, + "learning_rate": 2.1030703019088846e-07, + "loss": 0.87985027, + "num_input_tokens_seen": 154249015, + "step": 7128, + "time_per_iteration": 2.7430949211120605 + }, + { + "auxiliary_loss_clip": 0.01172779, + "auxiliary_loss_mlp": 0.01021469, + "balance_loss_clip": 1.04620147, + "balance_loss_mlp": 1.01447117, + "epoch": 0.8572115673660795, + "flos": 20048748433920.0, + "grad_norm": 1.8610287405712704, + "language_loss": 0.70783728, + "learning_rate": 2.099594532391291e-07, + "loss": 0.72977972, + "num_input_tokens_seen": 154267700, + "step": 7129, + "time_per_iteration": 2.6349527835845947 + }, + { + "auxiliary_loss_clip": 0.01162319, + "auxiliary_loss_mlp": 0.01028232, + "balance_loss_clip": 1.04164517, + "balance_loss_mlp": 1.02149391, + "epoch": 0.8573318102567186, + "flos": 27160389342720.0, + "grad_norm": 1.6206753612548004, + "language_loss": 0.79190129, + "learning_rate": 2.0961214783509806e-07, + "loss": 0.81380677, + "num_input_tokens_seen": 154290580, + "step": 7130, + "time_per_iteration": 3.607046127319336 + }, + { + "auxiliary_loss_clip": 0.01186807, + "auxiliary_loss_mlp": 0.01026083, + "balance_loss_clip": 1.04238594, + "balance_loss_mlp": 1.01927054, + "epoch": 0.8574520531473576, + "flos": 24936585402240.0, + "grad_norm": 3.431136111554211, + "language_loss": 0.74720955, + "learning_rate": 2.0926511403148051e-07, + "loss": 0.76933849, + "num_input_tokens_seen": 154309545, + "step": 7131, + "time_per_iteration": 2.658625602722168 + }, + { + "auxiliary_loss_clip": 0.01201568, + "auxiliary_loss_mlp": 0.01025954, + "balance_loss_clip": 1.04553795, + "balance_loss_mlp": 1.01855433, + "epoch": 0.8575722960379968, + "flos": 18771154513920.0, + "grad_norm": 1.8648783029336358, + "language_loss": 0.75660557, + "learning_rate": 2.0891835188092143e-07, + "loss": 0.77888083, + "num_input_tokens_seen": 154326545, + "step": 7132, + "time_per_iteration": 2.7109358310699463 + }, + { + "auxiliary_loss_clip": 0.0119979, + "auxiliary_loss_mlp": 0.01024309, + "balance_loss_clip": 1.04252577, + "balance_loss_mlp": 1.01706123, + "epoch": 0.8576925389286358, + "flos": 22200300167040.0, + "grad_norm": 1.812148775898313, + "language_loss": 0.81215858, + "learning_rate": 2.0857186143602434e-07, + "loss": 0.83439964, + "num_input_tokens_seen": 154345190, + "step": 7133, + "time_per_iteration": 2.656475782394409 + }, + { + "auxiliary_loss_clip": 0.01182364, + "auxiliary_loss_mlp": 0.01024957, + "balance_loss_clip": 1.04097176, + "balance_loss_mlp": 1.01721454, + "epoch": 0.8578127818192749, + "flos": 22894345733760.0, + "grad_norm": 1.7016122178430257, + "language_loss": 0.67391962, + "learning_rate": 2.0822564274935094e-07, + "loss": 0.69599283, + "num_input_tokens_seen": 154364615, + "step": 7134, + "time_per_iteration": 3.6721017360687256 + }, + { + "auxiliary_loss_clip": 0.01184973, + "auxiliary_loss_mlp": 0.01024614, + "balance_loss_clip": 1.04578924, + "balance_loss_mlp": 1.01708007, + "epoch": 0.8579330247099141, + "flos": 34824839541120.0, + "grad_norm": 1.8499089393749062, + "language_loss": 0.66926032, + "learning_rate": 2.078796958734239e-07, + "loss": 0.69135618, + "num_input_tokens_seen": 154387335, + "step": 7135, + "time_per_iteration": 2.8426337242126465 + }, + { + "auxiliary_loss_clip": 0.01175993, + "auxiliary_loss_mlp": 0.01025235, + "balance_loss_clip": 1.0460552, + "balance_loss_mlp": 1.0175879, + "epoch": 0.8580532676005531, + "flos": 19755681367680.0, + "grad_norm": 2.0851599051638705, + "language_loss": 0.74934149, + "learning_rate": 2.0753402086072124e-07, + "loss": 0.77135378, + "num_input_tokens_seen": 154405965, + "step": 7136, + "time_per_iteration": 2.5805420875549316 + }, + { + "auxiliary_loss_clip": 0.0123573, + "auxiliary_loss_mlp": 0.01034136, + "balance_loss_clip": 1.04229498, + "balance_loss_mlp": 1.02678108, + "epoch": 0.8581735104911922, + "flos": 22739318634240.0, + "grad_norm": 2.5884640531922862, + "language_loss": 0.75583231, + "learning_rate": 2.071886177636828e-07, + "loss": 0.77853096, + "num_input_tokens_seen": 154422750, + "step": 7137, + "time_per_iteration": 3.0280044078826904 + }, + { + "auxiliary_loss_clip": 0.01174296, + "auxiliary_loss_mlp": 0.01027706, + "balance_loss_clip": 1.04603708, + "balance_loss_mlp": 1.02091968, + "epoch": 0.8582937533818313, + "flos": 23149131880320.0, + "grad_norm": 2.174407486253911, + "language_loss": 0.82877946, + "learning_rate": 2.0684348663470575e-07, + "loss": 0.8507995, + "num_input_tokens_seen": 154442930, + "step": 7138, + "time_per_iteration": 3.036168336868286 + }, + { + "auxiliary_loss_clip": 0.01180039, + "auxiliary_loss_mlp": 0.01023145, + "balance_loss_clip": 1.04091108, + "balance_loss_mlp": 1.01501513, + "epoch": 0.8584139962724704, + "flos": 19498668577920.0, + "grad_norm": 1.8005745417251617, + "language_loss": 0.61642408, + "learning_rate": 2.0649862752614555e-07, + "loss": 0.63845599, + "num_input_tokens_seen": 154461640, + "step": 7139, + "time_per_iteration": 2.675964593887329 + }, + { + "auxiliary_loss_clip": 0.0108668, + "auxiliary_loss_mlp": 0.01001777, + "balance_loss_clip": 1.00987053, + "balance_loss_mlp": 1.00066209, + "epoch": 0.8585342391631094, + "flos": 71276577788160.0, + "grad_norm": 0.7500127560285578, + "language_loss": 0.57042646, + "learning_rate": 2.0615404049031838e-07, + "loss": 0.59131104, + "num_input_tokens_seen": 154518610, + "step": 7140, + "time_per_iteration": 3.2244555950164795 + }, + { + "auxiliary_loss_clip": 0.01176663, + "auxiliary_loss_mlp": 0.01029493, + "balance_loss_clip": 1.04732609, + "balance_loss_mlp": 1.02184319, + "epoch": 0.8586544820537486, + "flos": 10815432929280.0, + "grad_norm": 2.4675654385064663, + "language_loss": 0.7828151, + "learning_rate": 2.0580972557949616e-07, + "loss": 0.80487669, + "num_input_tokens_seen": 154533700, + "step": 7141, + "time_per_iteration": 2.5674803256988525 + }, + { + "auxiliary_loss_clip": 0.01073828, + "auxiliary_loss_mlp": 0.01001137, + "balance_loss_clip": 1.00931954, + "balance_loss_mlp": 1.00002277, + "epoch": 0.8587747249443877, + "flos": 64811184422400.0, + "grad_norm": 0.7933444307062252, + "language_loss": 0.54214656, + "learning_rate": 2.054656828459125e-07, + "loss": 0.56289619, + "num_input_tokens_seen": 154597810, + "step": 7142, + "time_per_iteration": 3.2164554595947266 + }, + { + "auxiliary_loss_clip": 0.01200025, + "auxiliary_loss_mlp": 0.01027469, + "balance_loss_clip": 1.04303527, + "balance_loss_mlp": 1.02025676, + "epoch": 0.8588949678350267, + "flos": 26834607964800.0, + "grad_norm": 1.7665978657880053, + "language_loss": 0.77218097, + "learning_rate": 2.051219123417578e-07, + "loss": 0.79445589, + "num_input_tokens_seen": 154617870, + "step": 7143, + "time_per_iteration": 2.7312700748443604 + }, + { + "auxiliary_loss_clip": 0.01169068, + "auxiliary_loss_mlp": 0.01030539, + "balance_loss_clip": 1.04687834, + "balance_loss_mlp": 1.02255821, + "epoch": 0.8590152107256659, + "flos": 26104256726400.0, + "grad_norm": 2.2935910087252114, + "language_loss": 0.60086554, + "learning_rate": 2.0477841411918196e-07, + "loss": 0.62286162, + "num_input_tokens_seen": 154637395, + "step": 7144, + "time_per_iteration": 2.648329734802246 + }, + { + "auxiliary_loss_clip": 0.01172713, + "auxiliary_loss_mlp": 0.01021486, + "balance_loss_clip": 1.04507279, + "balance_loss_mlp": 1.01439035, + "epoch": 0.859135453616305, + "flos": 26140885620480.0, + "grad_norm": 2.0868089297076864, + "language_loss": 0.74501348, + "learning_rate": 2.0443518823029326e-07, + "loss": 0.76695549, + "num_input_tokens_seen": 154657935, + "step": 7145, + "time_per_iteration": 2.644888401031494 + }, + { + "auxiliary_loss_clip": 0.01185329, + "auxiliary_loss_mlp": 0.01025139, + "balance_loss_clip": 1.04136491, + "balance_loss_mlp": 1.01782537, + "epoch": 0.859255696506944, + "flos": 12969319046400.0, + "grad_norm": 2.039309256556629, + "language_loss": 0.76620746, + "learning_rate": 2.0409223472715854e-07, + "loss": 0.78831208, + "num_input_tokens_seen": 154675080, + "step": 7146, + "time_per_iteration": 2.7939929962158203 + }, + { + "auxiliary_loss_clip": 0.01194647, + "auxiliary_loss_mlp": 0.01378332, + "balance_loss_clip": 1.04389501, + "balance_loss_mlp": 1.00017285, + "epoch": 0.8593759393975832, + "flos": 18475753063680.0, + "grad_norm": 1.8332772457582278, + "language_loss": 0.74852335, + "learning_rate": 2.0374955366180434e-07, + "loss": 0.77425313, + "num_input_tokens_seen": 154692720, + "step": 7147, + "time_per_iteration": 2.724923849105835 + }, + { + "auxiliary_loss_clip": 0.01195542, + "auxiliary_loss_mlp": 0.01026176, + "balance_loss_clip": 1.04069686, + "balance_loss_mlp": 1.0183233, + "epoch": 0.8594961822882222, + "flos": 22200156512640.0, + "grad_norm": 1.6897304159594784, + "language_loss": 0.7271266, + "learning_rate": 2.034071450862147e-07, + "loss": 0.74934387, + "num_input_tokens_seen": 154710190, + "step": 7148, + "time_per_iteration": 2.6866753101348877 + }, + { + "auxiliary_loss_clip": 0.01184787, + "auxiliary_loss_mlp": 0.0103282, + "balance_loss_clip": 1.04218888, + "balance_loss_mlp": 1.02477884, + "epoch": 0.8596164251788613, + "flos": 23294749616640.0, + "grad_norm": 1.7621099097842514, + "language_loss": 0.76755536, + "learning_rate": 2.030650090523327e-07, + "loss": 0.7897315, + "num_input_tokens_seen": 154729380, + "step": 7149, + "time_per_iteration": 2.765655517578125 + }, + { + "auxiliary_loss_clip": 0.01187108, + "auxiliary_loss_mlp": 0.01027215, + "balance_loss_clip": 1.04013753, + "balance_loss_mlp": 1.01963902, + "epoch": 0.8597366680695004, + "flos": 31649905416960.0, + "grad_norm": 2.0130873616353133, + "language_loss": 0.5914011, + "learning_rate": 2.0272314561205995e-07, + "loss": 0.61354434, + "num_input_tokens_seen": 154749775, + "step": 7150, + "time_per_iteration": 2.7943594455718994 + }, + { + "auxiliary_loss_clip": 0.0118306, + "auxiliary_loss_mlp": 0.01026675, + "balance_loss_clip": 1.03961182, + "balance_loss_mlp": 1.01956081, + "epoch": 0.8598569109601395, + "flos": 21287738211840.0, + "grad_norm": 2.191519758245832, + "language_loss": 0.72593439, + "learning_rate": 2.023815548172567e-07, + "loss": 0.7480318, + "num_input_tokens_seen": 154769845, + "step": 7151, + "time_per_iteration": 2.7386295795440674 + }, + { + "auxiliary_loss_clip": 0.01175719, + "auxiliary_loss_mlp": 0.01028625, + "balance_loss_clip": 1.04547489, + "balance_loss_mlp": 1.02179706, + "epoch": 0.8599771538507786, + "flos": 25447809720960.0, + "grad_norm": 1.5457589730743442, + "language_loss": 0.6604259, + "learning_rate": 2.0204023671974267e-07, + "loss": 0.68246937, + "num_input_tokens_seen": 154789230, + "step": 7152, + "time_per_iteration": 4.66881537437439 + }, + { + "auxiliary_loss_clip": 0.01171149, + "auxiliary_loss_mlp": 0.0102629, + "balance_loss_clip": 1.04397321, + "balance_loss_mlp": 1.01780868, + "epoch": 0.8600973967414177, + "flos": 16723958768640.0, + "grad_norm": 2.0260521817097557, + "language_loss": 0.81034279, + "learning_rate": 2.0169919137129532e-07, + "loss": 0.83231717, + "num_input_tokens_seen": 154807670, + "step": 7153, + "time_per_iteration": 2.5871827602386475 + }, + { + "auxiliary_loss_clip": 0.01179774, + "auxiliary_loss_mlp": 0.01026284, + "balance_loss_clip": 1.04822946, + "balance_loss_mlp": 1.01851141, + "epoch": 0.8602176396320568, + "flos": 25227928615680.0, + "grad_norm": 2.183605238494296, + "language_loss": 0.70695114, + "learning_rate": 2.013584188236508e-07, + "loss": 0.72901171, + "num_input_tokens_seen": 154825575, + "step": 7154, + "time_per_iteration": 2.6450397968292236 + }, + { + "auxiliary_loss_clip": 0.01171141, + "auxiliary_loss_mlp": 0.01022496, + "balance_loss_clip": 1.04812789, + "balance_loss_mlp": 1.01524258, + "epoch": 0.8603378825226958, + "flos": 20412236113920.0, + "grad_norm": 2.059095264479755, + "language_loss": 0.79590923, + "learning_rate": 2.0101791912850396e-07, + "loss": 0.81784564, + "num_input_tokens_seen": 154845115, + "step": 7155, + "time_per_iteration": 2.599410057067871 + }, + { + "auxiliary_loss_clip": 0.01186707, + "auxiliary_loss_mlp": 0.01024436, + "balance_loss_clip": 1.04553056, + "balance_loss_mlp": 1.01737022, + "epoch": 0.8604581254133349, + "flos": 34930201109760.0, + "grad_norm": 1.8414321101455968, + "language_loss": 0.6398629, + "learning_rate": 2.006776923375082e-07, + "loss": 0.66197437, + "num_input_tokens_seen": 154866770, + "step": 7156, + "time_per_iteration": 3.602290630340576 + }, + { + "auxiliary_loss_clip": 0.01168735, + "auxiliary_loss_mlp": 0.01021742, + "balance_loss_clip": 1.04796243, + "balance_loss_mlp": 1.01469958, + "epoch": 0.860578368303974, + "flos": 22596538072320.0, + "grad_norm": 1.691051387575935, + "language_loss": 0.71175313, + "learning_rate": 2.003377385022764e-07, + "loss": 0.73365784, + "num_input_tokens_seen": 154885595, + "step": 7157, + "time_per_iteration": 2.641815662384033 + }, + { + "auxiliary_loss_clip": 0.0118603, + "auxiliary_loss_mlp": 0.01021675, + "balance_loss_clip": 1.04470587, + "balance_loss_mlp": 1.01454079, + "epoch": 0.8606986111946131, + "flos": 21324331192320.0, + "grad_norm": 2.3682559359068827, + "language_loss": 0.77587867, + "learning_rate": 1.9999805767437826e-07, + "loss": 0.79795581, + "num_input_tokens_seen": 154904485, + "step": 7158, + "time_per_iteration": 2.6522397994995117 + }, + { + "auxiliary_loss_clip": 0.01174047, + "auxiliary_loss_mlp": 0.01023727, + "balance_loss_clip": 1.03996253, + "balance_loss_mlp": 1.01620746, + "epoch": 0.8608188540852522, + "flos": 28877206769280.0, + "grad_norm": 1.6177850184506875, + "language_loss": 0.71731305, + "learning_rate": 1.9965864990534386e-07, + "loss": 0.73929083, + "num_input_tokens_seen": 154925010, + "step": 7159, + "time_per_iteration": 2.698934316635132 + }, + { + "auxiliary_loss_clip": 0.01184858, + "auxiliary_loss_mlp": 0.01019007, + "balance_loss_clip": 1.04045057, + "balance_loss_mlp": 1.01215243, + "epoch": 0.8609390969758913, + "flos": 29716187713920.0, + "grad_norm": 1.5957632409639375, + "language_loss": 0.77668488, + "learning_rate": 1.9931951524666092e-07, + "loss": 0.79872358, + "num_input_tokens_seen": 154946100, + "step": 7160, + "time_per_iteration": 3.8182373046875 + }, + { + "auxiliary_loss_clip": 0.0118061, + "auxiliary_loss_mlp": 0.01378637, + "balance_loss_clip": 1.0465399, + "balance_loss_mlp": 1.00021076, + "epoch": 0.8610593398665304, + "flos": 21249349551360.0, + "grad_norm": 1.6561410434191726, + "language_loss": 0.8115105, + "learning_rate": 1.9898065374977534e-07, + "loss": 0.83710301, + "num_input_tokens_seen": 154966305, + "step": 7161, + "time_per_iteration": 2.6844890117645264 + }, + { + "auxiliary_loss_clip": 0.01190956, + "auxiliary_loss_mlp": 0.0102186, + "balance_loss_clip": 1.04154038, + "balance_loss_mlp": 1.01592886, + "epoch": 0.8611795827571694, + "flos": 14830102183680.0, + "grad_norm": 1.832964336233198, + "language_loss": 0.72905278, + "learning_rate": 1.9864206546609342e-07, + "loss": 0.75118095, + "num_input_tokens_seen": 154985145, + "step": 7162, + "time_per_iteration": 2.674529552459717 + }, + { + "auxiliary_loss_clip": 0.01164825, + "auxiliary_loss_mlp": 0.01026072, + "balance_loss_clip": 1.04450083, + "balance_loss_mlp": 1.01878786, + "epoch": 0.8612998256478086, + "flos": 24243258107520.0, + "grad_norm": 1.6859991385513218, + "language_loss": 0.8439768, + "learning_rate": 1.983037504469771e-07, + "loss": 0.86588573, + "num_input_tokens_seen": 155003855, + "step": 7163, + "time_per_iteration": 2.6051254272460938 + }, + { + "auxiliary_loss_clip": 0.01176618, + "auxiliary_loss_mlp": 0.01032376, + "balance_loss_clip": 1.04664457, + "balance_loss_mlp": 1.02487469, + "epoch": 0.8614200685384477, + "flos": 21252653602560.0, + "grad_norm": 1.825561012108549, + "language_loss": 0.66585064, + "learning_rate": 1.9796570874374984e-07, + "loss": 0.6879406, + "num_input_tokens_seen": 155023960, + "step": 7164, + "time_per_iteration": 2.6089839935302734 + }, + { + "auxiliary_loss_clip": 0.01186965, + "auxiliary_loss_mlp": 0.01025504, + "balance_loss_clip": 1.04344773, + "balance_loss_mlp": 1.01770735, + "epoch": 0.8615403114290867, + "flos": 20007738080640.0, + "grad_norm": 1.700738244354356, + "language_loss": 0.77578807, + "learning_rate": 1.976279404076917e-07, + "loss": 0.79791278, + "num_input_tokens_seen": 155043360, + "step": 7165, + "time_per_iteration": 2.707943916320801 + }, + { + "auxiliary_loss_clip": 0.01194457, + "auxiliary_loss_mlp": 0.0102437, + "balance_loss_clip": 1.04484355, + "balance_loss_mlp": 1.01744127, + "epoch": 0.8616605543197259, + "flos": 29789373674880.0, + "grad_norm": 1.7948689743562154, + "language_loss": 0.75779235, + "learning_rate": 1.9729044549004193e-07, + "loss": 0.77998054, + "num_input_tokens_seen": 155064745, + "step": 7166, + "time_per_iteration": 2.755448341369629 + }, + { + "auxiliary_loss_clip": 0.01173295, + "auxiliary_loss_mlp": 0.01023696, + "balance_loss_clip": 1.04597354, + "balance_loss_mlp": 1.01650715, + "epoch": 0.8617807972103649, + "flos": 28911609020160.0, + "grad_norm": 2.272878388694853, + "language_loss": 0.70219827, + "learning_rate": 1.9695322404199822e-07, + "loss": 0.72416812, + "num_input_tokens_seen": 155086790, + "step": 7167, + "time_per_iteration": 2.651374101638794 + }, + { + "auxiliary_loss_clip": 0.01187429, + "auxiliary_loss_mlp": 0.01022811, + "balance_loss_clip": 1.04558718, + "balance_loss_mlp": 1.01513946, + "epoch": 0.861901040101004, + "flos": 27673804391040.0, + "grad_norm": 2.2058067336901876, + "language_loss": 0.82041007, + "learning_rate": 1.9661627611471654e-07, + "loss": 0.84251249, + "num_input_tokens_seen": 155106585, + "step": 7168, + "time_per_iteration": 2.7176012992858887 + }, + { + "auxiliary_loss_clip": 0.01191901, + "auxiliary_loss_mlp": 0.01023288, + "balance_loss_clip": 1.04429591, + "balance_loss_mlp": 1.01562917, + "epoch": 0.8620212829916432, + "flos": 49748056755840.0, + "grad_norm": 2.493216537340909, + "language_loss": 0.7028324, + "learning_rate": 1.9627960175931246e-07, + "loss": 0.72498429, + "num_input_tokens_seen": 155131285, + "step": 7169, + "time_per_iteration": 2.840951919555664 + }, + { + "auxiliary_loss_clip": 0.01177854, + "auxiliary_loss_mlp": 0.01022985, + "balance_loss_clip": 1.04828691, + "balance_loss_mlp": 1.01604664, + "epoch": 0.8621415258822822, + "flos": 21138672769920.0, + "grad_norm": 1.8495234175249875, + "language_loss": 0.74202186, + "learning_rate": 1.9594320102685847e-07, + "loss": 0.76403028, + "num_input_tokens_seen": 155150555, + "step": 7170, + "time_per_iteration": 2.7051687240600586 + }, + { + "auxiliary_loss_clip": 0.01174144, + "auxiliary_loss_mlp": 0.01378282, + "balance_loss_clip": 1.04308534, + "balance_loss_mlp": 1.00006199, + "epoch": 0.8622617687729213, + "flos": 21689039934720.0, + "grad_norm": 2.2670492224975054, + "language_loss": 0.64489734, + "learning_rate": 1.956070739683864e-07, + "loss": 0.6704216, + "num_input_tokens_seen": 155169890, + "step": 7171, + "time_per_iteration": 2.6462459564208984 + }, + { + "auxiliary_loss_clip": 0.01175067, + "auxiliary_loss_mlp": 0.01024402, + "balance_loss_clip": 1.03957725, + "balance_loss_mlp": 1.01766622, + "epoch": 0.8623820116635604, + "flos": 26250592734720.0, + "grad_norm": 1.5082360584798313, + "language_loss": 0.74197203, + "learning_rate": 1.9527122063488678e-07, + "loss": 0.76396668, + "num_input_tokens_seen": 155191005, + "step": 7172, + "time_per_iteration": 2.7379324436187744 + }, + { + "auxiliary_loss_clip": 0.01181453, + "auxiliary_loss_mlp": 0.01020396, + "balance_loss_clip": 1.04095328, + "balance_loss_mlp": 1.01359797, + "epoch": 0.8625022545541995, + "flos": 19647554451840.0, + "grad_norm": 1.852613576758301, + "language_loss": 0.80374467, + "learning_rate": 1.9493564107730755e-07, + "loss": 0.82576311, + "num_input_tokens_seen": 155211005, + "step": 7173, + "time_per_iteration": 2.665686845779419 + }, + { + "auxiliary_loss_clip": 0.0117558, + "auxiliary_loss_mlp": 0.01027042, + "balance_loss_clip": 1.03978682, + "balance_loss_mlp": 1.01978803, + "epoch": 0.8626224974448385, + "flos": 21908382336000.0, + "grad_norm": 1.835011657621352, + "language_loss": 0.6101734, + "learning_rate": 1.9460033534655684e-07, + "loss": 0.63219965, + "num_input_tokens_seen": 155230365, + "step": 7174, + "time_per_iteration": 2.6863884925842285 + }, + { + "auxiliary_loss_clip": 0.01174286, + "auxiliary_loss_mlp": 0.01027907, + "balance_loss_clip": 1.0379529, + "balance_loss_mlp": 1.02078414, + "epoch": 0.8627427403354777, + "flos": 23331198942720.0, + "grad_norm": 1.4803697052629101, + "language_loss": 0.84020323, + "learning_rate": 1.9426530349349978e-07, + "loss": 0.86222517, + "num_input_tokens_seen": 155250815, + "step": 7175, + "time_per_iteration": 2.6592438220977783 + }, + { + "auxiliary_loss_clip": 0.01171216, + "auxiliary_loss_mlp": 0.01378448, + "balance_loss_clip": 1.04305851, + "balance_loss_mlp": 1.00014758, + "epoch": 0.8628629832261168, + "flos": 16362877299840.0, + "grad_norm": 1.995007721016844, + "language_loss": 0.64598638, + "learning_rate": 1.9393054556896038e-07, + "loss": 0.67148304, + "num_input_tokens_seen": 155268515, + "step": 7176, + "time_per_iteration": 2.593406915664673 + }, + { + "auxiliary_loss_clip": 0.01187591, + "auxiliary_loss_mlp": 0.01028822, + "balance_loss_clip": 1.0404408, + "balance_loss_mlp": 1.02099848, + "epoch": 0.8629832261167558, + "flos": 28103941756800.0, + "grad_norm": 2.6355380589517714, + "language_loss": 0.69536084, + "learning_rate": 1.9359606162372133e-07, + "loss": 0.71752495, + "num_input_tokens_seen": 155290120, + "step": 7177, + "time_per_iteration": 2.7239866256713867 + }, + { + "auxiliary_loss_clip": 0.01166747, + "auxiliary_loss_mlp": 0.0102351, + "balance_loss_clip": 1.04710317, + "balance_loss_mlp": 1.01648903, + "epoch": 0.863103469007395, + "flos": 20230061310720.0, + "grad_norm": 1.6091862169077402, + "language_loss": 0.70651686, + "learning_rate": 1.9326185170852293e-07, + "loss": 0.72841942, + "num_input_tokens_seen": 155309085, + "step": 7178, + "time_per_iteration": 3.5015180110931396 + }, + { + "auxiliary_loss_clip": 0.01172536, + "auxiliary_loss_mlp": 0.01025803, + "balance_loss_clip": 1.04440427, + "balance_loss_mlp": 1.0185281, + "epoch": 0.863223711898034, + "flos": 24498547044480.0, + "grad_norm": 1.9528575230497942, + "language_loss": 0.72353578, + "learning_rate": 1.9292791587406598e-07, + "loss": 0.74551916, + "num_input_tokens_seen": 155327945, + "step": 7179, + "time_per_iteration": 3.7390341758728027 + }, + { + "auxiliary_loss_clip": 0.0116945, + "auxiliary_loss_mlp": 0.01378801, + "balance_loss_clip": 1.04184628, + "balance_loss_mlp": 1.00014138, + "epoch": 0.8633439547886731, + "flos": 17675376261120.0, + "grad_norm": 2.4601348214363234, + "language_loss": 0.86993784, + "learning_rate": 1.9259425417100661e-07, + "loss": 0.89542031, + "num_input_tokens_seen": 155344060, + "step": 7180, + "time_per_iteration": 2.595442056655884 + }, + { + "auxiliary_loss_clip": 0.0120022, + "auxiliary_loss_mlp": 0.01024622, + "balance_loss_clip": 1.03315973, + "balance_loss_mlp": 1.01711822, + "epoch": 0.8634641976793123, + "flos": 12895055677440.0, + "grad_norm": 2.463015711872218, + "language_loss": 0.75004339, + "learning_rate": 1.9226086664996234e-07, + "loss": 0.77229184, + "num_input_tokens_seen": 155362305, + "step": 7181, + "time_per_iteration": 2.7839229106903076 + }, + { + "auxiliary_loss_clip": 0.01187752, + "auxiliary_loss_mlp": 0.0102492, + "balance_loss_clip": 1.04530168, + "balance_loss_mlp": 1.01816416, + "epoch": 0.8635844405699513, + "flos": 23878980328320.0, + "grad_norm": 2.480421125708257, + "language_loss": 0.74461877, + "learning_rate": 1.9192775336150712e-07, + "loss": 0.76674551, + "num_input_tokens_seen": 155382605, + "step": 7182, + "time_per_iteration": 3.5293643474578857 + }, + { + "auxiliary_loss_clip": 0.01071849, + "auxiliary_loss_mlp": 0.01009585, + "balance_loss_clip": 1.01098514, + "balance_loss_mlp": 1.00854814, + "epoch": 0.8637046834605904, + "flos": 60453387521280.0, + "grad_norm": 0.7679004256594674, + "language_loss": 0.56270564, + "learning_rate": 1.915949143561739e-07, + "loss": 0.58351994, + "num_input_tokens_seen": 155437280, + "step": 7183, + "time_per_iteration": 3.106386661529541 + }, + { + "auxiliary_loss_clip": 0.01177852, + "auxiliary_loss_mlp": 0.01025678, + "balance_loss_clip": 1.04752731, + "balance_loss_mlp": 1.01773322, + "epoch": 0.8638249263512295, + "flos": 20558751690240.0, + "grad_norm": 2.6329816667110166, + "language_loss": 0.78050405, + "learning_rate": 1.9126234968445498e-07, + "loss": 0.80253935, + "num_input_tokens_seen": 155456970, + "step": 7184, + "time_per_iteration": 2.5524871349334717 + }, + { + "auxiliary_loss_clip": 0.01166196, + "auxiliary_loss_mlp": 0.01020165, + "balance_loss_clip": 1.04561806, + "balance_loss_mlp": 1.01303303, + "epoch": 0.8639451692418686, + "flos": 26615768353920.0, + "grad_norm": 2.545621128939335, + "language_loss": 0.67618638, + "learning_rate": 1.9093005939679884e-07, + "loss": 0.69805002, + "num_input_tokens_seen": 155478925, + "step": 7185, + "time_per_iteration": 2.5960099697113037 + }, + { + "auxiliary_loss_clip": 0.01177962, + "auxiliary_loss_mlp": 0.01026735, + "balance_loss_clip": 1.04815078, + "balance_loss_mlp": 1.01900983, + "epoch": 0.8640654121325076, + "flos": 15122450977920.0, + "grad_norm": 5.002631317441076, + "language_loss": 0.76668924, + "learning_rate": 1.9059804354361452e-07, + "loss": 0.78873622, + "num_input_tokens_seen": 155496700, + "step": 7186, + "time_per_iteration": 3.491151809692383 + }, + { + "auxiliary_loss_clip": 0.01174734, + "auxiliary_loss_mlp": 0.01028397, + "balance_loss_clip": 1.04046774, + "balance_loss_mlp": 1.02148342, + "epoch": 0.8641856550231467, + "flos": 31869068250240.0, + "grad_norm": 1.6437683508145038, + "language_loss": 0.70141453, + "learning_rate": 1.902663021752684e-07, + "loss": 0.72344589, + "num_input_tokens_seen": 155518130, + "step": 7187, + "time_per_iteration": 2.725128650665283 + }, + { + "auxiliary_loss_clip": 0.01170978, + "auxiliary_loss_mlp": 0.01026231, + "balance_loss_clip": 1.0499289, + "balance_loss_mlp": 1.01890564, + "epoch": 0.8643058979137859, + "flos": 14976545932800.0, + "grad_norm": 3.544060049301615, + "language_loss": 0.82151997, + "learning_rate": 1.8993483534208556e-07, + "loss": 0.84349203, + "num_input_tokens_seen": 155537040, + "step": 7188, + "time_per_iteration": 2.5965988636016846 + }, + { + "auxiliary_loss_clip": 0.01178592, + "auxiliary_loss_mlp": 0.01027669, + "balance_loss_clip": 1.04375494, + "balance_loss_mlp": 1.020123, + "epoch": 0.8644261408044249, + "flos": 13115726881920.0, + "grad_norm": 3.059171826256663, + "language_loss": 0.74642098, + "learning_rate": 1.8960364309434884e-07, + "loss": 0.76848352, + "num_input_tokens_seen": 155554535, + "step": 7189, + "time_per_iteration": 2.6271920204162598 + }, + { + "auxiliary_loss_clip": 0.01205578, + "auxiliary_loss_mlp": 0.01378689, + "balance_loss_clip": 1.03897572, + "balance_loss_mlp": 1.00012541, + "epoch": 0.864546383695064, + "flos": 20850920916480.0, + "grad_norm": 1.6300881535301288, + "language_loss": 0.78427613, + "learning_rate": 1.8927272548229967e-07, + "loss": 0.81011879, + "num_input_tokens_seen": 155574225, + "step": 7190, + "time_per_iteration": 2.8375039100646973 + }, + { + "auxiliary_loss_clip": 0.01204965, + "auxiliary_loss_mlp": 0.01026637, + "balance_loss_clip": 1.04257345, + "balance_loss_mlp": 1.01953816, + "epoch": 0.8646666265857031, + "flos": 21324582587520.0, + "grad_norm": 1.5263540836725549, + "language_loss": 0.82973152, + "learning_rate": 1.8894208255613876e-07, + "loss": 0.85204756, + "num_input_tokens_seen": 155593540, + "step": 7191, + "time_per_iteration": 2.8507258892059326 + }, + { + "auxiliary_loss_clip": 0.01168849, + "auxiliary_loss_mlp": 0.01025494, + "balance_loss_clip": 1.04846215, + "balance_loss_mlp": 1.01861835, + "epoch": 0.8647868694763422, + "flos": 19750833031680.0, + "grad_norm": 2.2300083458499502, + "language_loss": 0.77941066, + "learning_rate": 1.8861171436602397e-07, + "loss": 0.80135411, + "num_input_tokens_seen": 155610655, + "step": 7192, + "time_per_iteration": 2.5609889030456543 + }, + { + "auxiliary_loss_clip": 0.01178812, + "auxiliary_loss_mlp": 0.01024189, + "balance_loss_clip": 1.04746354, + "balance_loss_mlp": 1.01714957, + "epoch": 0.8649071123669813, + "flos": 26176760328960.0, + "grad_norm": 2.691644834012578, + "language_loss": 0.80593866, + "learning_rate": 1.882816209620719e-07, + "loss": 0.82796866, + "num_input_tokens_seen": 155627365, + "step": 7193, + "time_per_iteration": 2.677671432495117 + }, + { + "auxiliary_loss_clip": 0.01189559, + "auxiliary_loss_mlp": 0.01028916, + "balance_loss_clip": 1.04796553, + "balance_loss_mlp": 1.02169168, + "epoch": 0.8650273552576204, + "flos": 20302888135680.0, + "grad_norm": 1.8710852907547146, + "language_loss": 0.76950264, + "learning_rate": 1.8795180239435738e-07, + "loss": 0.79168737, + "num_input_tokens_seen": 155646220, + "step": 7194, + "time_per_iteration": 2.6719818115234375 + }, + { + "auxiliary_loss_clip": 0.01190615, + "auxiliary_loss_mlp": 0.01026144, + "balance_loss_clip": 1.04491997, + "balance_loss_mlp": 1.01919448, + "epoch": 0.8651475981482595, + "flos": 23951088881280.0, + "grad_norm": 4.665974859784938, + "language_loss": 0.75610662, + "learning_rate": 1.8762225871291348e-07, + "loss": 0.77827418, + "num_input_tokens_seen": 155662095, + "step": 7195, + "time_per_iteration": 2.6947288513183594 + }, + { + "auxiliary_loss_clip": 0.01166296, + "auxiliary_loss_mlp": 0.01378631, + "balance_loss_clip": 1.04578578, + "balance_loss_mlp": 1.00010777, + "epoch": 0.8652678410388985, + "flos": 21684622561920.0, + "grad_norm": 1.7926458588913905, + "language_loss": 0.80949831, + "learning_rate": 1.8729298996773201e-07, + "loss": 0.83494759, + "num_input_tokens_seen": 155680845, + "step": 7196, + "time_per_iteration": 2.5647521018981934 + }, + { + "auxiliary_loss_clip": 0.01069068, + "auxiliary_loss_mlp": 0.00999123, + "balance_loss_clip": 1.00947952, + "balance_loss_mlp": 0.99804461, + "epoch": 0.8653880839295377, + "flos": 65224660855680.0, + "grad_norm": 0.8365065864839969, + "language_loss": 0.60942996, + "learning_rate": 1.8696399620876301e-07, + "loss": 0.63011187, + "num_input_tokens_seen": 155737875, + "step": 7197, + "time_per_iteration": 3.1314289569854736 + }, + { + "auxiliary_loss_clip": 0.01184195, + "auxiliary_loss_mlp": 0.0102555, + "balance_loss_clip": 1.03621411, + "balance_loss_mlp": 1.01819181, + "epoch": 0.8655083268201768, + "flos": 17749172753280.0, + "grad_norm": 33.286736805775696, + "language_loss": 0.79279006, + "learning_rate": 1.866352774859141e-07, + "loss": 0.81488746, + "num_input_tokens_seen": 155753100, + "step": 7198, + "time_per_iteration": 2.6322288513183594 + }, + { + "auxiliary_loss_clip": 0.01195866, + "auxiliary_loss_mlp": 0.01023879, + "balance_loss_clip": 1.04151356, + "balance_loss_mlp": 1.01667857, + "epoch": 0.8656285697108158, + "flos": 20703974376960.0, + "grad_norm": 2.1487568355937245, + "language_loss": 0.6900202, + "learning_rate": 1.8630683384905188e-07, + "loss": 0.71221763, + "num_input_tokens_seen": 155772430, + "step": 7199, + "time_per_iteration": 2.6764400005340576 + }, + { + "auxiliary_loss_clip": 0.01168912, + "auxiliary_loss_mlp": 0.0137867, + "balance_loss_clip": 1.04812503, + "balance_loss_mlp": 1.00011027, + "epoch": 0.865748812601455, + "flos": 18653833716480.0, + "grad_norm": 1.8997036292952754, + "language_loss": 0.88753736, + "learning_rate": 1.8597866534800045e-07, + "loss": 0.91301322, + "num_input_tokens_seen": 155787545, + "step": 7200, + "time_per_iteration": 2.5566468238830566 + }, + { + "auxiliary_loss_clip": 0.01180864, + "auxiliary_loss_mlp": 0.0137856, + "balance_loss_clip": 1.04799938, + "balance_loss_mlp": 1.00012481, + "epoch": 0.865869055492094, + "flos": 70652554807680.0, + "grad_norm": 2.0329028763156134, + "language_loss": 0.74377441, + "learning_rate": 1.8565077203254398e-07, + "loss": 0.76936865, + "num_input_tokens_seen": 155813005, + "step": 7201, + "time_per_iteration": 3.0560481548309326 + }, + { + "auxiliary_loss_clip": 0.01199174, + "auxiliary_loss_mlp": 0.0102596, + "balance_loss_clip": 1.04806602, + "balance_loss_mlp": 1.01835728, + "epoch": 0.8659892983827331, + "flos": 17383961220480.0, + "grad_norm": 2.4301118712398084, + "language_loss": 0.72534502, + "learning_rate": 1.8532315395242203e-07, + "loss": 0.74759632, + "num_input_tokens_seen": 155829455, + "step": 7202, + "time_per_iteration": 2.7212588787078857 + }, + { + "auxiliary_loss_clip": 0.01195307, + "auxiliary_loss_mlp": 0.01026557, + "balance_loss_clip": 1.04206014, + "balance_loss_mlp": 1.01982737, + "epoch": 0.8661095412733723, + "flos": 17895221452800.0, + "grad_norm": 2.030576250328214, + "language_loss": 0.7215482, + "learning_rate": 1.849958111573353e-07, + "loss": 0.7437669, + "num_input_tokens_seen": 155848060, + "step": 7203, + "time_per_iteration": 2.6702630519866943 + }, + { + "auxiliary_loss_clip": 0.01163497, + "auxiliary_loss_mlp": 0.01024604, + "balance_loss_clip": 1.04428816, + "balance_loss_mlp": 1.01741278, + "epoch": 0.8662297841640113, + "flos": 18224163227520.0, + "grad_norm": 1.7531005723312572, + "language_loss": 0.63986218, + "learning_rate": 1.8466874369694074e-07, + "loss": 0.66174316, + "num_input_tokens_seen": 155865755, + "step": 7204, + "time_per_iteration": 3.601560592651367 + }, + { + "auxiliary_loss_clip": 0.01188487, + "auxiliary_loss_mlp": 0.01022398, + "balance_loss_clip": 1.03819799, + "balance_loss_mlp": 1.01562953, + "epoch": 0.8663500270546504, + "flos": 16362159027840.0, + "grad_norm": 2.238910149063721, + "language_loss": 0.70115542, + "learning_rate": 1.843419516208542e-07, + "loss": 0.72326428, + "num_input_tokens_seen": 155882680, + "step": 7205, + "time_per_iteration": 3.887641668319702 + }, + { + "auxiliary_loss_clip": 0.01179942, + "auxiliary_loss_mlp": 0.01025011, + "balance_loss_clip": 1.04934108, + "balance_loss_mlp": 1.0173099, + "epoch": 0.8664702699452895, + "flos": 17894431353600.0, + "grad_norm": 2.466207063152627, + "language_loss": 0.80148244, + "learning_rate": 1.8401543497865047e-07, + "loss": 0.82353199, + "num_input_tokens_seen": 155900680, + "step": 7206, + "time_per_iteration": 2.660895824432373 + }, + { + "auxiliary_loss_clip": 0.01177987, + "auxiliary_loss_mlp": 0.01378632, + "balance_loss_clip": 1.04467058, + "balance_loss_mlp": 1.00019765, + "epoch": 0.8665905128359286, + "flos": 30736373794560.0, + "grad_norm": 2.540586803196661, + "language_loss": 0.64309925, + "learning_rate": 1.836891938198608e-07, + "loss": 0.66866541, + "num_input_tokens_seen": 155921105, + "step": 7207, + "time_per_iteration": 2.78216290473938 + }, + { + "auxiliary_loss_clip": 0.01184877, + "auxiliary_loss_mlp": 0.01027475, + "balance_loss_clip": 1.045403, + "balance_loss_mlp": 1.02060246, + "epoch": 0.8667107557265676, + "flos": 18656419495680.0, + "grad_norm": 5.181233191156189, + "language_loss": 0.71205127, + "learning_rate": 1.8336322819397677e-07, + "loss": 0.73417479, + "num_input_tokens_seen": 155938640, + "step": 7208, + "time_per_iteration": 3.5368492603302 + }, + { + "auxiliary_loss_clip": 0.01194105, + "auxiliary_loss_mlp": 0.01027303, + "balance_loss_clip": 1.03935421, + "balance_loss_mlp": 1.01969755, + "epoch": 0.8668309986172068, + "flos": 20083725302400.0, + "grad_norm": 1.8961441896286975, + "language_loss": 0.624879, + "learning_rate": 1.8303753815044654e-07, + "loss": 0.64709306, + "num_input_tokens_seen": 155957945, + "step": 7209, + "time_per_iteration": 2.668632984161377 + }, + { + "auxiliary_loss_clip": 0.01193625, + "auxiliary_loss_mlp": 0.01026205, + "balance_loss_clip": 1.04287434, + "balance_loss_mlp": 1.01803303, + "epoch": 0.8669512415078459, + "flos": 21615099788160.0, + "grad_norm": 2.458908840758716, + "language_loss": 0.70289183, + "learning_rate": 1.827121237386773e-07, + "loss": 0.72509015, + "num_input_tokens_seen": 155975390, + "step": 7210, + "time_per_iteration": 2.6681222915649414 + }, + { + "auxiliary_loss_clip": 0.01189094, + "auxiliary_loss_mlp": 0.01026502, + "balance_loss_clip": 1.04442096, + "balance_loss_mlp": 1.01878905, + "epoch": 0.8670714843984849, + "flos": 17703601372800.0, + "grad_norm": 2.192316326096098, + "language_loss": 0.74624729, + "learning_rate": 1.8238698500803374e-07, + "loss": 0.76840329, + "num_input_tokens_seen": 155988155, + "step": 7211, + "time_per_iteration": 2.58876633644104 + }, + { + "auxiliary_loss_clip": 0.01073811, + "auxiliary_loss_mlp": 0.01004787, + "balance_loss_clip": 1.00883198, + "balance_loss_mlp": 1.00371456, + "epoch": 0.8671917272891241, + "flos": 60705483125760.0, + "grad_norm": 0.7178011539862161, + "language_loss": 0.56235081, + "learning_rate": 1.820621220078391e-07, + "loss": 0.5831368, + "num_input_tokens_seen": 156052065, + "step": 7212, + "time_per_iteration": 4.184757471084595 + }, + { + "auxiliary_loss_clip": 0.01166389, + "auxiliary_loss_mlp": 0.01025174, + "balance_loss_clip": 1.04556799, + "balance_loss_mlp": 1.01769388, + "epoch": 0.8673119701797631, + "flos": 20451881750400.0, + "grad_norm": 1.546090529745323, + "language_loss": 0.67634928, + "learning_rate": 1.8173753478737553e-07, + "loss": 0.6982649, + "num_input_tokens_seen": 156072500, + "step": 7213, + "time_per_iteration": 2.715595245361328 + }, + { + "auxiliary_loss_clip": 0.01165618, + "auxiliary_loss_mlp": 0.01026182, + "balance_loss_clip": 1.04542279, + "balance_loss_mlp": 1.019104, + "epoch": 0.8674322130704022, + "flos": 19647410797440.0, + "grad_norm": 2.8460712814388383, + "language_loss": 0.799182, + "learning_rate": 1.8141322339588205e-07, + "loss": 0.8211, + "num_input_tokens_seen": 156089840, + "step": 7214, + "time_per_iteration": 2.563384771347046 + }, + { + "auxiliary_loss_clip": 0.01166336, + "auxiliary_loss_mlp": 0.01026674, + "balance_loss_clip": 1.04736662, + "balance_loss_mlp": 1.01928258, + "epoch": 0.8675524559610414, + "flos": 26025001367040.0, + "grad_norm": 1.8344894977214798, + "language_loss": 0.70417035, + "learning_rate": 1.810891878825569e-07, + "loss": 0.72610044, + "num_input_tokens_seen": 156109815, + "step": 7215, + "time_per_iteration": 2.6905932426452637 + }, + { + "auxiliary_loss_clip": 0.01183301, + "auxiliary_loss_mlp": 0.01025266, + "balance_loss_clip": 1.04276657, + "balance_loss_mlp": 1.01800013, + "epoch": 0.8676726988516804, + "flos": 15049444584960.0, + "grad_norm": 1.9958389788266506, + "language_loss": 0.71709907, + "learning_rate": 1.8076542829655561e-07, + "loss": 0.7391848, + "num_input_tokens_seen": 156128620, + "step": 7216, + "time_per_iteration": 2.5847201347351074 + }, + { + "auxiliary_loss_clip": 0.01188272, + "auxiliary_loss_mlp": 0.01028425, + "balance_loss_clip": 1.04702377, + "balance_loss_mlp": 1.02046227, + "epoch": 0.8677929417423195, + "flos": 16288111140480.0, + "grad_norm": 1.986545522327044, + "language_loss": 0.79287517, + "learning_rate": 1.8044194468699203e-07, + "loss": 0.81504214, + "num_input_tokens_seen": 156145930, + "step": 7217, + "time_per_iteration": 2.63161301612854 + }, + { + "auxiliary_loss_clip": 0.0118658, + "auxiliary_loss_mlp": 0.01021645, + "balance_loss_clip": 1.04797232, + "balance_loss_mlp": 1.01393783, + "epoch": 0.8679131846329585, + "flos": 18844160906880.0, + "grad_norm": 2.5892653002931234, + "language_loss": 0.7574271, + "learning_rate": 1.8011873710293912e-07, + "loss": 0.77950931, + "num_input_tokens_seen": 156164435, + "step": 7218, + "time_per_iteration": 2.6129302978515625 + }, + { + "auxiliary_loss_clip": 0.01173939, + "auxiliary_loss_mlp": 0.01028153, + "balance_loss_clip": 1.04603267, + "balance_loss_mlp": 1.02058899, + "epoch": 0.8680334275235977, + "flos": 33620718890880.0, + "grad_norm": 2.085022061536851, + "language_loss": 0.69521391, + "learning_rate": 1.7979580559342677e-07, + "loss": 0.71723479, + "num_input_tokens_seen": 156185165, + "step": 7219, + "time_per_iteration": 2.7703866958618164 + }, + { + "auxiliary_loss_clip": 0.01184635, + "auxiliary_loss_mlp": 0.01024048, + "balance_loss_clip": 1.04540133, + "balance_loss_mlp": 1.01609111, + "epoch": 0.8681536704142367, + "flos": 24681152810880.0, + "grad_norm": 1.6139056621381247, + "language_loss": 0.67001295, + "learning_rate": 1.7947315020744358e-07, + "loss": 0.69209975, + "num_input_tokens_seen": 156206260, + "step": 7220, + "time_per_iteration": 2.717933177947998 + }, + { + "auxiliary_loss_clip": 0.01182972, + "auxiliary_loss_mlp": 0.01019031, + "balance_loss_clip": 1.04325891, + "balance_loss_mlp": 1.01209867, + "epoch": 0.8682739133048758, + "flos": 20011042131840.0, + "grad_norm": 1.9214566874971963, + "language_loss": 0.80621916, + "learning_rate": 1.7915077099393594e-07, + "loss": 0.8282392, + "num_input_tokens_seen": 156222860, + "step": 7221, + "time_per_iteration": 2.6268367767333984 + }, + { + "auxiliary_loss_clip": 0.01176682, + "auxiliary_loss_mlp": 0.01028193, + "balance_loss_clip": 1.04471111, + "balance_loss_mlp": 1.02104676, + "epoch": 0.868394156195515, + "flos": 16654759217280.0, + "grad_norm": 2.1197472119560965, + "language_loss": 0.73610795, + "learning_rate": 1.788286680018083e-07, + "loss": 0.75815678, + "num_input_tokens_seen": 156241570, + "step": 7222, + "time_per_iteration": 2.6270341873168945 + }, + { + "auxiliary_loss_clip": 0.01188677, + "auxiliary_loss_mlp": 0.01024302, + "balance_loss_clip": 1.04519939, + "balance_loss_mlp": 1.01719368, + "epoch": 0.868514399086154, + "flos": 28001381448960.0, + "grad_norm": 1.5528038613584987, + "language_loss": 0.7233606, + "learning_rate": 1.7850684127992443e-07, + "loss": 0.74549037, + "num_input_tokens_seen": 156261315, + "step": 7223, + "time_per_iteration": 2.7905571460723877 + }, + { + "auxiliary_loss_clip": 0.0119556, + "auxiliary_loss_mlp": 0.01030776, + "balance_loss_clip": 1.04656982, + "balance_loss_mlp": 1.02381766, + "epoch": 0.8686346419767931, + "flos": 20084587228800.0, + "grad_norm": 2.628491315586782, + "language_loss": 0.70451909, + "learning_rate": 1.7818529087710378e-07, + "loss": 0.72678244, + "num_input_tokens_seen": 156281670, + "step": 7224, + "time_per_iteration": 2.766166925430298 + }, + { + "auxiliary_loss_clip": 0.01178099, + "auxiliary_loss_mlp": 0.01378637, + "balance_loss_clip": 1.04684746, + "balance_loss_mlp": 1.00012124, + "epoch": 0.8687548848674322, + "flos": 18223516782720.0, + "grad_norm": 1.697032216788881, + "language_loss": 0.84136468, + "learning_rate": 1.7786401684212637e-07, + "loss": 0.86693203, + "num_input_tokens_seen": 156300500, + "step": 7225, + "time_per_iteration": 2.696822166442871 + }, + { + "auxiliary_loss_clip": 0.01096856, + "auxiliary_loss_mlp": 0.0100141, + "balance_loss_clip": 1.01066279, + "balance_loss_mlp": 1.00020647, + "epoch": 0.8688751277580713, + "flos": 70457885049600.0, + "grad_norm": 0.846912444692541, + "language_loss": 0.55975115, + "learning_rate": 1.7754301922372883e-07, + "loss": 0.58073378, + "num_input_tokens_seen": 156350145, + "step": 7226, + "time_per_iteration": 3.2059450149536133 + }, + { + "auxiliary_loss_clip": 0.01215561, + "auxiliary_loss_mlp": 0.01025881, + "balance_loss_clip": 1.03887868, + "balance_loss_mlp": 1.01871014, + "epoch": 0.8689953706487104, + "flos": 26906788344960.0, + "grad_norm": 2.2186618859724923, + "language_loss": 0.80865633, + "learning_rate": 1.7722229807060617e-07, + "loss": 0.83107078, + "num_input_tokens_seen": 156368725, + "step": 7227, + "time_per_iteration": 2.845895767211914 + }, + { + "auxiliary_loss_clip": 0.01181343, + "auxiliary_loss_mlp": 0.010196, + "balance_loss_clip": 1.03979754, + "balance_loss_mlp": 1.01247716, + "epoch": 0.8691156135393495, + "flos": 34637385438720.0, + "grad_norm": 2.1546721741491863, + "language_loss": 0.81777203, + "learning_rate": 1.7690185343141172e-07, + "loss": 0.83978152, + "num_input_tokens_seen": 156388640, + "step": 7228, + "time_per_iteration": 2.921241044998169 + }, + { + "auxiliary_loss_clip": 0.01184343, + "auxiliary_loss_mlp": 0.01026481, + "balance_loss_clip": 1.04293847, + "balance_loss_mlp": 1.01984978, + "epoch": 0.8692358564299886, + "flos": 18989814556800.0, + "grad_norm": 3.188541309614685, + "language_loss": 0.69901711, + "learning_rate": 1.7658168535475615e-07, + "loss": 0.72112536, + "num_input_tokens_seen": 156406425, + "step": 7229, + "time_per_iteration": 3.624180793762207 + }, + { + "auxiliary_loss_clip": 0.01188944, + "auxiliary_loss_mlp": 0.01019221, + "balance_loss_clip": 1.04611182, + "balance_loss_mlp": 1.01221108, + "epoch": 0.8693560993206276, + "flos": 30370839039360.0, + "grad_norm": 1.8257990244110938, + "language_loss": 0.64527923, + "learning_rate": 1.7626179388920948e-07, + "loss": 0.66736084, + "num_input_tokens_seen": 156427705, + "step": 7230, + "time_per_iteration": 2.7954015731811523 + }, + { + "auxiliary_loss_clip": 0.01185193, + "auxiliary_loss_mlp": 0.01378445, + "balance_loss_clip": 1.04342687, + "balance_loss_mlp": 1.00015748, + "epoch": 0.8694763422112668, + "flos": 27200430028800.0, + "grad_norm": 1.7513593041347253, + "language_loss": 0.80442071, + "learning_rate": 1.7594217908329866e-07, + "loss": 0.83005714, + "num_input_tokens_seen": 156449890, + "step": 7231, + "time_per_iteration": 3.698904275894165 + }, + { + "auxiliary_loss_clip": 0.01174945, + "auxiliary_loss_mlp": 0.01025051, + "balance_loss_clip": 1.04388559, + "balance_loss_mlp": 1.01834869, + "epoch": 0.8695965851019059, + "flos": 26139161767680.0, + "grad_norm": 1.8283809055678328, + "language_loss": 0.73843884, + "learning_rate": 1.7562284098550895e-07, + "loss": 0.76043886, + "num_input_tokens_seen": 156469600, + "step": 7232, + "time_per_iteration": 2.703537940979004 + }, + { + "auxiliary_loss_clip": 0.01087004, + "auxiliary_loss_mlp": 0.01004214, + "balance_loss_clip": 1.01579154, + "balance_loss_mlp": 1.00317693, + "epoch": 0.8697168279925449, + "flos": 67332616456320.0, + "grad_norm": 0.8335965025099858, + "language_loss": 0.62242663, + "learning_rate": 1.753037796442838e-07, + "loss": 0.6433388, + "num_input_tokens_seen": 156529040, + "step": 7233, + "time_per_iteration": 3.2589590549468994 + }, + { + "auxiliary_loss_clip": 0.01165703, + "auxiliary_loss_mlp": 0.01026622, + "balance_loss_clip": 1.04546094, + "balance_loss_mlp": 1.01913321, + "epoch": 0.8698370708831841, + "flos": 19718693337600.0, + "grad_norm": 2.297239464147866, + "language_loss": 0.75093615, + "learning_rate": 1.74984995108024e-07, + "loss": 0.77285939, + "num_input_tokens_seen": 156546970, + "step": 7234, + "time_per_iteration": 3.6753902435302734 + }, + { + "auxiliary_loss_clip": 0.0117952, + "auxiliary_loss_mlp": 0.01024486, + "balance_loss_clip": 1.04656851, + "balance_loss_mlp": 1.01719046, + "epoch": 0.8699573137738231, + "flos": 12859971068160.0, + "grad_norm": 1.9312651983865832, + "language_loss": 0.83338749, + "learning_rate": 1.7466648742508981e-07, + "loss": 0.8554275, + "num_input_tokens_seen": 156563155, + "step": 7235, + "time_per_iteration": 2.659724235534668 + }, + { + "auxiliary_loss_clip": 0.0118393, + "auxiliary_loss_mlp": 0.01020795, + "balance_loss_clip": 1.0457201, + "balance_loss_mlp": 1.01370192, + "epoch": 0.8700775566644622, + "flos": 17420733768960.0, + "grad_norm": 1.9763093328575276, + "language_loss": 0.84454012, + "learning_rate": 1.7434825664379837e-07, + "loss": 0.86658734, + "num_input_tokens_seen": 156581660, + "step": 7236, + "time_per_iteration": 2.6878228187561035 + }, + { + "auxiliary_loss_clip": 0.01177002, + "auxiliary_loss_mlp": 0.01023909, + "balance_loss_clip": 1.04571319, + "balance_loss_mlp": 1.01648855, + "epoch": 0.8701977995551013, + "flos": 13735221770880.0, + "grad_norm": 3.410033462909526, + "language_loss": 0.8656311, + "learning_rate": 1.740303028124246e-07, + "loss": 0.88764012, + "num_input_tokens_seen": 156597720, + "step": 7237, + "time_per_iteration": 2.61660099029541 + }, + { + "auxiliary_loss_clip": 0.01200302, + "auxiliary_loss_mlp": 0.01027447, + "balance_loss_clip": 1.0373714, + "balance_loss_mlp": 1.02027655, + "epoch": 0.8703180424457404, + "flos": 30555707362560.0, + "grad_norm": 1.8497454629367214, + "language_loss": 0.75589812, + "learning_rate": 1.7371262597920212e-07, + "loss": 0.77817565, + "num_input_tokens_seen": 156619780, + "step": 7238, + "time_per_iteration": 2.8568360805511475 + }, + { + "auxiliary_loss_clip": 0.01201021, + "auxiliary_loss_mlp": 0.01026308, + "balance_loss_clip": 1.04416513, + "balance_loss_mlp": 1.01895523, + "epoch": 0.8704382853363795, + "flos": 19608986223360.0, + "grad_norm": 1.6417029108523127, + "language_loss": 0.76197338, + "learning_rate": 1.7339522619232195e-07, + "loss": 0.78424668, + "num_input_tokens_seen": 156638160, + "step": 7239, + "time_per_iteration": 3.629950761795044 + }, + { + "auxiliary_loss_clip": 0.01191673, + "auxiliary_loss_mlp": 0.01027528, + "balance_loss_clip": 1.04377568, + "balance_loss_mlp": 1.02034271, + "epoch": 0.8705585282270186, + "flos": 26613900846720.0, + "grad_norm": 2.5719960073256316, + "language_loss": 0.75900805, + "learning_rate": 1.730781034999338e-07, + "loss": 0.78120005, + "num_input_tokens_seen": 156659740, + "step": 7240, + "time_per_iteration": 2.7407608032226562 + }, + { + "auxiliary_loss_clip": 0.01165186, + "auxiliary_loss_mlp": 0.01022649, + "balance_loss_clip": 1.04889977, + "balance_loss_mlp": 1.01615191, + "epoch": 0.8706787711176577, + "flos": 34090465979520.0, + "grad_norm": 1.908673728891641, + "language_loss": 0.73628759, + "learning_rate": 1.7276125795014497e-07, + "loss": 0.75816596, + "num_input_tokens_seen": 156678190, + "step": 7241, + "time_per_iteration": 2.6765425205230713 + }, + { + "auxiliary_loss_clip": 0.01185362, + "auxiliary_loss_mlp": 0.01022883, + "balance_loss_clip": 1.04175174, + "balance_loss_mlp": 1.015504, + "epoch": 0.8707990140082967, + "flos": 14611513968000.0, + "grad_norm": 1.7768712250715935, + "language_loss": 0.67326951, + "learning_rate": 1.7244468959102054e-07, + "loss": 0.69535196, + "num_input_tokens_seen": 156695245, + "step": 7242, + "time_per_iteration": 2.5804386138916016 + }, + { + "auxiliary_loss_clip": 0.01177181, + "auxiliary_loss_mlp": 0.01023266, + "balance_loss_clip": 1.04833579, + "balance_loss_mlp": 1.01558852, + "epoch": 0.8709192568989359, + "flos": 20084156265600.0, + "grad_norm": 2.0617273992233627, + "language_loss": 0.85104734, + "learning_rate": 1.7212839847058348e-07, + "loss": 0.87305182, + "num_input_tokens_seen": 156710375, + "step": 7243, + "time_per_iteration": 2.5976903438568115 + }, + { + "auxiliary_loss_clip": 0.01218897, + "auxiliary_loss_mlp": 0.01027213, + "balance_loss_clip": 1.03972769, + "balance_loss_mlp": 1.01984882, + "epoch": 0.871039499789575, + "flos": 16727083251840.0, + "grad_norm": 1.8666565536841477, + "language_loss": 0.73778284, + "learning_rate": 1.718123846368147e-07, + "loss": 0.76024389, + "num_input_tokens_seen": 156729420, + "step": 7244, + "time_per_iteration": 2.7136552333831787 + }, + { + "auxiliary_loss_clip": 0.01184894, + "auxiliary_loss_mlp": 0.0137821, + "balance_loss_clip": 1.0458914, + "balance_loss_mlp": 1.00010741, + "epoch": 0.871159742680214, + "flos": 21068790860160.0, + "grad_norm": 2.0358265782608775, + "language_loss": 0.71463394, + "learning_rate": 1.714966481376543e-07, + "loss": 0.74026501, + "num_input_tokens_seen": 156746100, + "step": 7245, + "time_per_iteration": 2.721162796020508 + }, + { + "auxiliary_loss_clip": 0.01175067, + "auxiliary_loss_mlp": 0.01025979, + "balance_loss_clip": 1.0446254, + "balance_loss_mlp": 1.01877594, + "epoch": 0.8712799855708532, + "flos": 28256526731520.0, + "grad_norm": 1.901165979829582, + "language_loss": 0.8299588, + "learning_rate": 1.7118118902099797e-07, + "loss": 0.85196924, + "num_input_tokens_seen": 156764185, + "step": 7246, + "time_per_iteration": 2.6416308879852295 + }, + { + "auxiliary_loss_clip": 0.01177379, + "auxiliary_loss_mlp": 0.01026034, + "balance_loss_clip": 1.04567623, + "balance_loss_mlp": 1.01866937, + "epoch": 0.8714002284614922, + "flos": 22236677665920.0, + "grad_norm": 1.5432724390862884, + "language_loss": 0.80494356, + "learning_rate": 1.7086600733470146e-07, + "loss": 0.82697773, + "num_input_tokens_seen": 156784855, + "step": 7247, + "time_per_iteration": 2.616748094558716 + }, + { + "auxiliary_loss_clip": 0.01173045, + "auxiliary_loss_mlp": 0.01027422, + "balance_loss_clip": 1.04495537, + "balance_loss_mlp": 1.02004623, + "epoch": 0.8715204713521313, + "flos": 21431919404160.0, + "grad_norm": 1.8678763191243803, + "language_loss": 0.76728272, + "learning_rate": 1.7055110312657738e-07, + "loss": 0.78928739, + "num_input_tokens_seen": 156804350, + "step": 7248, + "time_per_iteration": 2.5978589057922363 + }, + { + "auxiliary_loss_clip": 0.01175997, + "auxiliary_loss_mlp": 0.01026969, + "balance_loss_clip": 1.04184461, + "balance_loss_mlp": 1.01948845, + "epoch": 0.8716407142427703, + "flos": 23440439180160.0, + "grad_norm": 2.428778985467845, + "language_loss": 0.74150157, + "learning_rate": 1.702364764443962e-07, + "loss": 0.76353127, + "num_input_tokens_seen": 156823425, + "step": 7249, + "time_per_iteration": 2.673776388168335 + }, + { + "auxiliary_loss_clip": 0.01204113, + "auxiliary_loss_mlp": 0.01027047, + "balance_loss_clip": 1.03624105, + "balance_loss_mlp": 1.0197041, + "epoch": 0.8717609571334095, + "flos": 27958683156480.0, + "grad_norm": 1.9697741239818682, + "language_loss": 0.72833532, + "learning_rate": 1.6992212733588685e-07, + "loss": 0.75064695, + "num_input_tokens_seen": 156843090, + "step": 7250, + "time_per_iteration": 2.7823479175567627 + }, + { + "auxiliary_loss_clip": 0.01178721, + "auxiliary_loss_mlp": 0.01023596, + "balance_loss_clip": 1.04208732, + "balance_loss_mlp": 1.01620197, + "epoch": 0.8718812000240486, + "flos": 25479482538240.0, + "grad_norm": 1.7392214696981876, + "language_loss": 0.75336123, + "learning_rate": 1.6960805584873538e-07, + "loss": 0.77538437, + "num_input_tokens_seen": 156861090, + "step": 7251, + "time_per_iteration": 2.694648027420044 + }, + { + "auxiliary_loss_clip": 0.01200928, + "auxiliary_loss_mlp": 0.01020558, + "balance_loss_clip": 1.0381912, + "balance_loss_mlp": 1.01362562, + "epoch": 0.8720014429146876, + "flos": 23403056100480.0, + "grad_norm": 1.7998469759034141, + "language_loss": 0.78238845, + "learning_rate": 1.6929426203058684e-07, + "loss": 0.8046034, + "num_input_tokens_seen": 156881515, + "step": 7252, + "time_per_iteration": 2.738595724105835 + }, + { + "auxiliary_loss_clip": 0.01170946, + "auxiliary_loss_mlp": 0.01379233, + "balance_loss_clip": 1.04644096, + "balance_loss_mlp": 1.00015628, + "epoch": 0.8721216858053268, + "flos": 24352821567360.0, + "grad_norm": 2.0620221804564336, + "language_loss": 0.79835784, + "learning_rate": 1.689807459290431e-07, + "loss": 0.82385963, + "num_input_tokens_seen": 156900170, + "step": 7253, + "time_per_iteration": 2.59834623336792 + }, + { + "auxiliary_loss_clip": 0.01186677, + "auxiliary_loss_mlp": 0.01021032, + "balance_loss_clip": 1.04637158, + "balance_loss_mlp": 1.01387668, + "epoch": 0.8722419286959658, + "flos": 33869687034240.0, + "grad_norm": 2.4781741248319125, + "language_loss": 0.7059375, + "learning_rate": 1.6866750759166437e-07, + "loss": 0.72801453, + "num_input_tokens_seen": 156920150, + "step": 7254, + "time_per_iteration": 2.728612184524536 + }, + { + "auxiliary_loss_clip": 0.01183731, + "auxiliary_loss_mlp": 0.01023344, + "balance_loss_clip": 1.03910661, + "balance_loss_mlp": 1.01604211, + "epoch": 0.8723621715866049, + "flos": 18369385914240.0, + "grad_norm": 3.661286677079552, + "language_loss": 0.77474397, + "learning_rate": 1.6835454706596865e-07, + "loss": 0.79681474, + "num_input_tokens_seen": 156937980, + "step": 7255, + "time_per_iteration": 3.568647623062134 + }, + { + "auxiliary_loss_clip": 0.01168492, + "auxiliary_loss_mlp": 0.01029584, + "balance_loss_clip": 1.0484376, + "balance_loss_mlp": 1.02196085, + "epoch": 0.8724824144772441, + "flos": 22013348855040.0, + "grad_norm": 2.3819375172414166, + "language_loss": 0.73873162, + "learning_rate": 1.680418643994317e-07, + "loss": 0.76071239, + "num_input_tokens_seen": 156956550, + "step": 7256, + "time_per_iteration": 2.7810885906219482 + }, + { + "auxiliary_loss_clip": 0.0105996, + "auxiliary_loss_mlp": 0.01001302, + "balance_loss_clip": 1.00889754, + "balance_loss_mlp": 1.00027049, + "epoch": 0.8726026573678831, + "flos": 66698720213760.0, + "grad_norm": 0.8906428709249403, + "language_loss": 0.64554757, + "learning_rate": 1.6772945963948738e-07, + "loss": 0.66616023, + "num_input_tokens_seen": 157014715, + "step": 7257, + "time_per_iteration": 4.0464723110198975 + }, + { + "auxiliary_loss_clip": 0.01184063, + "auxiliary_loss_mlp": 0.01026776, + "balance_loss_clip": 1.04570162, + "balance_loss_mlp": 1.01901555, + "epoch": 0.8727229002585222, + "flos": 13370908078080.0, + "grad_norm": 2.1741765149821455, + "language_loss": 0.77956998, + "learning_rate": 1.6741733283352733e-07, + "loss": 0.80167836, + "num_input_tokens_seen": 157032320, + "step": 7258, + "time_per_iteration": 2.6003270149230957 + }, + { + "auxiliary_loss_clip": 0.01210067, + "auxiliary_loss_mlp": 0.01026333, + "balance_loss_clip": 1.04368961, + "balance_loss_mlp": 1.01834631, + "epoch": 0.8728431431491613, + "flos": 21796987282560.0, + "grad_norm": 1.6084600680359764, + "language_loss": 0.83878195, + "learning_rate": 1.6710548402890102e-07, + "loss": 0.86114597, + "num_input_tokens_seen": 157052845, + "step": 7259, + "time_per_iteration": 2.741001844406128 + }, + { + "auxiliary_loss_clip": 0.0117164, + "auxiliary_loss_mlp": 0.01024509, + "balance_loss_clip": 1.04731345, + "balance_loss_mlp": 1.01672494, + "epoch": 0.8729633860398004, + "flos": 36173823742080.0, + "grad_norm": 4.823105408429739, + "language_loss": 0.66787946, + "learning_rate": 1.6679391327291527e-07, + "loss": 0.68984103, + "num_input_tokens_seen": 157074050, + "step": 7260, + "time_per_iteration": 3.5702385902404785 + }, + { + "auxiliary_loss_clip": 0.01183907, + "auxiliary_loss_mlp": 0.01027393, + "balance_loss_clip": 1.04175878, + "balance_loss_mlp": 1.02099991, + "epoch": 0.8730836289304394, + "flos": 16359680989440.0, + "grad_norm": 3.1011437458631286, + "language_loss": 0.68349755, + "learning_rate": 1.6648262061283492e-07, + "loss": 0.70561057, + "num_input_tokens_seen": 157089350, + "step": 7261, + "time_per_iteration": 2.605816602706909 + }, + { + "auxiliary_loss_clip": 0.01191889, + "auxiliary_loss_mlp": 0.01025709, + "balance_loss_clip": 1.03907728, + "balance_loss_mlp": 1.01825857, + "epoch": 0.8732038718210786, + "flos": 21215126868480.0, + "grad_norm": 2.3392790833299015, + "language_loss": 0.73369104, + "learning_rate": 1.6617160609588353e-07, + "loss": 0.755867, + "num_input_tokens_seen": 157108525, + "step": 7262, + "time_per_iteration": 2.7223169803619385 + }, + { + "auxiliary_loss_clip": 0.01188554, + "auxiliary_loss_mlp": 0.01023042, + "balance_loss_clip": 1.04365969, + "balance_loss_mlp": 1.01549935, + "epoch": 0.8733241147117177, + "flos": 16610696208000.0, + "grad_norm": 1.9506832879796079, + "language_loss": 0.72102022, + "learning_rate": 1.6586086976924163e-07, + "loss": 0.74313617, + "num_input_tokens_seen": 157124025, + "step": 7263, + "time_per_iteration": 2.5950417518615723 + }, + { + "auxiliary_loss_clip": 0.0117566, + "auxiliary_loss_mlp": 0.01025253, + "balance_loss_clip": 1.04452467, + "balance_loss_mlp": 1.01784408, + "epoch": 0.8734443576023567, + "flos": 20193935207040.0, + "grad_norm": 1.8793339987046458, + "language_loss": 0.78378791, + "learning_rate": 1.6555041168004747e-07, + "loss": 0.8057971, + "num_input_tokens_seen": 157143345, + "step": 7264, + "time_per_iteration": 3.51594614982605 + }, + { + "auxiliary_loss_clip": 0.01180402, + "auxiliary_loss_mlp": 0.01028809, + "balance_loss_clip": 1.04408574, + "balance_loss_mlp": 1.02158451, + "epoch": 0.8735646004929959, + "flos": 18041162411520.0, + "grad_norm": 1.8201651722819785, + "language_loss": 0.68629569, + "learning_rate": 1.6524023187539715e-07, + "loss": 0.70838785, + "num_input_tokens_seen": 157161630, + "step": 7265, + "time_per_iteration": 2.615610122680664 + }, + { + "auxiliary_loss_clip": 0.01182089, + "auxiliary_loss_mlp": 0.01025356, + "balance_loss_clip": 1.04270053, + "balance_loss_mlp": 1.01784003, + "epoch": 0.873684843383635, + "flos": 20262344659200.0, + "grad_norm": 3.216044947189598, + "language_loss": 0.74739426, + "learning_rate": 1.649303304023446e-07, + "loss": 0.76946872, + "num_input_tokens_seen": 157181385, + "step": 7266, + "time_per_iteration": 2.753169536590576 + }, + { + "auxiliary_loss_clip": 0.01189837, + "auxiliary_loss_mlp": 0.01022585, + "balance_loss_clip": 1.04556394, + "balance_loss_mlp": 1.01572692, + "epoch": 0.873805086274274, + "flos": 16947287579520.0, + "grad_norm": 1.9471834462554252, + "language_loss": 0.78825724, + "learning_rate": 1.6462070730790246e-07, + "loss": 0.81038147, + "num_input_tokens_seen": 157200545, + "step": 7267, + "time_per_iteration": 2.7460689544677734 + }, + { + "auxiliary_loss_clip": 0.01177155, + "auxiliary_loss_mlp": 0.01025877, + "balance_loss_clip": 1.03982806, + "balance_loss_mlp": 1.01885879, + "epoch": 0.8739253291649132, + "flos": 18041270152320.0, + "grad_norm": 2.4960818197020758, + "language_loss": 0.78577811, + "learning_rate": 1.6431136263903912e-07, + "loss": 0.80780846, + "num_input_tokens_seen": 157219545, + "step": 7268, + "time_per_iteration": 2.6041836738586426 + }, + { + "auxiliary_loss_clip": 0.01177017, + "auxiliary_loss_mlp": 0.01378394, + "balance_loss_clip": 1.04356086, + "balance_loss_mlp": 1.00011444, + "epoch": 0.8740455720555522, + "flos": 21325085377920.0, + "grad_norm": 2.0687566999760985, + "language_loss": 0.73655128, + "learning_rate": 1.6400229644268282e-07, + "loss": 0.76210541, + "num_input_tokens_seen": 157237900, + "step": 7269, + "time_per_iteration": 2.6104671955108643 + }, + { + "auxiliary_loss_clip": 0.01186424, + "auxiliary_loss_mlp": 0.01028459, + "balance_loss_clip": 1.04334664, + "balance_loss_mlp": 1.0213306, + "epoch": 0.8741658149461913, + "flos": 15158684822400.0, + "grad_norm": 2.6737009616422354, + "language_loss": 0.81248099, + "learning_rate": 1.6369350876571852e-07, + "loss": 0.83462983, + "num_input_tokens_seen": 157256055, + "step": 7270, + "time_per_iteration": 2.6680099964141846 + }, + { + "auxiliary_loss_clip": 0.01198585, + "auxiliary_loss_mlp": 0.01024415, + "balance_loss_clip": 1.04030573, + "balance_loss_mlp": 1.01734018, + "epoch": 0.8742860578368304, + "flos": 23039855729280.0, + "grad_norm": 2.4224857038113843, + "language_loss": 0.82017183, + "learning_rate": 1.6338499965498874e-07, + "loss": 0.84240186, + "num_input_tokens_seen": 157274785, + "step": 7271, + "time_per_iteration": 2.818089723587036 + }, + { + "auxiliary_loss_clip": 0.01184834, + "auxiliary_loss_mlp": 0.01022072, + "balance_loss_clip": 1.04065633, + "balance_loss_mlp": 1.01463056, + "epoch": 0.8744063007274695, + "flos": 28145347159680.0, + "grad_norm": 1.5385785411750326, + "language_loss": 0.77555221, + "learning_rate": 1.630767691572943e-07, + "loss": 0.79762125, + "num_input_tokens_seen": 157294805, + "step": 7272, + "time_per_iteration": 2.7120749950408936 + }, + { + "auxiliary_loss_clip": 0.01088169, + "auxiliary_loss_mlp": 0.01002408, + "balance_loss_clip": 1.00979018, + "balance_loss_mlp": 1.00131691, + "epoch": 0.8745265436181086, + "flos": 64034076654720.0, + "grad_norm": 0.7501098454935762, + "language_loss": 0.53513438, + "learning_rate": 1.6276881731939306e-07, + "loss": 0.55604011, + "num_input_tokens_seen": 157356695, + "step": 7273, + "time_per_iteration": 3.268181800842285 + }, + { + "auxiliary_loss_clip": 0.01174879, + "auxiliary_loss_mlp": 0.01025645, + "balance_loss_clip": 1.047189, + "balance_loss_mlp": 1.01893604, + "epoch": 0.8746467865087477, + "flos": 28658618553600.0, + "grad_norm": 11.274905810868464, + "language_loss": 0.75219381, + "learning_rate": 1.6246114418800193e-07, + "loss": 0.77419907, + "num_input_tokens_seen": 157376975, + "step": 7274, + "time_per_iteration": 2.662719249725342 + }, + { + "auxiliary_loss_clip": 0.01167943, + "auxiliary_loss_mlp": 0.0102745, + "balance_loss_clip": 1.04386437, + "balance_loss_mlp": 1.02006531, + "epoch": 0.8747670293993868, + "flos": 23985850268160.0, + "grad_norm": 1.6509274907006173, + "language_loss": 0.76657265, + "learning_rate": 1.6215374980979423e-07, + "loss": 0.78852665, + "num_input_tokens_seen": 157397385, + "step": 7275, + "time_per_iteration": 2.6605656147003174 + }, + { + "auxiliary_loss_clip": 0.01171005, + "auxiliary_loss_mlp": 0.01025041, + "balance_loss_clip": 1.0457077, + "balance_loss_mlp": 1.01804376, + "epoch": 0.8748872722900258, + "flos": 45221624478720.0, + "grad_norm": 2.007360980849209, + "language_loss": 0.69120038, + "learning_rate": 1.6184663423140133e-07, + "loss": 0.71316081, + "num_input_tokens_seen": 157417685, + "step": 7276, + "time_per_iteration": 2.7882163524627686 + }, + { + "auxiliary_loss_clip": 0.01204481, + "auxiliary_loss_mlp": 0.01024785, + "balance_loss_clip": 1.04185009, + "balance_loss_mlp": 1.01753652, + "epoch": 0.875007515180665, + "flos": 19754280737280.0, + "grad_norm": 1.8964052667101834, + "language_loss": 0.6405822, + "learning_rate": 1.615397974994126e-07, + "loss": 0.66287488, + "num_input_tokens_seen": 157435490, + "step": 7277, + "time_per_iteration": 2.7517430782318115 + }, + { + "auxiliary_loss_clip": 0.01165231, + "auxiliary_loss_mlp": 0.0102475, + "balance_loss_clip": 1.04655457, + "balance_loss_mlp": 1.01812243, + "epoch": 0.875127758071304, + "flos": 22710734386560.0, + "grad_norm": 1.4917410432073963, + "language_loss": 0.80819875, + "learning_rate": 1.6123323966037438e-07, + "loss": 0.83009851, + "num_input_tokens_seen": 157454010, + "step": 7278, + "time_per_iteration": 2.601825714111328 + }, + { + "auxiliary_loss_clip": 0.01166574, + "auxiliary_loss_mlp": 0.01025215, + "balance_loss_clip": 1.04735768, + "balance_loss_mlp": 1.01814342, + "epoch": 0.8752480009619431, + "flos": 23403846199680.0, + "grad_norm": 2.1624010069253594, + "language_loss": 0.78415495, + "learning_rate": 1.6092696076079216e-07, + "loss": 0.80607283, + "num_input_tokens_seen": 157472385, + "step": 7279, + "time_per_iteration": 2.6425113677978516 + }, + { + "auxiliary_loss_clip": 0.01180465, + "auxiliary_loss_mlp": 0.01023114, + "balance_loss_clip": 1.0410049, + "balance_loss_mlp": 1.01618838, + "epoch": 0.8753682438525822, + "flos": 26213101914240.0, + "grad_norm": 1.849009227522498, + "language_loss": 0.73459101, + "learning_rate": 1.6062096084712785e-07, + "loss": 0.75662684, + "num_input_tokens_seen": 157493735, + "step": 7280, + "time_per_iteration": 2.7133419513702393 + }, + { + "auxiliary_loss_clip": 0.0117345, + "auxiliary_loss_mlp": 0.01378709, + "balance_loss_clip": 1.04043519, + "balance_loss_mlp": 1.00019145, + "epoch": 0.8754884867432213, + "flos": 23326745656320.0, + "grad_norm": 1.794536776195457, + "language_loss": 0.70591855, + "learning_rate": 1.6031523996580098e-07, + "loss": 0.73144007, + "num_input_tokens_seen": 157511295, + "step": 7281, + "time_per_iteration": 3.6714658737182617 + }, + { + "auxiliary_loss_clip": 0.01200702, + "auxiliary_loss_mlp": 0.01026438, + "balance_loss_clip": 1.0434773, + "balance_loss_mlp": 1.01884389, + "epoch": 0.8756087296338604, + "flos": 12495226412160.0, + "grad_norm": 2.276959398900431, + "language_loss": 0.6614064, + "learning_rate": 1.6000979816318981e-07, + "loss": 0.68367779, + "num_input_tokens_seen": 157529760, + "step": 7282, + "time_per_iteration": 2.6784303188323975 + }, + { + "auxiliary_loss_clip": 0.0117419, + "auxiliary_loss_mlp": 0.01026478, + "balance_loss_clip": 1.04760981, + "balance_loss_mlp": 1.019454, + "epoch": 0.8757289725244994, + "flos": 18952898353920.0, + "grad_norm": 2.161380584267379, + "language_loss": 0.74875665, + "learning_rate": 1.5970463548562886e-07, + "loss": 0.77076328, + "num_input_tokens_seen": 157548915, + "step": 7283, + "time_per_iteration": 2.564711570739746 + }, + { + "auxiliary_loss_clip": 0.01182111, + "auxiliary_loss_mlp": 0.01027955, + "balance_loss_clip": 1.04454482, + "balance_loss_mlp": 1.02071905, + "epoch": 0.8758492154151386, + "flos": 25265958140160.0, + "grad_norm": 2.4010615055570987, + "language_loss": 0.71167254, + "learning_rate": 1.5939975197941192e-07, + "loss": 0.73377317, + "num_input_tokens_seen": 157570570, + "step": 7284, + "time_per_iteration": 3.6236226558685303 + }, + { + "auxiliary_loss_clip": 0.01088038, + "auxiliary_loss_mlp": 0.01001543, + "balance_loss_clip": 1.00967669, + "balance_loss_mlp": 1.00041044, + "epoch": 0.8759694583057777, + "flos": 65571664193280.0, + "grad_norm": 0.8051173783340823, + "language_loss": 0.53379309, + "learning_rate": 1.5909514769078892e-07, + "loss": 0.55468893, + "num_input_tokens_seen": 157635675, + "step": 7285, + "time_per_iteration": 3.307117462158203 + }, + { + "auxiliary_loss_clip": 0.01184141, + "auxiliary_loss_mlp": 0.01023587, + "balance_loss_clip": 1.04406762, + "balance_loss_mlp": 1.01654816, + "epoch": 0.8760897011964167, + "flos": 25446193608960.0, + "grad_norm": 1.5432720584543427, + "language_loss": 0.77695161, + "learning_rate": 1.5879082266596867e-07, + "loss": 0.79902887, + "num_input_tokens_seen": 157657015, + "step": 7286, + "time_per_iteration": 3.656680107116699 + }, + { + "auxiliary_loss_clip": 0.01175902, + "auxiliary_loss_mlp": 0.01025188, + "balance_loss_clip": 1.03913522, + "balance_loss_mlp": 1.0181998, + "epoch": 0.8762099440870559, + "flos": 28984830894720.0, + "grad_norm": 1.8782811933854955, + "language_loss": 0.71664369, + "learning_rate": 1.5848677695111645e-07, + "loss": 0.73865461, + "num_input_tokens_seen": 157678615, + "step": 7287, + "time_per_iteration": 2.685652732849121 + }, + { + "auxiliary_loss_clip": 0.01201717, + "auxiliary_loss_mlp": 0.01022553, + "balance_loss_clip": 1.04555595, + "balance_loss_mlp": 1.01503074, + "epoch": 0.8763301869776949, + "flos": 21609461352960.0, + "grad_norm": 2.406395541540044, + "language_loss": 0.69967717, + "learning_rate": 1.5818301059235562e-07, + "loss": 0.72191989, + "num_input_tokens_seen": 157693790, + "step": 7288, + "time_per_iteration": 2.6350317001342773 + }, + { + "auxiliary_loss_clip": 0.01187587, + "auxiliary_loss_mlp": 0.01029044, + "balance_loss_clip": 1.04620695, + "balance_loss_mlp": 1.02123606, + "epoch": 0.876450429868334, + "flos": 24644416176000.0, + "grad_norm": 1.6797937151670361, + "language_loss": 0.81560308, + "learning_rate": 1.578795236357684e-07, + "loss": 0.83776939, + "num_input_tokens_seen": 157715255, + "step": 7289, + "time_per_iteration": 2.674849271774292 + }, + { + "auxiliary_loss_clip": 0.01187029, + "auxiliary_loss_mlp": 0.01026797, + "balance_loss_clip": 1.04578948, + "balance_loss_mlp": 1.01946521, + "epoch": 0.8765706727589732, + "flos": 20260046188800.0, + "grad_norm": 2.0359319568810363, + "language_loss": 0.85601318, + "learning_rate": 1.5757631612739218e-07, + "loss": 0.87815142, + "num_input_tokens_seen": 157728800, + "step": 7290, + "time_per_iteration": 3.559126138687134 + }, + { + "auxiliary_loss_clip": 0.01059997, + "auxiliary_loss_mlp": 0.01001207, + "balance_loss_clip": 1.00880277, + "balance_loss_mlp": 1.00018775, + "epoch": 0.8766909156496122, + "flos": 71371165276800.0, + "grad_norm": 0.7789036985728506, + "language_loss": 0.61431265, + "learning_rate": 1.572733881132242e-07, + "loss": 0.63492465, + "num_input_tokens_seen": 157789445, + "step": 7291, + "time_per_iteration": 3.2181265354156494 + }, + { + "auxiliary_loss_clip": 0.01100496, + "auxiliary_loss_mlp": 0.01002053, + "balance_loss_clip": 1.01672757, + "balance_loss_mlp": 1.00099206, + "epoch": 0.8768111585402513, + "flos": 69523490603520.0, + "grad_norm": 0.780428475381197, + "language_loss": 0.58514804, + "learning_rate": 1.5697073963921814e-07, + "loss": 0.60617357, + "num_input_tokens_seen": 157848685, + "step": 7292, + "time_per_iteration": 3.172748565673828 + }, + { + "auxiliary_loss_clip": 0.01178052, + "auxiliary_loss_mlp": 0.01025156, + "balance_loss_clip": 1.04798353, + "balance_loss_mlp": 1.01731551, + "epoch": 0.8769314014308904, + "flos": 18838558385280.0, + "grad_norm": 2.399172552361193, + "language_loss": 0.84889913, + "learning_rate": 1.566683707512857e-07, + "loss": 0.87093121, + "num_input_tokens_seen": 157866360, + "step": 7293, + "time_per_iteration": 2.5842695236206055 + }, + { + "auxiliary_loss_clip": 0.01178991, + "auxiliary_loss_mlp": 0.01027488, + "balance_loss_clip": 1.04199934, + "balance_loss_mlp": 1.02020717, + "epoch": 0.8770516443215295, + "flos": 14976402278400.0, + "grad_norm": 1.9523232515932754, + "language_loss": 0.79220265, + "learning_rate": 1.5636628149529553e-07, + "loss": 0.8142674, + "num_input_tokens_seen": 157884150, + "step": 7294, + "time_per_iteration": 2.615009307861328 + }, + { + "auxiliary_loss_clip": 0.01182319, + "auxiliary_loss_mlp": 0.01024509, + "balance_loss_clip": 1.04270756, + "balance_loss_mlp": 1.01746964, + "epoch": 0.8771718872121685, + "flos": 31649654021760.0, + "grad_norm": 2.3031878455713115, + "language_loss": 0.79527259, + "learning_rate": 1.560644719170743e-07, + "loss": 0.81734085, + "num_input_tokens_seen": 157905020, + "step": 7295, + "time_per_iteration": 2.704932689666748 + }, + { + "auxiliary_loss_clip": 0.01188718, + "auxiliary_loss_mlp": 0.01031528, + "balance_loss_clip": 1.04172206, + "balance_loss_mlp": 1.02343416, + "epoch": 0.8772921301028077, + "flos": 36095466222720.0, + "grad_norm": 1.8245095018789097, + "language_loss": 0.71908259, + "learning_rate": 1.5576294206240692e-07, + "loss": 0.74128503, + "num_input_tokens_seen": 157924545, + "step": 7296, + "time_per_iteration": 2.798795461654663 + }, + { + "auxiliary_loss_clip": 0.01176709, + "auxiliary_loss_mlp": 0.01026812, + "balance_loss_clip": 1.04165089, + "balance_loss_mlp": 1.01943564, + "epoch": 0.8774123729934468, + "flos": 57116961849600.0, + "grad_norm": 4.509222522097547, + "language_loss": 0.67507005, + "learning_rate": 1.5546169197703507e-07, + "loss": 0.69710517, + "num_input_tokens_seen": 157950820, + "step": 7297, + "time_per_iteration": 2.9733803272247314 + }, + { + "auxiliary_loss_clip": 0.01185937, + "auxiliary_loss_mlp": 0.01026721, + "balance_loss_clip": 1.0408926, + "balance_loss_mlp": 1.01925588, + "epoch": 0.8775326158840858, + "flos": 23914495900800.0, + "grad_norm": 4.055278674691733, + "language_loss": 0.77601725, + "learning_rate": 1.5516072170665774e-07, + "loss": 0.79814386, + "num_input_tokens_seen": 157968790, + "step": 7298, + "time_per_iteration": 2.653512477874756 + }, + { + "auxiliary_loss_clip": 0.01177173, + "auxiliary_loss_mlp": 0.01025058, + "balance_loss_clip": 1.04661703, + "balance_loss_mlp": 1.01812279, + "epoch": 0.877652858774725, + "flos": 17123285243520.0, + "grad_norm": 1.8145308006500576, + "language_loss": 0.87242329, + "learning_rate": 1.5486003129693214e-07, + "loss": 0.89444566, + "num_input_tokens_seen": 157986155, + "step": 7299, + "time_per_iteration": 2.6503355503082275 + }, + { + "auxiliary_loss_clip": 0.01179784, + "auxiliary_loss_mlp": 0.01026241, + "balance_loss_clip": 1.0461154, + "balance_loss_mlp": 1.01919007, + "epoch": 0.877773101665364, + "flos": 16508961912960.0, + "grad_norm": 1.9970586344038073, + "language_loss": 0.77971959, + "learning_rate": 1.545596207934725e-07, + "loss": 0.80177981, + "num_input_tokens_seen": 158004640, + "step": 7300, + "time_per_iteration": 2.6905453205108643 + }, + { + "auxiliary_loss_clip": 0.01176186, + "auxiliary_loss_mlp": 0.01020855, + "balance_loss_clip": 1.0421555, + "balance_loss_mlp": 1.01447392, + "epoch": 0.8778933445560031, + "flos": 22053209973120.0, + "grad_norm": 1.7032419177740334, + "language_loss": 0.77868086, + "learning_rate": 1.5425949024185147e-07, + "loss": 0.80065119, + "num_input_tokens_seen": 158024665, + "step": 7301, + "time_per_iteration": 2.685203790664673 + }, + { + "auxiliary_loss_clip": 0.01182754, + "auxiliary_loss_mlp": 0.0102589, + "balance_loss_clip": 1.04078388, + "balance_loss_mlp": 1.01885056, + "epoch": 0.8780135874466423, + "flos": 22564757514240.0, + "grad_norm": 1.8780917441612075, + "language_loss": 0.67709374, + "learning_rate": 1.5395963968759818e-07, + "loss": 0.69918019, + "num_input_tokens_seen": 158044940, + "step": 7302, + "time_per_iteration": 2.597313642501831 + }, + { + "auxiliary_loss_clip": 0.01182639, + "auxiliary_loss_mlp": 0.01027719, + "balance_loss_clip": 1.04118896, + "balance_loss_mlp": 1.02039671, + "epoch": 0.8781338303372813, + "flos": 61531999073280.0, + "grad_norm": 1.4628255408634392, + "language_loss": 0.64382982, + "learning_rate": 1.536600691761998e-07, + "loss": 0.66593337, + "num_input_tokens_seen": 158070770, + "step": 7303, + "time_per_iteration": 2.966463088989258 + }, + { + "auxiliary_loss_clip": 0.01197932, + "auxiliary_loss_mlp": 0.01025046, + "balance_loss_clip": 1.04635477, + "balance_loss_mlp": 1.01762795, + "epoch": 0.8782540732279204, + "flos": 22674751937280.0, + "grad_norm": 1.982664390351109, + "language_loss": 0.71634746, + "learning_rate": 1.5336077875310084e-07, + "loss": 0.73857725, + "num_input_tokens_seen": 158089995, + "step": 7304, + "time_per_iteration": 2.673495054244995 + }, + { + "auxiliary_loss_clip": 0.01207274, + "auxiliary_loss_mlp": 0.01029377, + "balance_loss_clip": 1.04194093, + "balance_loss_mlp": 1.02228403, + "epoch": 0.8783743161185595, + "flos": 16070348937600.0, + "grad_norm": 2.5196904930995703, + "language_loss": 0.74125087, + "learning_rate": 1.5306176846370321e-07, + "loss": 0.7636174, + "num_input_tokens_seen": 158108140, + "step": 7305, + "time_per_iteration": 2.6952719688415527 + }, + { + "auxiliary_loss_clip": 0.01192924, + "auxiliary_loss_mlp": 0.01025811, + "balance_loss_clip": 1.04443121, + "balance_loss_mlp": 1.01827133, + "epoch": 0.8784945590091986, + "flos": 26067879227520.0, + "grad_norm": 2.559105616578526, + "language_loss": 0.73835021, + "learning_rate": 1.5276303835336712e-07, + "loss": 0.76053751, + "num_input_tokens_seen": 158128680, + "step": 7306, + "time_per_iteration": 2.6750681400299072 + }, + { + "auxiliary_loss_clip": 0.01073789, + "auxiliary_loss_mlp": 0.01000267, + "balance_loss_clip": 1.00866866, + "balance_loss_mlp": 0.99920028, + "epoch": 0.8786148018998376, + "flos": 62720643939840.0, + "grad_norm": 0.7721219409084142, + "language_loss": 0.53565735, + "learning_rate": 1.524645884674094e-07, + "loss": 0.55639791, + "num_input_tokens_seen": 158185610, + "step": 7307, + "time_per_iteration": 3.176966428756714 + }, + { + "auxiliary_loss_clip": 0.0116894, + "auxiliary_loss_mlp": 0.01379127, + "balance_loss_clip": 1.04684472, + "balance_loss_mlp": 1.00010777, + "epoch": 0.8787350447904768, + "flos": 21652734263040.0, + "grad_norm": 3.478730533930734, + "language_loss": 0.79383743, + "learning_rate": 1.521664188511047e-07, + "loss": 0.81931812, + "num_input_tokens_seen": 158205635, + "step": 7308, + "time_per_iteration": 3.4768598079681396 + }, + { + "auxiliary_loss_clip": 0.01186378, + "auxiliary_loss_mlp": 0.01378559, + "balance_loss_clip": 1.04786885, + "balance_loss_mlp": 1.00014591, + "epoch": 0.8788552876811159, + "flos": 25478476957440.0, + "grad_norm": 2.2149231498487025, + "language_loss": 0.80344141, + "learning_rate": 1.518685295496851e-07, + "loss": 0.82909083, + "num_input_tokens_seen": 158223495, + "step": 7309, + "time_per_iteration": 3.6622462272644043 + }, + { + "auxiliary_loss_clip": 0.01175579, + "auxiliary_loss_mlp": 0.01024142, + "balance_loss_clip": 1.04427958, + "balance_loss_mlp": 1.01704919, + "epoch": 0.8789755305717549, + "flos": 22310222762880.0, + "grad_norm": 1.7529549100973352, + "language_loss": 0.85395658, + "learning_rate": 1.5157092060833975e-07, + "loss": 0.87595385, + "num_input_tokens_seen": 158243145, + "step": 7310, + "time_per_iteration": 2.6128602027893066 + }, + { + "auxiliary_loss_clip": 0.01181514, + "auxiliary_loss_mlp": 0.01022958, + "balance_loss_clip": 1.04238808, + "balance_loss_mlp": 1.01566267, + "epoch": 0.879095773462394, + "flos": 29310971408640.0, + "grad_norm": 1.7190686574665965, + "language_loss": 0.66018867, + "learning_rate": 1.5127359207221658e-07, + "loss": 0.68223345, + "num_input_tokens_seen": 158262625, + "step": 7311, + "time_per_iteration": 2.7613584995269775 + }, + { + "auxiliary_loss_clip": 0.01195404, + "auxiliary_loss_mlp": 0.01024554, + "balance_loss_clip": 1.03585231, + "balance_loss_mlp": 1.01688337, + "epoch": 0.8792160163530331, + "flos": 16690023394560.0, + "grad_norm": 1.8773159505581314, + "language_loss": 0.73236459, + "learning_rate": 1.5097654398641923e-07, + "loss": 0.75456417, + "num_input_tokens_seen": 158280530, + "step": 7312, + "time_per_iteration": 3.6278367042541504 + }, + { + "auxiliary_loss_clip": 0.01181758, + "auxiliary_loss_mlp": 0.01024975, + "balance_loss_clip": 1.04719949, + "balance_loss_mlp": 1.01810288, + "epoch": 0.8793362592436722, + "flos": 24499301230080.0, + "grad_norm": 1.4977694681521854, + "language_loss": 0.73263186, + "learning_rate": 1.5067977639601014e-07, + "loss": 0.75469923, + "num_input_tokens_seen": 158303290, + "step": 7313, + "time_per_iteration": 2.6734817028045654 + }, + { + "auxiliary_loss_clip": 0.01185582, + "auxiliary_loss_mlp": 0.01023063, + "balance_loss_clip": 1.04691005, + "balance_loss_mlp": 1.01609766, + "epoch": 0.8794565021343113, + "flos": 14538399834240.0, + "grad_norm": 2.169369106052993, + "language_loss": 0.70833647, + "learning_rate": 1.5038328934600864e-07, + "loss": 0.73042291, + "num_input_tokens_seen": 158319925, + "step": 7314, + "time_per_iteration": 2.622122049331665 + }, + { + "auxiliary_loss_clip": 0.01186446, + "auxiliary_loss_mlp": 0.01024683, + "balance_loss_clip": 1.04539847, + "balance_loss_mlp": 1.01723886, + "epoch": 0.8795767450249504, + "flos": 39530286224640.0, + "grad_norm": 1.8128788689646316, + "language_loss": 0.70103031, + "learning_rate": 1.5008708288139161e-07, + "loss": 0.72314161, + "num_input_tokens_seen": 158342285, + "step": 7315, + "time_per_iteration": 2.802382469177246 + }, + { + "auxiliary_loss_clip": 0.01174859, + "auxiliary_loss_mlp": 0.0102086, + "balance_loss_clip": 1.04599833, + "balance_loss_mlp": 1.01367164, + "epoch": 0.8796969879155895, + "flos": 22960672197120.0, + "grad_norm": 1.9244724842259093, + "language_loss": 0.73517585, + "learning_rate": 1.497911570470931e-07, + "loss": 0.75713307, + "num_input_tokens_seen": 158362290, + "step": 7316, + "time_per_iteration": 3.434054136276245 + }, + { + "auxiliary_loss_clip": 0.01180276, + "auxiliary_loss_mlp": 0.01020282, + "balance_loss_clip": 1.04168916, + "balance_loss_mlp": 1.01317692, + "epoch": 0.8798172308062285, + "flos": 28362427004160.0, + "grad_norm": 1.6036236209610082, + "language_loss": 0.8550899, + "learning_rate": 1.494955118880048e-07, + "loss": 0.87709546, + "num_input_tokens_seen": 158383275, + "step": 7317, + "time_per_iteration": 2.713838577270508 + }, + { + "auxiliary_loss_clip": 0.01177405, + "auxiliary_loss_mlp": 0.01020162, + "balance_loss_clip": 1.04581642, + "balance_loss_mlp": 1.01343012, + "epoch": 0.8799374736968677, + "flos": 23988974751360.0, + "grad_norm": 1.726609644866625, + "language_loss": 0.72847372, + "learning_rate": 1.4920014744897634e-07, + "loss": 0.75044936, + "num_input_tokens_seen": 158402690, + "step": 7318, + "time_per_iteration": 2.6660258769989014 + }, + { + "auxiliary_loss_clip": 0.01174469, + "auxiliary_loss_mlp": 0.01028485, + "balance_loss_clip": 1.04184461, + "balance_loss_mlp": 1.0213685, + "epoch": 0.8800577165875068, + "flos": 25630271832960.0, + "grad_norm": 2.389418964804201, + "language_loss": 0.86436582, + "learning_rate": 1.4890506377481392e-07, + "loss": 0.88639534, + "num_input_tokens_seen": 158421780, + "step": 7319, + "time_per_iteration": 2.6656229496002197 + }, + { + "auxiliary_loss_clip": 0.01208207, + "auxiliary_loss_mlp": 0.01029219, + "balance_loss_clip": 1.03953314, + "balance_loss_mlp": 1.0224421, + "epoch": 0.8801779594781458, + "flos": 23440331439360.0, + "grad_norm": 1.4811124534323636, + "language_loss": 0.63915753, + "learning_rate": 1.486102609102815e-07, + "loss": 0.66153181, + "num_input_tokens_seen": 158442330, + "step": 7320, + "time_per_iteration": 2.7721874713897705 + }, + { + "auxiliary_loss_clip": 0.01179297, + "auxiliary_loss_mlp": 0.01023296, + "balance_loss_clip": 1.04332566, + "balance_loss_mlp": 1.01631927, + "epoch": 0.880298202368785, + "flos": 11508580656000.0, + "grad_norm": 2.4062984932993374, + "language_loss": 0.85712862, + "learning_rate": 1.483157389001004e-07, + "loss": 0.87915444, + "num_input_tokens_seen": 158459890, + "step": 7321, + "time_per_iteration": 2.6257822513580322 + }, + { + "auxiliary_loss_clip": 0.01180791, + "auxiliary_loss_mlp": 0.0102654, + "balance_loss_clip": 1.04139686, + "balance_loss_mlp": 1.01832032, + "epoch": 0.880418445259424, + "flos": 22671447886080.0, + "grad_norm": 2.6636306798468894, + "language_loss": 0.78679335, + "learning_rate": 1.4802149778894933e-07, + "loss": 0.80886662, + "num_input_tokens_seen": 158478680, + "step": 7322, + "time_per_iteration": 2.7027547359466553 + }, + { + "auxiliary_loss_clip": 0.01164569, + "auxiliary_loss_mlp": 0.01024548, + "balance_loss_clip": 1.04084945, + "balance_loss_mlp": 1.01796472, + "epoch": 0.8805386881500631, + "flos": 20522158709760.0, + "grad_norm": 1.8272067878184277, + "language_loss": 0.87414122, + "learning_rate": 1.4772753762146484e-07, + "loss": 0.89603233, + "num_input_tokens_seen": 158497935, + "step": 7323, + "time_per_iteration": 2.5885307788848877 + }, + { + "auxiliary_loss_clip": 0.01166035, + "auxiliary_loss_mlp": 0.01024401, + "balance_loss_clip": 1.04226923, + "balance_loss_mlp": 1.01704013, + "epoch": 0.8806589310407023, + "flos": 36538891620480.0, + "grad_norm": 1.6043008428756957, + "language_loss": 0.70543814, + "learning_rate": 1.474338584422401e-07, + "loss": 0.72734249, + "num_input_tokens_seen": 158523145, + "step": 7324, + "time_per_iteration": 2.7258338928222656 + }, + { + "auxiliary_loss_clip": 0.01174168, + "auxiliary_loss_mlp": 0.01025618, + "balance_loss_clip": 1.04754055, + "balance_loss_mlp": 1.01878452, + "epoch": 0.8807791739313413, + "flos": 23440187784960.0, + "grad_norm": 1.7513400393423226, + "language_loss": 0.75914848, + "learning_rate": 1.4714046029582595e-07, + "loss": 0.78114635, + "num_input_tokens_seen": 158542210, + "step": 7325, + "time_per_iteration": 2.597198724746704 + }, + { + "auxiliary_loss_clip": 0.01195373, + "auxiliary_loss_mlp": 0.01024872, + "balance_loss_clip": 1.04178131, + "balance_loss_mlp": 1.01756144, + "epoch": 0.8808994168219804, + "flos": 25956843310080.0, + "grad_norm": 1.7591864506344714, + "language_loss": 0.76050436, + "learning_rate": 1.46847343226731e-07, + "loss": 0.78270686, + "num_input_tokens_seen": 158563250, + "step": 7326, + "time_per_iteration": 2.7344555854797363 + }, + { + "auxiliary_loss_clip": 0.01178847, + "auxiliary_loss_mlp": 0.01026695, + "balance_loss_clip": 1.04527104, + "balance_loss_mlp": 1.01952195, + "epoch": 0.8810196597126195, + "flos": 17092079303040.0, + "grad_norm": 1.8507575614857172, + "language_loss": 0.69688565, + "learning_rate": 1.465545072794203e-07, + "loss": 0.71894109, + "num_input_tokens_seen": 158581125, + "step": 7327, + "time_per_iteration": 2.6359012126922607 + }, + { + "auxiliary_loss_clip": 0.01201412, + "auxiliary_loss_mlp": 0.01028632, + "balance_loss_clip": 1.04479265, + "balance_loss_mlp": 1.02122593, + "epoch": 0.8811399026032586, + "flos": 23002831785600.0, + "grad_norm": 1.5510129070234104, + "language_loss": 0.7563563, + "learning_rate": 1.4626195249831774e-07, + "loss": 0.77865672, + "num_input_tokens_seen": 158602025, + "step": 7328, + "time_per_iteration": 2.7227346897125244 + }, + { + "auxiliary_loss_clip": 0.01170974, + "auxiliary_loss_mlp": 0.01022941, + "balance_loss_clip": 1.04402995, + "balance_loss_mlp": 1.01580942, + "epoch": 0.8812601454938976, + "flos": 14463813242880.0, + "grad_norm": 1.736794347187787, + "language_loss": 0.71937591, + "learning_rate": 1.4596967892780244e-07, + "loss": 0.74131507, + "num_input_tokens_seen": 158618355, + "step": 7329, + "time_per_iteration": 2.728924512863159 + }, + { + "auxiliary_loss_clip": 0.01166933, + "auxiliary_loss_mlp": 0.01025699, + "balance_loss_clip": 1.04744124, + "balance_loss_mlp": 1.01838005, + "epoch": 0.8813803883845368, + "flos": 22493223578880.0, + "grad_norm": 1.8412010328509336, + "language_loss": 0.74690306, + "learning_rate": 1.4567768661221314e-07, + "loss": 0.76882935, + "num_input_tokens_seen": 158638925, + "step": 7330, + "time_per_iteration": 2.693150520324707 + }, + { + "auxiliary_loss_clip": 0.01180097, + "auxiliary_loss_mlp": 0.01378605, + "balance_loss_clip": 1.04715252, + "balance_loss_mlp": 1.00011301, + "epoch": 0.8815006312751759, + "flos": 21506901045120.0, + "grad_norm": 2.100329448597674, + "language_loss": 0.74432349, + "learning_rate": 1.4538597559584442e-07, + "loss": 0.76991051, + "num_input_tokens_seen": 158656715, + "step": 7331, + "time_per_iteration": 2.6743552684783936 + }, + { + "auxiliary_loss_clip": 0.01180044, + "auxiliary_loss_mlp": 0.01027627, + "balance_loss_clip": 1.04317379, + "balance_loss_mlp": 1.01980138, + "epoch": 0.8816208741658149, + "flos": 22784566792320.0, + "grad_norm": 2.1225567438489534, + "language_loss": 0.7907061, + "learning_rate": 1.4509454592294823e-07, + "loss": 0.81278276, + "num_input_tokens_seen": 158677200, + "step": 7332, + "time_per_iteration": 2.7432472705841064 + }, + { + "auxiliary_loss_clip": 0.01193785, + "auxiliary_loss_mlp": 0.01378511, + "balance_loss_clip": 1.04315817, + "balance_loss_mlp": 1.00015914, + "epoch": 0.8817411170564541, + "flos": 17779409026560.0, + "grad_norm": 4.717143089561719, + "language_loss": 0.78826433, + "learning_rate": 1.448033976377354e-07, + "loss": 0.81398737, + "num_input_tokens_seen": 158692185, + "step": 7333, + "time_per_iteration": 3.6650617122650146 + }, + { + "auxiliary_loss_clip": 0.01178337, + "auxiliary_loss_mlp": 0.01024412, + "balance_loss_clip": 1.04506135, + "balance_loss_mlp": 1.01775146, + "epoch": 0.8818613599470931, + "flos": 18551812112640.0, + "grad_norm": 1.9200295729959933, + "language_loss": 0.74360836, + "learning_rate": 1.445125307843713e-07, + "loss": 0.76563585, + "num_input_tokens_seen": 158710410, + "step": 7334, + "time_per_iteration": 2.6488420963287354 + }, + { + "auxiliary_loss_clip": 0.01173715, + "auxiliary_loss_mlp": 0.01023058, + "balance_loss_clip": 1.0455724, + "balance_loss_mlp": 1.01649213, + "epoch": 0.8819816028377322, + "flos": 27599792417280.0, + "grad_norm": 1.9543751717542275, + "language_loss": 0.75327617, + "learning_rate": 1.442219454069813e-07, + "loss": 0.77524388, + "num_input_tokens_seen": 158731435, + "step": 7335, + "time_per_iteration": 3.554194927215576 + }, + { + "auxiliary_loss_clip": 0.01203139, + "auxiliary_loss_mlp": 0.01029142, + "balance_loss_clip": 1.04032731, + "balance_loss_mlp": 1.02243686, + "epoch": 0.8821018457283714, + "flos": 23404600385280.0, + "grad_norm": 2.2673398150158377, + "language_loss": 0.66551381, + "learning_rate": 1.4393164154964676e-07, + "loss": 0.68783659, + "num_input_tokens_seen": 158750965, + "step": 7336, + "time_per_iteration": 2.7181203365325928 + }, + { + "auxiliary_loss_clip": 0.01175158, + "auxiliary_loss_mlp": 0.01021296, + "balance_loss_clip": 1.04794979, + "balance_loss_mlp": 1.01426828, + "epoch": 0.8822220886190104, + "flos": 29132459792640.0, + "grad_norm": 1.8594561402837815, + "language_loss": 0.9417156, + "learning_rate": 1.4364161925640649e-07, + "loss": 0.96368015, + "num_input_tokens_seen": 158772365, + "step": 7337, + "time_per_iteration": 3.7601704597473145 + }, + { + "auxiliary_loss_clip": 0.01165406, + "auxiliary_loss_mlp": 0.01025007, + "balance_loss_clip": 1.04729867, + "balance_loss_mlp": 1.01855755, + "epoch": 0.8823423315096495, + "flos": 20485422074880.0, + "grad_norm": 1.7498097709008966, + "language_loss": 0.85142791, + "learning_rate": 1.4335187857125663e-07, + "loss": 0.87333208, + "num_input_tokens_seen": 158791065, + "step": 7338, + "time_per_iteration": 2.5361196994781494 + }, + { + "auxiliary_loss_clip": 0.01177653, + "auxiliary_loss_mlp": 0.0101991, + "balance_loss_clip": 1.04585922, + "balance_loss_mlp": 1.01268005, + "epoch": 0.8824625744002886, + "flos": 24206377818240.0, + "grad_norm": 1.5866848193995569, + "language_loss": 0.75093627, + "learning_rate": 1.4306241953815023e-07, + "loss": 0.77291185, + "num_input_tokens_seen": 158812125, + "step": 7339, + "time_per_iteration": 2.6967074871063232 + }, + { + "auxiliary_loss_clip": 0.01174909, + "auxiliary_loss_mlp": 0.01019192, + "balance_loss_clip": 1.04351997, + "balance_loss_mlp": 1.01239395, + "epoch": 0.8825828172909277, + "flos": 24679500785280.0, + "grad_norm": 1.8033636658949819, + "language_loss": 0.71369624, + "learning_rate": 1.4277324220099862e-07, + "loss": 0.73563719, + "num_input_tokens_seen": 158834035, + "step": 7340, + "time_per_iteration": 2.6388938426971436 + }, + { + "auxiliary_loss_clip": 0.01186907, + "auxiliary_loss_mlp": 0.01024097, + "balance_loss_clip": 1.04216433, + "balance_loss_mlp": 1.01653302, + "epoch": 0.8827030601815667, + "flos": 22456163721600.0, + "grad_norm": 1.798654797191921, + "language_loss": 0.74500501, + "learning_rate": 1.4248434660366938e-07, + "loss": 0.767115, + "num_input_tokens_seen": 158853510, + "step": 7341, + "time_per_iteration": 2.7154080867767334 + }, + { + "auxiliary_loss_clip": 0.01184532, + "auxiliary_loss_mlp": 0.01026074, + "balance_loss_clip": 1.04443026, + "balance_loss_mlp": 1.01889205, + "epoch": 0.8828233030722058, + "flos": 19865639877120.0, + "grad_norm": 1.6853473795168727, + "language_loss": 0.70156407, + "learning_rate": 1.4219573278998808e-07, + "loss": 0.72367018, + "num_input_tokens_seen": 158871970, + "step": 7342, + "time_per_iteration": 3.537043571472168 + }, + { + "auxiliary_loss_clip": 0.01181726, + "auxiliary_loss_mlp": 0.01029732, + "balance_loss_clip": 1.04330456, + "balance_loss_mlp": 1.02190018, + "epoch": 0.882943545962845, + "flos": 39347213581440.0, + "grad_norm": 2.050354813434126, + "language_loss": 0.64487821, + "learning_rate": 1.4190740080373685e-07, + "loss": 0.66699278, + "num_input_tokens_seen": 158892250, + "step": 7343, + "time_per_iteration": 2.7875964641571045 + }, + { + "auxiliary_loss_clip": 0.011995, + "auxiliary_loss_mlp": 0.0102299, + "balance_loss_clip": 1.0420022, + "balance_loss_mlp": 1.0154326, + "epoch": 0.883063788853484, + "flos": 19054524908160.0, + "grad_norm": 2.9284680851240057, + "language_loss": 0.83941901, + "learning_rate": 1.4161935068865538e-07, + "loss": 0.86164391, + "num_input_tokens_seen": 158907395, + "step": 7344, + "time_per_iteration": 2.728574752807617 + }, + { + "auxiliary_loss_clip": 0.01166611, + "auxiliary_loss_mlp": 0.01025576, + "balance_loss_clip": 1.04545832, + "balance_loss_mlp": 1.01807141, + "epoch": 0.8831840317441231, + "flos": 18733196816640.0, + "grad_norm": 2.1149295087322617, + "language_loss": 0.7578243, + "learning_rate": 1.4133158248844113e-07, + "loss": 0.77974612, + "num_input_tokens_seen": 158926300, + "step": 7345, + "time_per_iteration": 2.5244994163513184 + }, + { + "auxiliary_loss_clip": 0.01195241, + "auxiliary_loss_mlp": 0.01023955, + "balance_loss_clip": 1.04200602, + "balance_loss_mlp": 1.01635253, + "epoch": 0.8833042746347622, + "flos": 26827712553600.0, + "grad_norm": 1.6994547136621878, + "language_loss": 0.7342335, + "learning_rate": 1.4104409624674785e-07, + "loss": 0.75642538, + "num_input_tokens_seen": 158946085, + "step": 7346, + "time_per_iteration": 2.7217562198638916 + }, + { + "auxiliary_loss_clip": 0.01178801, + "auxiliary_loss_mlp": 0.01023899, + "balance_loss_clip": 1.04763842, + "balance_loss_mlp": 1.01644266, + "epoch": 0.8834245175254013, + "flos": 26104077158400.0, + "grad_norm": 1.7886079263567818, + "language_loss": 0.78384566, + "learning_rate": 1.407568920071873e-07, + "loss": 0.80587268, + "num_input_tokens_seen": 158964950, + "step": 7347, + "time_per_iteration": 2.640796184539795 + }, + { + "auxiliary_loss_clip": 0.01172848, + "auxiliary_loss_mlp": 0.0102831, + "balance_loss_clip": 1.04874301, + "balance_loss_mlp": 1.02049601, + "epoch": 0.8835447604160404, + "flos": 30629036977920.0, + "grad_norm": 4.243243065808063, + "language_loss": 0.68888491, + "learning_rate": 1.4046996981332782e-07, + "loss": 0.71089649, + "num_input_tokens_seen": 158984835, + "step": 7348, + "time_per_iteration": 2.712657928466797 + }, + { + "auxiliary_loss_clip": 0.01196759, + "auxiliary_loss_mlp": 0.01024133, + "balance_loss_clip": 1.04346538, + "balance_loss_mlp": 1.01650977, + "epoch": 0.8836650033066795, + "flos": 24718356322560.0, + "grad_norm": 2.1163400619848556, + "language_loss": 0.7818886, + "learning_rate": 1.4018332970869516e-07, + "loss": 0.80409753, + "num_input_tokens_seen": 159002775, + "step": 7349, + "time_per_iteration": 2.6825602054595947 + }, + { + "auxiliary_loss_clip": 0.01178781, + "auxiliary_loss_mlp": 0.01025584, + "balance_loss_clip": 1.04501534, + "balance_loss_mlp": 1.01805258, + "epoch": 0.8837852461973186, + "flos": 25413371556480.0, + "grad_norm": 1.9972697548819585, + "language_loss": 0.85169947, + "learning_rate": 1.398969717367733e-07, + "loss": 0.87374312, + "num_input_tokens_seen": 159024100, + "step": 7350, + "time_per_iteration": 2.7025532722473145 + }, + { + "auxiliary_loss_clip": 0.01202317, + "auxiliary_loss_mlp": 0.01021963, + "balance_loss_clip": 1.04513288, + "balance_loss_mlp": 1.01529646, + "epoch": 0.8839054890879576, + "flos": 17822574195840.0, + "grad_norm": 1.6109809561357658, + "language_loss": 0.76482797, + "learning_rate": 1.396108959410014e-07, + "loss": 0.78707075, + "num_input_tokens_seen": 159043315, + "step": 7351, + "time_per_iteration": 2.6539552211761475 + }, + { + "auxiliary_loss_clip": 0.01178292, + "auxiliary_loss_mlp": 0.01378995, + "balance_loss_clip": 1.04934406, + "balance_loss_mlp": 1.0001514, + "epoch": 0.8840257319785968, + "flos": 23769021818880.0, + "grad_norm": 1.5710632818262997, + "language_loss": 0.81747693, + "learning_rate": 1.3932510236477745e-07, + "loss": 0.84304976, + "num_input_tokens_seen": 159063985, + "step": 7352, + "time_per_iteration": 2.663520336151123 + }, + { + "auxiliary_loss_clip": 0.01173129, + "auxiliary_loss_mlp": 0.01023743, + "balance_loss_clip": 1.04405951, + "balance_loss_mlp": 1.01574111, + "epoch": 0.8841459748692359, + "flos": 29059776622080.0, + "grad_norm": 2.400961057363552, + "language_loss": 0.55919695, + "learning_rate": 1.3903959105145636e-07, + "loss": 0.58116567, + "num_input_tokens_seen": 159084475, + "step": 7353, + "time_per_iteration": 2.6306583881378174 + }, + { + "auxiliary_loss_clip": 0.01165997, + "auxiliary_loss_mlp": 0.01021947, + "balance_loss_clip": 1.0455575, + "balance_loss_mlp": 1.01488686, + "epoch": 0.8842662177598749, + "flos": 24311523905280.0, + "grad_norm": 2.1409634550910033, + "language_loss": 0.83177328, + "learning_rate": 1.387543620443492e-07, + "loss": 0.85365272, + "num_input_tokens_seen": 159101320, + "step": 7354, + "time_per_iteration": 2.5818240642547607 + }, + { + "auxiliary_loss_clip": 0.01166292, + "auxiliary_loss_mlp": 0.0102549, + "balance_loss_clip": 1.04712677, + "balance_loss_mlp": 1.01837945, + "epoch": 0.8843864606505141, + "flos": 25007867942400.0, + "grad_norm": 1.8866647726055286, + "language_loss": 0.840693, + "learning_rate": 1.3846941538672606e-07, + "loss": 0.86261082, + "num_input_tokens_seen": 159120025, + "step": 7355, + "time_per_iteration": 2.5896573066711426 + }, + { + "auxiliary_loss_clip": 0.01212697, + "auxiliary_loss_mlp": 0.01027421, + "balance_loss_clip": 1.04799581, + "balance_loss_mlp": 1.01979208, + "epoch": 0.8845067035411531, + "flos": 28183915388160.0, + "grad_norm": 2.462149526892667, + "language_loss": 0.80732578, + "learning_rate": 1.3818475112181193e-07, + "loss": 0.82972693, + "num_input_tokens_seen": 159138820, + "step": 7356, + "time_per_iteration": 2.802492141723633 + }, + { + "auxiliary_loss_clip": 0.01184481, + "auxiliary_loss_mlp": 0.0101825, + "balance_loss_clip": 1.043419, + "balance_loss_mlp": 1.01192284, + "epoch": 0.8846269464317922, + "flos": 12853219311360.0, + "grad_norm": 2.0644699842911924, + "language_loss": 0.79978621, + "learning_rate": 1.3790036929279091e-07, + "loss": 0.82181352, + "num_input_tokens_seen": 159155975, + "step": 7357, + "time_per_iteration": 2.5971906185150146 + }, + { + "auxiliary_loss_clip": 0.01179499, + "auxiliary_loss_mlp": 0.01378711, + "balance_loss_clip": 1.04735458, + "balance_loss_mlp": 1.00014186, + "epoch": 0.8847471893224313, + "flos": 18624351628800.0, + "grad_norm": 2.311811128674643, + "language_loss": 0.58782417, + "learning_rate": 1.3761626994280363e-07, + "loss": 0.6134063, + "num_input_tokens_seen": 159173445, + "step": 7358, + "time_per_iteration": 2.5912368297576904 + }, + { + "auxiliary_loss_clip": 0.01198909, + "auxiliary_loss_mlp": 0.01027361, + "balance_loss_clip": 1.04339075, + "balance_loss_mlp": 1.0199554, + "epoch": 0.8848674322130704, + "flos": 35769433449600.0, + "grad_norm": 1.7047633207720114, + "language_loss": 0.73456371, + "learning_rate": 1.3733245311494735e-07, + "loss": 0.7568264, + "num_input_tokens_seen": 159196100, + "step": 7359, + "time_per_iteration": 3.748932361602783 + }, + { + "auxiliary_loss_clip": 0.01175899, + "auxiliary_loss_mlp": 0.01027772, + "balance_loss_clip": 1.04546738, + "balance_loss_mlp": 1.0207479, + "epoch": 0.8849876751037095, + "flos": 24243760897920.0, + "grad_norm": 1.873305534639298, + "language_loss": 0.70657736, + "learning_rate": 1.3704891885227676e-07, + "loss": 0.72861409, + "num_input_tokens_seen": 159216145, + "step": 7360, + "time_per_iteration": 2.6071970462799072 + }, + { + "auxiliary_loss_clip": 0.01188406, + "auxiliary_loss_mlp": 0.0102567, + "balance_loss_clip": 1.03939307, + "balance_loss_mlp": 1.01835692, + "epoch": 0.8851079179943486, + "flos": 21500580251520.0, + "grad_norm": 2.015117475730502, + "language_loss": 0.77831888, + "learning_rate": 1.367656671978037e-07, + "loss": 0.80045968, + "num_input_tokens_seen": 159233610, + "step": 7361, + "time_per_iteration": 3.594113826751709 + }, + { + "auxiliary_loss_clip": 0.01191254, + "auxiliary_loss_mlp": 0.01029271, + "balance_loss_clip": 1.04378474, + "balance_loss_mlp": 1.02241373, + "epoch": 0.8852281608849877, + "flos": 15300711198720.0, + "grad_norm": 1.8715836934891488, + "language_loss": 0.73642492, + "learning_rate": 1.36482698194498e-07, + "loss": 0.7586301, + "num_input_tokens_seen": 159250155, + "step": 7362, + "time_per_iteration": 2.6378016471862793 + }, + { + "auxiliary_loss_clip": 0.01181737, + "auxiliary_loss_mlp": 0.01025357, + "balance_loss_clip": 1.04274678, + "balance_loss_mlp": 1.01763809, + "epoch": 0.8853484037756267, + "flos": 23295719283840.0, + "grad_norm": 1.8797465360152403, + "language_loss": 0.71908939, + "learning_rate": 1.3620001188528506e-07, + "loss": 0.74116033, + "num_input_tokens_seen": 159270875, + "step": 7363, + "time_per_iteration": 2.722538948059082 + }, + { + "auxiliary_loss_clip": 0.01178349, + "auxiliary_loss_mlp": 0.01025456, + "balance_loss_clip": 1.0439955, + "balance_loss_mlp": 1.01787162, + "epoch": 0.8854686466662659, + "flos": 25114773795840.0, + "grad_norm": 2.7185065995919886, + "language_loss": 0.73837936, + "learning_rate": 1.3591760831304865e-07, + "loss": 0.7604174, + "num_input_tokens_seen": 159288565, + "step": 7364, + "time_per_iteration": 3.5681538581848145 + }, + { + "auxiliary_loss_clip": 0.01165404, + "auxiliary_loss_mlp": 0.01023734, + "balance_loss_clip": 1.04614902, + "balance_loss_mlp": 1.01671231, + "epoch": 0.885588889556905, + "flos": 21390873137280.0, + "grad_norm": 1.8584597096373479, + "language_loss": 0.7945838, + "learning_rate": 1.356354875206287e-07, + "loss": 0.81647515, + "num_input_tokens_seen": 159306400, + "step": 7365, + "time_per_iteration": 2.5735485553741455 + }, + { + "auxiliary_loss_clip": 0.01194092, + "auxiliary_loss_mlp": 0.01025322, + "balance_loss_clip": 1.04536378, + "balance_loss_mlp": 1.0181154, + "epoch": 0.885709132447544, + "flos": 26906752431360.0, + "grad_norm": 2.1350792510974417, + "language_loss": 0.69903159, + "learning_rate": 1.3535364955082296e-07, + "loss": 0.72122568, + "num_input_tokens_seen": 159326250, + "step": 7366, + "time_per_iteration": 2.786877393722534 + }, + { + "auxiliary_loss_clip": 0.01167533, + "auxiliary_loss_mlp": 0.01024727, + "balance_loss_clip": 1.04886866, + "balance_loss_mlp": 1.01790524, + "epoch": 0.8858293753381832, + "flos": 26103394800000.0, + "grad_norm": 2.1459733861424124, + "language_loss": 0.65053248, + "learning_rate": 1.3507209444638613e-07, + "loss": 0.67245507, + "num_input_tokens_seen": 159348250, + "step": 7367, + "time_per_iteration": 2.6100564002990723 + }, + { + "auxiliary_loss_clip": 0.01177103, + "auxiliary_loss_mlp": 0.01023284, + "balance_loss_clip": 1.04561496, + "balance_loss_mlp": 1.01630473, + "epoch": 0.8859496182288222, + "flos": 23292810282240.0, + "grad_norm": 1.8165388844904067, + "language_loss": 0.73955333, + "learning_rate": 1.347908222500298e-07, + "loss": 0.76155716, + "num_input_tokens_seen": 159368325, + "step": 7368, + "time_per_iteration": 3.53145170211792 + }, + { + "auxiliary_loss_clip": 0.01182257, + "auxiliary_loss_mlp": 0.01028495, + "balance_loss_clip": 1.04409492, + "balance_loss_mlp": 1.02143228, + "epoch": 0.8860698611194613, + "flos": 16872916469760.0, + "grad_norm": 1.9274799145449644, + "language_loss": 0.69788635, + "learning_rate": 1.3450983300442276e-07, + "loss": 0.71999395, + "num_input_tokens_seen": 159387555, + "step": 7369, + "time_per_iteration": 2.6569013595581055 + }, + { + "auxiliary_loss_clip": 0.01177509, + "auxiliary_loss_mlp": 0.01027582, + "balance_loss_clip": 1.04554355, + "balance_loss_mlp": 1.02086473, + "epoch": 0.8861901040101005, + "flos": 24681404206080.0, + "grad_norm": 1.9008083225285577, + "language_loss": 0.73373789, + "learning_rate": 1.3422912675219068e-07, + "loss": 0.7557888, + "num_input_tokens_seen": 159407310, + "step": 7370, + "time_per_iteration": 2.619171380996704 + }, + { + "auxiliary_loss_clip": 0.01165941, + "auxiliary_loss_mlp": 0.01025333, + "balance_loss_clip": 1.04821277, + "balance_loss_mlp": 1.01880062, + "epoch": 0.8863103469007395, + "flos": 24423026699520.0, + "grad_norm": 1.6057690356277305, + "language_loss": 0.79001343, + "learning_rate": 1.339487035359166e-07, + "loss": 0.81192619, + "num_input_tokens_seen": 159427680, + "step": 7371, + "time_per_iteration": 2.6359591484069824 + }, + { + "auxiliary_loss_clip": 0.01187573, + "auxiliary_loss_mlp": 0.0137797, + "balance_loss_clip": 1.04678333, + "balance_loss_mlp": 1.00013268, + "epoch": 0.8864305897913786, + "flos": 22053964158720.0, + "grad_norm": 2.95259671677694, + "language_loss": 0.84696656, + "learning_rate": 1.336685633981409e-07, + "loss": 0.87262201, + "num_input_tokens_seen": 159448765, + "step": 7372, + "time_per_iteration": 2.666940212249756 + }, + { + "auxiliary_loss_clip": 0.01175696, + "auxiliary_loss_mlp": 0.01026469, + "balance_loss_clip": 1.04332757, + "balance_loss_mlp": 1.01943278, + "epoch": 0.8865508326820177, + "flos": 19099449843840.0, + "grad_norm": 2.0484677786487424, + "language_loss": 0.74923378, + "learning_rate": 1.333887063813597e-07, + "loss": 0.77125543, + "num_input_tokens_seen": 159466870, + "step": 7373, + "time_per_iteration": 2.636983633041382 + }, + { + "auxiliary_loss_clip": 0.01186433, + "auxiliary_loss_mlp": 0.01025605, + "balance_loss_clip": 1.04327345, + "balance_loss_mlp": 1.01897705, + "epoch": 0.8866710755726568, + "flos": 15414189240960.0, + "grad_norm": 1.83210734302761, + "language_loss": 0.66642314, + "learning_rate": 1.331091325280278e-07, + "loss": 0.68854356, + "num_input_tokens_seen": 159485840, + "step": 7374, + "time_per_iteration": 2.625487804412842 + }, + { + "auxiliary_loss_clip": 0.01192189, + "auxiliary_loss_mlp": 0.01021129, + "balance_loss_clip": 1.03951192, + "balance_loss_mlp": 1.013762, + "epoch": 0.8867913184632958, + "flos": 20083689388800.0, + "grad_norm": 2.02847116547195, + "language_loss": 0.7865746, + "learning_rate": 1.3282984188055625e-07, + "loss": 0.80870777, + "num_input_tokens_seen": 159505630, + "step": 7375, + "time_per_iteration": 2.7502191066741943 + }, + { + "auxiliary_loss_clip": 0.01166365, + "auxiliary_loss_mlp": 0.01022565, + "balance_loss_clip": 1.04566443, + "balance_loss_mlp": 1.01540029, + "epoch": 0.8869115613539349, + "flos": 23365852588800.0, + "grad_norm": 1.7997364232188762, + "language_loss": 0.79786032, + "learning_rate": 1.3255083448131288e-07, + "loss": 0.81974959, + "num_input_tokens_seen": 159524675, + "step": 7376, + "time_per_iteration": 2.5677154064178467 + }, + { + "auxiliary_loss_clip": 0.01178738, + "auxiliary_loss_mlp": 0.01028567, + "balance_loss_clip": 1.0437479, + "balance_loss_mlp": 1.02115214, + "epoch": 0.8870318042445741, + "flos": 21286840371840.0, + "grad_norm": 2.066103381200194, + "language_loss": 0.7938745, + "learning_rate": 1.3227211037262365e-07, + "loss": 0.81594753, + "num_input_tokens_seen": 159541915, + "step": 7377, + "time_per_iteration": 2.6184146404266357 + }, + { + "auxiliary_loss_clip": 0.01204871, + "auxiliary_loss_mlp": 0.01029581, + "balance_loss_clip": 1.04140151, + "balance_loss_mlp": 1.0219934, + "epoch": 0.8871520471352131, + "flos": 20010862563840.0, + "grad_norm": 2.0755957566603276, + "language_loss": 0.85425806, + "learning_rate": 1.319936695967696e-07, + "loss": 0.87660259, + "num_input_tokens_seen": 159559740, + "step": 7378, + "time_per_iteration": 2.7842392921447754 + }, + { + "auxiliary_loss_clip": 0.01172961, + "auxiliary_loss_mlp": 0.01026228, + "balance_loss_clip": 1.04812217, + "balance_loss_mlp": 1.01824045, + "epoch": 0.8872722900258522, + "flos": 22601422321920.0, + "grad_norm": 2.362109464981751, + "language_loss": 0.82032371, + "learning_rate": 1.3171551219599097e-07, + "loss": 0.84231561, + "num_input_tokens_seen": 159578265, + "step": 7379, + "time_per_iteration": 2.662667989730835 + }, + { + "auxiliary_loss_clip": 0.01170409, + "auxiliary_loss_mlp": 0.01025123, + "balance_loss_clip": 1.0500617, + "balance_loss_mlp": 1.01763642, + "epoch": 0.8873925329164913, + "flos": 22163276223360.0, + "grad_norm": 2.2259510828984888, + "language_loss": 0.78146607, + "learning_rate": 1.3143763821248377e-07, + "loss": 0.80342138, + "num_input_tokens_seen": 159595350, + "step": 7380, + "time_per_iteration": 2.5904130935668945 + }, + { + "auxiliary_loss_clip": 0.01165013, + "auxiliary_loss_mlp": 0.01019985, + "balance_loss_clip": 1.04653859, + "balance_loss_mlp": 1.01302326, + "epoch": 0.8875127758071304, + "flos": 19208223204480.0, + "grad_norm": 1.7044928225371923, + "language_loss": 0.72284824, + "learning_rate": 1.3116004768840118e-07, + "loss": 0.74469823, + "num_input_tokens_seen": 159613725, + "step": 7381, + "time_per_iteration": 2.613051176071167 + }, + { + "auxiliary_loss_clip": 0.01168244, + "auxiliary_loss_mlp": 0.01026134, + "balance_loss_clip": 1.0468967, + "balance_loss_mlp": 1.01870441, + "epoch": 0.8876330186977694, + "flos": 18110900666880.0, + "grad_norm": 1.5868506017396817, + "language_loss": 0.74377203, + "learning_rate": 1.3088274066585348e-07, + "loss": 0.76571584, + "num_input_tokens_seen": 159631335, + "step": 7382, + "time_per_iteration": 2.5135111808776855 + }, + { + "auxiliary_loss_clip": 0.01196242, + "auxiliary_loss_mlp": 0.0102388, + "balance_loss_clip": 1.04116035, + "balance_loss_mlp": 1.01691794, + "epoch": 0.8877532615884086, + "flos": 22009434272640.0, + "grad_norm": 6.2630809523059225, + "language_loss": 0.90356499, + "learning_rate": 1.3060571718690749e-07, + "loss": 0.92576611, + "num_input_tokens_seen": 159648830, + "step": 7383, + "time_per_iteration": 2.702427387237549 + }, + { + "auxiliary_loss_clip": 0.01098758, + "auxiliary_loss_mlp": 0.01373946, + "balance_loss_clip": 1.00839508, + "balance_loss_mlp": 0.99993783, + "epoch": 0.8878735044790477, + "flos": 72136924346880.0, + "grad_norm": 0.7457422095418295, + "language_loss": 0.56859118, + "learning_rate": 1.3032897729358805e-07, + "loss": 0.59331822, + "num_input_tokens_seen": 159709785, + "step": 7384, + "time_per_iteration": 3.282365560531616 + }, + { + "auxiliary_loss_clip": 0.01193057, + "auxiliary_loss_mlp": 0.01378838, + "balance_loss_clip": 1.03519654, + "balance_loss_mlp": 1.00007915, + "epoch": 0.8879937473696867, + "flos": 27526355061120.0, + "grad_norm": 1.8659815124617, + "language_loss": 0.79900873, + "learning_rate": 1.3005252102787645e-07, + "loss": 0.82472765, + "num_input_tokens_seen": 159728725, + "step": 7385, + "time_per_iteration": 3.6172330379486084 + }, + { + "auxiliary_loss_clip": 0.01178884, + "auxiliary_loss_mlp": 0.01028912, + "balance_loss_clip": 1.04502964, + "balance_loss_mlp": 1.02182817, + "epoch": 0.8881139902603259, + "flos": 22234091886720.0, + "grad_norm": 1.6334760136961528, + "language_loss": 0.73538315, + "learning_rate": 1.297763484317105e-07, + "loss": 0.75746113, + "num_input_tokens_seen": 159747020, + "step": 7386, + "time_per_iteration": 2.6297647953033447 + }, + { + "auxiliary_loss_clip": 0.01198143, + "auxiliary_loss_mlp": 0.01378985, + "balance_loss_clip": 1.03999805, + "balance_loss_mlp": 1.00014389, + "epoch": 0.888234233150965, + "flos": 20299548170880.0, + "grad_norm": 2.8606072310084234, + "language_loss": 0.70459718, + "learning_rate": 1.2950045954698551e-07, + "loss": 0.73036844, + "num_input_tokens_seen": 159764855, + "step": 7387, + "time_per_iteration": 3.5202105045318604 + }, + { + "auxiliary_loss_clip": 0.01186594, + "auxiliary_loss_mlp": 0.01027266, + "balance_loss_clip": 1.0443604, + "balance_loss_mlp": 1.01964247, + "epoch": 0.888354476041604, + "flos": 18147996437760.0, + "grad_norm": 1.7532136643144114, + "language_loss": 0.75507212, + "learning_rate": 1.2922485441555343e-07, + "loss": 0.77721077, + "num_input_tokens_seen": 159783935, + "step": 7388, + "time_per_iteration": 2.755295991897583 + }, + { + "auxiliary_loss_clip": 0.01166322, + "auxiliary_loss_mlp": 0.01025381, + "balance_loss_clip": 1.0456773, + "balance_loss_mlp": 1.01845193, + "epoch": 0.8884747189322432, + "flos": 22014282608640.0, + "grad_norm": 1.999410366055808, + "language_loss": 0.81732213, + "learning_rate": 1.2894953307922363e-07, + "loss": 0.83923912, + "num_input_tokens_seen": 159802895, + "step": 7389, + "time_per_iteration": 2.7021172046661377 + }, + { + "auxiliary_loss_clip": 0.01189722, + "auxiliary_loss_mlp": 0.01026917, + "balance_loss_clip": 1.0436852, + "balance_loss_mlp": 1.01905549, + "epoch": 0.8885949618228822, + "flos": 19786779567360.0, + "grad_norm": 1.9695055592530557, + "language_loss": 0.83727825, + "learning_rate": 1.2867449557976208e-07, + "loss": 0.85944468, + "num_input_tokens_seen": 159820995, + "step": 7390, + "time_per_iteration": 3.6513917446136475 + }, + { + "auxiliary_loss_clip": 0.01178675, + "auxiliary_loss_mlp": 0.01029159, + "balance_loss_clip": 1.0488472, + "balance_loss_mlp": 1.02174735, + "epoch": 0.8887152047135213, + "flos": 20047599198720.0, + "grad_norm": 3.480290628222127, + "language_loss": 0.75636756, + "learning_rate": 1.283997419588916e-07, + "loss": 0.77844584, + "num_input_tokens_seen": 159840465, + "step": 7391, + "time_per_iteration": 2.634431838989258 + }, + { + "auxiliary_loss_clip": 0.01179842, + "auxiliary_loss_mlp": 0.01023309, + "balance_loss_clip": 1.04602838, + "balance_loss_mlp": 1.01642442, + "epoch": 0.8888354476041604, + "flos": 18588117784320.0, + "grad_norm": 2.0761552165549184, + "language_loss": 0.61889327, + "learning_rate": 1.2812527225829216e-07, + "loss": 0.64092481, + "num_input_tokens_seen": 159858690, + "step": 7392, + "time_per_iteration": 2.628875970840454 + }, + { + "auxiliary_loss_clip": 0.0118369, + "auxiliary_loss_mlp": 0.0102639, + "balance_loss_clip": 1.04756045, + "balance_loss_mlp": 1.01862907, + "epoch": 0.8889556904947995, + "flos": 21689794120320.0, + "grad_norm": 1.9400449819537413, + "language_loss": 0.76475799, + "learning_rate": 1.2785108651960052e-07, + "loss": 0.78685886, + "num_input_tokens_seen": 159880325, + "step": 7393, + "time_per_iteration": 2.649075984954834 + }, + { + "auxiliary_loss_clip": 0.01176344, + "auxiliary_loss_mlp": 0.01027554, + "balance_loss_clip": 1.04321992, + "balance_loss_mlp": 1.02060699, + "epoch": 0.8890759333854386, + "flos": 27381204201600.0, + "grad_norm": 1.8998508261495568, + "language_loss": 0.80817562, + "learning_rate": 1.2757718478441094e-07, + "loss": 0.83021462, + "num_input_tokens_seen": 159901070, + "step": 7394, + "time_per_iteration": 3.577308416366577 + }, + { + "auxiliary_loss_clip": 0.01180731, + "auxiliary_loss_mlp": 0.01027689, + "balance_loss_clip": 1.04159009, + "balance_loss_mlp": 1.02072144, + "epoch": 0.8891961762760777, + "flos": 24498834353280.0, + "grad_norm": 1.7463529977047385, + "language_loss": 0.77182752, + "learning_rate": 1.2730356709427302e-07, + "loss": 0.79391176, + "num_input_tokens_seen": 159919750, + "step": 7395, + "time_per_iteration": 2.66304612159729 + }, + { + "auxiliary_loss_clip": 0.01175546, + "auxiliary_loss_mlp": 0.01020099, + "balance_loss_clip": 1.04790187, + "balance_loss_mlp": 1.01288414, + "epoch": 0.8893164191667168, + "flos": 41499770895360.0, + "grad_norm": 1.6135337164607195, + "language_loss": 0.60062933, + "learning_rate": 1.2703023349069542e-07, + "loss": 0.62258577, + "num_input_tokens_seen": 159944600, + "step": 7396, + "time_per_iteration": 2.7782623767852783 + }, + { + "auxiliary_loss_clip": 0.01172743, + "auxiliary_loss_mlp": 0.01027005, + "balance_loss_clip": 1.04651046, + "balance_loss_mlp": 1.01997185, + "epoch": 0.8894366620573558, + "flos": 33583623120000.0, + "grad_norm": 1.7572668222789085, + "language_loss": 0.61287713, + "learning_rate": 1.2675718401514223e-07, + "loss": 0.63487458, + "num_input_tokens_seen": 159968780, + "step": 7397, + "time_per_iteration": 2.694873094558716 + }, + { + "auxiliary_loss_clip": 0.0118427, + "auxiliary_loss_mlp": 0.01027049, + "balance_loss_clip": 1.04501486, + "balance_loss_mlp": 1.01969934, + "epoch": 0.889556904947995, + "flos": 16909832672640.0, + "grad_norm": 2.3167611814029896, + "language_loss": 0.74653429, + "learning_rate": 1.264844187090346e-07, + "loss": 0.76864749, + "num_input_tokens_seen": 159985905, + "step": 7398, + "time_per_iteration": 2.6373131275177 + }, + { + "auxiliary_loss_clip": 0.01176422, + "auxiliary_loss_mlp": 0.01026819, + "balance_loss_clip": 1.0411706, + "balance_loss_mlp": 1.01985168, + "epoch": 0.889677147838634, + "flos": 26030855283840.0, + "grad_norm": 1.7575386409130986, + "language_loss": 0.75115788, + "learning_rate": 1.262119376137516e-07, + "loss": 0.77319032, + "num_input_tokens_seen": 160006965, + "step": 7399, + "time_per_iteration": 2.592769145965576 + }, + { + "auxiliary_loss_clip": 0.01164807, + "auxiliary_loss_mlp": 0.01025012, + "balance_loss_clip": 1.04394388, + "balance_loss_mlp": 1.01839924, + "epoch": 0.8897973907292731, + "flos": 26468283110400.0, + "grad_norm": 1.5014156640886667, + "language_loss": 0.85054493, + "learning_rate": 1.2593974077062707e-07, + "loss": 0.87244308, + "num_input_tokens_seen": 160028585, + "step": 7400, + "time_per_iteration": 2.633498191833496 + }, + { + "auxiliary_loss_clip": 0.01182627, + "auxiliary_loss_mlp": 0.0102512, + "balance_loss_clip": 1.04072309, + "balance_loss_mlp": 1.01781225, + "epoch": 0.8899176336199123, + "flos": 26249694894720.0, + "grad_norm": 1.720590906681602, + "language_loss": 0.63772368, + "learning_rate": 1.2566782822095423e-07, + "loss": 0.65980124, + "num_input_tokens_seen": 160048840, + "step": 7401, + "time_per_iteration": 2.6741483211517334 + }, + { + "auxiliary_loss_clip": 0.01200485, + "auxiliary_loss_mlp": 0.01022825, + "balance_loss_clip": 1.04642689, + "balance_loss_mlp": 1.01554739, + "epoch": 0.8900378765105513, + "flos": 20811742156800.0, + "grad_norm": 2.002373336473828, + "language_loss": 0.71302891, + "learning_rate": 1.2539620000598162e-07, + "loss": 0.73526204, + "num_input_tokens_seen": 160068175, + "step": 7402, + "time_per_iteration": 2.6676383018493652 + }, + { + "auxiliary_loss_clip": 0.01165042, + "auxiliary_loss_mlp": 0.01021271, + "balance_loss_clip": 1.04572129, + "balance_loss_mlp": 1.01388884, + "epoch": 0.8901581194011904, + "flos": 16472333018880.0, + "grad_norm": 1.685891770235067, + "language_loss": 0.79687679, + "learning_rate": 1.2512485616691492e-07, + "loss": 0.81873995, + "num_input_tokens_seen": 160085230, + "step": 7403, + "time_per_iteration": 2.6093616485595703 + }, + { + "auxiliary_loss_clip": 0.01195047, + "auxiliary_loss_mlp": 0.01028621, + "balance_loss_clip": 1.04350567, + "balance_loss_mlp": 1.02086091, + "epoch": 0.8902783622918296, + "flos": 35155253773440.0, + "grad_norm": 1.3736011914071533, + "language_loss": 0.80775571, + "learning_rate": 1.2485379674491681e-07, + "loss": 0.82999235, + "num_input_tokens_seen": 160111425, + "step": 7404, + "time_per_iteration": 2.8503661155700684 + }, + { + "auxiliary_loss_clip": 0.01183453, + "auxiliary_loss_mlp": 0.01031929, + "balance_loss_clip": 1.04541516, + "balance_loss_mlp": 1.02429342, + "epoch": 0.8903986051824686, + "flos": 17201068145280.0, + "grad_norm": 2.158206234170756, + "language_loss": 0.7952621, + "learning_rate": 1.2458302178110657e-07, + "loss": 0.81741589, + "num_input_tokens_seen": 160129790, + "step": 7405, + "time_per_iteration": 2.6450576782226562 + }, + { + "auxiliary_loss_clip": 0.01181436, + "auxiliary_loss_mlp": 0.01021386, + "balance_loss_clip": 1.03913355, + "balance_loss_mlp": 1.01457906, + "epoch": 0.8905188480731077, + "flos": 25483863997440.0, + "grad_norm": 1.9918627327146465, + "language_loss": 0.82195342, + "learning_rate": 1.2431253131656118e-07, + "loss": 0.84398162, + "num_input_tokens_seen": 160149265, + "step": 7406, + "time_per_iteration": 2.6762728691101074 + }, + { + "auxiliary_loss_clip": 0.0117875, + "auxiliary_loss_mlp": 0.01021911, + "balance_loss_clip": 1.04482043, + "balance_loss_mlp": 1.01426351, + "epoch": 0.8906390909637467, + "flos": 23365888502400.0, + "grad_norm": 1.9256208505027166, + "language_loss": 0.76480693, + "learning_rate": 1.240423253923133e-07, + "loss": 0.7868135, + "num_input_tokens_seen": 160168870, + "step": 7407, + "time_per_iteration": 2.660203218460083 + }, + { + "auxiliary_loss_clip": 0.01177415, + "auxiliary_loss_mlp": 0.01024827, + "balance_loss_clip": 1.04615068, + "balance_loss_mlp": 1.01724505, + "epoch": 0.8907593338543859, + "flos": 21068790860160.0, + "grad_norm": 2.083479158068002, + "language_loss": 0.69386148, + "learning_rate": 1.237724040493533e-07, + "loss": 0.71588391, + "num_input_tokens_seen": 160187495, + "step": 7408, + "time_per_iteration": 2.588282823562622 + }, + { + "auxiliary_loss_clip": 0.01171336, + "auxiliary_loss_mlp": 0.01027937, + "balance_loss_clip": 1.04841447, + "balance_loss_mlp": 1.02035856, + "epoch": 0.8908795767450249, + "flos": 21869562712320.0, + "grad_norm": 2.572236680968726, + "language_loss": 0.72995782, + "learning_rate": 1.2350276732862773e-07, + "loss": 0.7519505, + "num_input_tokens_seen": 160208520, + "step": 7409, + "time_per_iteration": 2.6511847972869873 + }, + { + "auxiliary_loss_clip": 0.01075176, + "auxiliary_loss_mlp": 0.01000459, + "balance_loss_clip": 1.00985503, + "balance_loss_mlp": 0.99940366, + "epoch": 0.890999819635664, + "flos": 66307869348480.0, + "grad_norm": 0.8401247729077759, + "language_loss": 0.56700546, + "learning_rate": 1.2323341527103993e-07, + "loss": 0.58776182, + "num_input_tokens_seen": 160263720, + "step": 7410, + "time_per_iteration": 3.1018033027648926 + }, + { + "auxiliary_loss_clip": 0.01164169, + "auxiliary_loss_mlp": 0.01028739, + "balance_loss_clip": 1.04512095, + "balance_loss_mlp": 1.02177167, + "epoch": 0.8911200625263032, + "flos": 26869908055680.0, + "grad_norm": 1.8571832971786437, + "language_loss": 0.85171092, + "learning_rate": 1.2296434791745135e-07, + "loss": 0.87364006, + "num_input_tokens_seen": 160282170, + "step": 7411, + "time_per_iteration": 3.4831736087799072 + }, + { + "auxiliary_loss_clip": 0.01178409, + "auxiliary_loss_mlp": 0.01017926, + "balance_loss_clip": 1.0466516, + "balance_loss_mlp": 1.01088619, + "epoch": 0.8912403054169422, + "flos": 20885825957760.0, + "grad_norm": 1.6190006431069472, + "language_loss": 0.76615876, + "learning_rate": 1.2269556530867875e-07, + "loss": 0.78812212, + "num_input_tokens_seen": 160300725, + "step": 7412, + "time_per_iteration": 2.623248338699341 + }, + { + "auxiliary_loss_clip": 0.01173206, + "auxiliary_loss_mlp": 0.01027318, + "balance_loss_clip": 1.04908824, + "balance_loss_mlp": 1.0190444, + "epoch": 0.8913605483075813, + "flos": 27016567286400.0, + "grad_norm": 2.0140849534224907, + "language_loss": 0.82182312, + "learning_rate": 1.2242706748549614e-07, + "loss": 0.84382832, + "num_input_tokens_seen": 160318720, + "step": 7413, + "time_per_iteration": 2.5859858989715576 + }, + { + "auxiliary_loss_clip": 0.01183528, + "auxiliary_loss_mlp": 0.01022165, + "balance_loss_clip": 1.04118884, + "balance_loss_mlp": 1.01482153, + "epoch": 0.8914807911982204, + "flos": 23621500661760.0, + "grad_norm": 1.7062902525277388, + "language_loss": 0.82140023, + "learning_rate": 1.2215885448863473e-07, + "loss": 0.84345716, + "num_input_tokens_seen": 160339595, + "step": 7414, + "time_per_iteration": 3.5764291286468506 + }, + { + "auxiliary_loss_clip": 0.01182782, + "auxiliary_loss_mlp": 0.01025615, + "balance_loss_clip": 1.04561353, + "balance_loss_mlp": 1.01831043, + "epoch": 0.8916010340888595, + "flos": 24462277286400.0, + "grad_norm": 1.8937611498701286, + "language_loss": 0.80313706, + "learning_rate": 1.2189092635878152e-07, + "loss": 0.82522106, + "num_input_tokens_seen": 160361045, + "step": 7415, + "time_per_iteration": 2.6415579319000244 + }, + { + "auxiliary_loss_clip": 0.01182257, + "auxiliary_loss_mlp": 0.01023623, + "balance_loss_clip": 1.04018557, + "balance_loss_mlp": 1.01645231, + "epoch": 0.8917212769794985, + "flos": 21215773313280.0, + "grad_norm": 1.6495821635269519, + "language_loss": 0.77280533, + "learning_rate": 1.216232831365822e-07, + "loss": 0.79486412, + "num_input_tokens_seen": 160379990, + "step": 7416, + "time_per_iteration": 3.613638162612915 + }, + { + "auxiliary_loss_clip": 0.01189508, + "auxiliary_loss_mlp": 0.01027689, + "balance_loss_clip": 1.04442656, + "balance_loss_mlp": 1.020468, + "epoch": 0.8918415198701377, + "flos": 25513992529920.0, + "grad_norm": 1.7614300164173118, + "language_loss": 0.81084305, + "learning_rate": 1.2135592486263678e-07, + "loss": 0.83301502, + "num_input_tokens_seen": 160399240, + "step": 7417, + "time_per_iteration": 2.67952036857605 + }, + { + "auxiliary_loss_clip": 0.01184601, + "auxiliary_loss_mlp": 0.01024449, + "balance_loss_clip": 1.04338527, + "balance_loss_mlp": 1.01694441, + "epoch": 0.8919617627607768, + "flos": 37853006693760.0, + "grad_norm": 1.5771466415601945, + "language_loss": 0.60856557, + "learning_rate": 1.2108885157750415e-07, + "loss": 0.63065606, + "num_input_tokens_seen": 160421600, + "step": 7418, + "time_per_iteration": 2.7600128650665283 + }, + { + "auxiliary_loss_clip": 0.01190687, + "auxiliary_loss_mlp": 0.01378598, + "balance_loss_clip": 1.04554451, + "balance_loss_mlp": 1.00020015, + "epoch": 0.8920820056514158, + "flos": 26213676531840.0, + "grad_norm": 2.3848109080579434, + "language_loss": 0.80188328, + "learning_rate": 1.2082206332169897e-07, + "loss": 0.82757616, + "num_input_tokens_seen": 160441695, + "step": 7419, + "time_per_iteration": 2.7143588066101074 + }, + { + "auxiliary_loss_clip": 0.01183089, + "auxiliary_loss_mlp": 0.01020013, + "balance_loss_clip": 1.04720676, + "balance_loss_mlp": 1.01307261, + "epoch": 0.892202248542055, + "flos": 17383135207680.0, + "grad_norm": 2.882124659706719, + "language_loss": 0.73135942, + "learning_rate": 1.2055556013569225e-07, + "loss": 0.75339043, + "num_input_tokens_seen": 160457205, + "step": 7420, + "time_per_iteration": 2.594412326812744 + }, + { + "auxiliary_loss_clip": 0.01186538, + "auxiliary_loss_mlp": 0.01024903, + "balance_loss_clip": 1.04501438, + "balance_loss_mlp": 1.01801598, + "epoch": 0.892322491432694, + "flos": 21324223451520.0, + "grad_norm": 3.2859929662833616, + "language_loss": 0.82249165, + "learning_rate": 1.2028934205991315e-07, + "loss": 0.84460604, + "num_input_tokens_seen": 160476525, + "step": 7421, + "time_per_iteration": 3.9776580333709717 + }, + { + "auxiliary_loss_clip": 0.01174279, + "auxiliary_loss_mlp": 0.01029597, + "balance_loss_clip": 1.04394126, + "balance_loss_mlp": 1.02225637, + "epoch": 0.8924427343233331, + "flos": 24029374573440.0, + "grad_norm": 1.4108477649698936, + "language_loss": 0.76849848, + "learning_rate": 1.2002340913474607e-07, + "loss": 0.79053724, + "num_input_tokens_seen": 160500160, + "step": 7422, + "time_per_iteration": 2.7666804790496826 + }, + { + "auxiliary_loss_clip": 0.01167339, + "auxiliary_loss_mlp": 0.01026056, + "balance_loss_clip": 1.04624093, + "balance_loss_mlp": 1.01812911, + "epoch": 0.8925629772139723, + "flos": 30008069631360.0, + "grad_norm": 3.1944179071981993, + "language_loss": 0.74352407, + "learning_rate": 1.1975776140053317e-07, + "loss": 0.76545799, + "num_input_tokens_seen": 160520130, + "step": 7423, + "time_per_iteration": 2.6614272594451904 + }, + { + "auxiliary_loss_clip": 0.0120764, + "auxiliary_loss_mlp": 0.01030712, + "balance_loss_clip": 1.04650521, + "balance_loss_mlp": 1.0230763, + "epoch": 0.8926832201046113, + "flos": 22601709630720.0, + "grad_norm": 2.070419210631777, + "language_loss": 0.73411828, + "learning_rate": 1.194923988975729e-07, + "loss": 0.75650185, + "num_input_tokens_seen": 160539730, + "step": 7424, + "time_per_iteration": 2.916212797164917 + }, + { + "auxiliary_loss_clip": 0.01194656, + "auxiliary_loss_mlp": 0.0102071, + "balance_loss_clip": 1.04533219, + "balance_loss_mlp": 1.01325369, + "epoch": 0.8928034629952504, + "flos": 13297722117120.0, + "grad_norm": 2.1401762367101025, + "language_loss": 0.73438609, + "learning_rate": 1.192273216661206e-07, + "loss": 0.7565397, + "num_input_tokens_seen": 160557820, + "step": 7425, + "time_per_iteration": 2.6202428340911865 + }, + { + "auxiliary_loss_clip": 0.01122417, + "auxiliary_loss_mlp": 0.01002636, + "balance_loss_clip": 1.00820935, + "balance_loss_mlp": 1.00142002, + "epoch": 0.8929237058858895, + "flos": 54854556744960.0, + "grad_norm": 0.7698501768112455, + "language_loss": 0.57496953, + "learning_rate": 1.189625297463881e-07, + "loss": 0.59622008, + "num_input_tokens_seen": 160619510, + "step": 7426, + "time_per_iteration": 3.2270359992980957 + }, + { + "auxiliary_loss_clip": 0.01201496, + "auxiliary_loss_mlp": 0.01025997, + "balance_loss_clip": 1.03638339, + "balance_loss_mlp": 1.01904738, + "epoch": 0.8930439487765286, + "flos": 28883850785280.0, + "grad_norm": 1.8164641309599012, + "language_loss": 0.79683149, + "learning_rate": 1.1869802317854394e-07, + "loss": 0.81910646, + "num_input_tokens_seen": 160643295, + "step": 7427, + "time_per_iteration": 2.830596923828125 + }, + { + "auxiliary_loss_clip": 0.01206336, + "auxiliary_loss_mlp": 0.01023303, + "balance_loss_clip": 1.04240048, + "balance_loss_mlp": 1.01614785, + "epoch": 0.8931641916671677, + "flos": 22419283432320.0, + "grad_norm": 1.8216899755749554, + "language_loss": 0.72192508, + "learning_rate": 1.1843380200271425e-07, + "loss": 0.74422145, + "num_input_tokens_seen": 160662495, + "step": 7428, + "time_per_iteration": 2.7438416481018066 + }, + { + "auxiliary_loss_clip": 0.01184344, + "auxiliary_loss_mlp": 0.01024735, + "balance_loss_clip": 1.04144049, + "balance_loss_mlp": 1.01707625, + "epoch": 0.8932844345578068, + "flos": 25843149786240.0, + "grad_norm": 1.726863294727006, + "language_loss": 0.80434352, + "learning_rate": 1.181698662589805e-07, + "loss": 0.82643431, + "num_input_tokens_seen": 160682080, + "step": 7429, + "time_per_iteration": 2.7409932613372803 + }, + { + "auxiliary_loss_clip": 0.01173749, + "auxiliary_loss_mlp": 0.01022887, + "balance_loss_clip": 1.04362071, + "balance_loss_mlp": 1.01568663, + "epoch": 0.8934046774484459, + "flos": 22925803069440.0, + "grad_norm": 4.234771750484045, + "language_loss": 0.75868094, + "learning_rate": 1.1790621598738249e-07, + "loss": 0.78064728, + "num_input_tokens_seen": 160700395, + "step": 7430, + "time_per_iteration": 2.6202774047851562 + }, + { + "auxiliary_loss_clip": 0.01165311, + "auxiliary_loss_mlp": 0.01023832, + "balance_loss_clip": 1.04722643, + "balance_loss_mlp": 1.01705825, + "epoch": 0.8935249203390849, + "flos": 24462097718400.0, + "grad_norm": 2.1518258956312626, + "language_loss": 0.74722552, + "learning_rate": 1.1764285122791461e-07, + "loss": 0.76911694, + "num_input_tokens_seen": 160721115, + "step": 7431, + "time_per_iteration": 2.584111213684082 + }, + { + "auxiliary_loss_clip": 0.01174392, + "auxiliary_loss_mlp": 0.01024186, + "balance_loss_clip": 1.04467094, + "balance_loss_mlp": 1.01728976, + "epoch": 0.8936451632297241, + "flos": 15742735966080.0, + "grad_norm": 1.6582480372015713, + "language_loss": 0.77310348, + "learning_rate": 1.173797720205294e-07, + "loss": 0.7950893, + "num_input_tokens_seen": 160739150, + "step": 7432, + "time_per_iteration": 2.6073930263519287 + }, + { + "auxiliary_loss_clip": 0.0117906, + "auxiliary_loss_mlp": 0.01026671, + "balance_loss_clip": 1.04773211, + "balance_loss_mlp": 1.01894331, + "epoch": 0.8937654061203631, + "flos": 35115500396160.0, + "grad_norm": 5.967764380177802, + "language_loss": 0.7170893, + "learning_rate": 1.1711697840513602e-07, + "loss": 0.73914659, + "num_input_tokens_seen": 160758585, + "step": 7433, + "time_per_iteration": 2.7476460933685303 + }, + { + "auxiliary_loss_clip": 0.01167657, + "auxiliary_loss_mlp": 0.01022229, + "balance_loss_clip": 1.0432117, + "balance_loss_mlp": 1.01509976, + "epoch": 0.8938856490110022, + "flos": 16107444708480.0, + "grad_norm": 2.1415465565246286, + "language_loss": 0.70673239, + "learning_rate": 1.1685447042160012e-07, + "loss": 0.72863126, + "num_input_tokens_seen": 160776620, + "step": 7434, + "time_per_iteration": 2.6209700107574463 + }, + { + "auxiliary_loss_clip": 0.01169381, + "auxiliary_loss_mlp": 0.01024762, + "balance_loss_clip": 1.04701591, + "balance_loss_mlp": 1.01725197, + "epoch": 0.8940058919016414, + "flos": 20704189858560.0, + "grad_norm": 2.6936942024579085, + "language_loss": 0.71411002, + "learning_rate": 1.1659224810974367e-07, + "loss": 0.73605144, + "num_input_tokens_seen": 160796580, + "step": 7435, + "time_per_iteration": 2.623328447341919 + }, + { + "auxiliary_loss_clip": 0.01184044, + "auxiliary_loss_mlp": 0.01025231, + "balance_loss_clip": 1.04689169, + "balance_loss_mlp": 1.01764309, + "epoch": 0.8941261347922804, + "flos": 25229041937280.0, + "grad_norm": 1.5625286025195215, + "language_loss": 0.68314362, + "learning_rate": 1.1633031150934591e-07, + "loss": 0.70523632, + "num_input_tokens_seen": 160819610, + "step": 7436, + "time_per_iteration": 2.730044364929199 + }, + { + "auxiliary_loss_clip": 0.01180115, + "auxiliary_loss_mlp": 0.01029984, + "balance_loss_clip": 1.04741251, + "balance_loss_mlp": 1.02238178, + "epoch": 0.8942463776829195, + "flos": 19537236806400.0, + "grad_norm": 1.7370757123899574, + "language_loss": 0.79935783, + "learning_rate": 1.1606866066014176e-07, + "loss": 0.82145882, + "num_input_tokens_seen": 160838660, + "step": 7437, + "time_per_iteration": 3.794613838195801 + }, + { + "auxiliary_loss_clip": 0.01191771, + "auxiliary_loss_mlp": 0.01027287, + "balance_loss_clip": 1.04155278, + "balance_loss_mlp": 1.02032876, + "epoch": 0.8943666205735585, + "flos": 22301567585280.0, + "grad_norm": 2.5759923141350898, + "language_loss": 0.75427139, + "learning_rate": 1.1580729560182434e-07, + "loss": 0.77646202, + "num_input_tokens_seen": 160854515, + "step": 7438, + "time_per_iteration": 2.6374528408050537 + }, + { + "auxiliary_loss_clip": 0.01165505, + "auxiliary_loss_mlp": 0.01378402, + "balance_loss_clip": 1.04652905, + "balance_loss_mlp": 1.00014138, + "epoch": 0.8944868634641977, + "flos": 18912893581440.0, + "grad_norm": 1.9029775747329989, + "language_loss": 0.71022898, + "learning_rate": 1.1554621637404171e-07, + "loss": 0.73566806, + "num_input_tokens_seen": 160872605, + "step": 7439, + "time_per_iteration": 2.559951066970825 + }, + { + "auxiliary_loss_clip": 0.01175943, + "auxiliary_loss_mlp": 0.01023308, + "balance_loss_clip": 1.04459882, + "balance_loss_mlp": 1.01653099, + "epoch": 0.8946071063548368, + "flos": 14460904241280.0, + "grad_norm": 2.211220095510334, + "language_loss": 0.60921812, + "learning_rate": 1.1528542301639999e-07, + "loss": 0.63121068, + "num_input_tokens_seen": 160889395, + "step": 7440, + "time_per_iteration": 3.577423095703125 + }, + { + "auxiliary_loss_clip": 0.01194005, + "auxiliary_loss_mlp": 0.01027162, + "balance_loss_clip": 1.04050684, + "balance_loss_mlp": 1.01983666, + "epoch": 0.8947273492454758, + "flos": 20084084438400.0, + "grad_norm": 2.7818347307424247, + "language_loss": 0.82399631, + "learning_rate": 1.1502491556846105e-07, + "loss": 0.84620792, + "num_input_tokens_seen": 160907890, + "step": 7441, + "time_per_iteration": 2.6469967365264893 + }, + { + "auxiliary_loss_clip": 0.01183779, + "auxiliary_loss_mlp": 0.01024962, + "balance_loss_clip": 1.04429269, + "balance_loss_mlp": 1.01797915, + "epoch": 0.894847592136115, + "flos": 18550555136640.0, + "grad_norm": 6.04793629553166, + "language_loss": 0.81440175, + "learning_rate": 1.1476469406974331e-07, + "loss": 0.83648908, + "num_input_tokens_seen": 160923490, + "step": 7442, + "time_per_iteration": 3.64709210395813 + }, + { + "auxiliary_loss_clip": 0.01164071, + "auxiliary_loss_mlp": 0.0102246, + "balance_loss_clip": 1.04664183, + "balance_loss_mlp": 1.01586223, + "epoch": 0.894967835026754, + "flos": 23478468704640.0, + "grad_norm": 1.6573343186391825, + "language_loss": 0.77000892, + "learning_rate": 1.1450475855972341e-07, + "loss": 0.79187417, + "num_input_tokens_seen": 160944280, + "step": 7443, + "time_per_iteration": 2.575216293334961 + }, + { + "auxiliary_loss_clip": 0.01182172, + "auxiliary_loss_mlp": 0.01378797, + "balance_loss_clip": 1.04254007, + "balance_loss_mlp": 1.00015461, + "epoch": 0.8950880779173931, + "flos": 15188310564480.0, + "grad_norm": 2.239408490708878, + "language_loss": 0.71017784, + "learning_rate": 1.1424510907783158e-07, + "loss": 0.73578751, + "num_input_tokens_seen": 160961560, + "step": 7444, + "time_per_iteration": 2.6507067680358887 + }, + { + "auxiliary_loss_clip": 0.01187032, + "auxiliary_loss_mlp": 0.01022937, + "balance_loss_clip": 1.04205227, + "balance_loss_mlp": 1.01591837, + "epoch": 0.8952083208080323, + "flos": 22091957769600.0, + "grad_norm": 1.6389381112664456, + "language_loss": 0.82690215, + "learning_rate": 1.1398574566345787e-07, + "loss": 0.84900188, + "num_input_tokens_seen": 160982195, + "step": 7445, + "time_per_iteration": 2.6711061000823975 + }, + { + "auxiliary_loss_clip": 0.01188791, + "auxiliary_loss_mlp": 0.01024102, + "balance_loss_clip": 1.04275155, + "balance_loss_mlp": 1.01639545, + "epoch": 0.8953285636986713, + "flos": 23254026572160.0, + "grad_norm": 2.3540334275420616, + "language_loss": 0.82500613, + "learning_rate": 1.1372666835594702e-07, + "loss": 0.84713507, + "num_input_tokens_seen": 161000520, + "step": 7446, + "time_per_iteration": 2.6456663608551025 + }, + { + "auxiliary_loss_clip": 0.01184571, + "auxiliary_loss_mlp": 0.01026572, + "balance_loss_clip": 1.04451513, + "balance_loss_mlp": 1.01922894, + "epoch": 0.8954488065893104, + "flos": 16362661818240.0, + "grad_norm": 1.8882037843754202, + "language_loss": 0.7207514, + "learning_rate": 1.1346787719460071e-07, + "loss": 0.74286282, + "num_input_tokens_seen": 161019405, + "step": 7447, + "time_per_iteration": 3.5136406421661377 + }, + { + "auxiliary_loss_clip": 0.01183425, + "auxiliary_loss_mlp": 0.01023026, + "balance_loss_clip": 1.04418683, + "balance_loss_mlp": 1.0161562, + "epoch": 0.8955690494799495, + "flos": 18257883120000.0, + "grad_norm": 1.7031757943590398, + "language_loss": 0.72130209, + "learning_rate": 1.1320937221867732e-07, + "loss": 0.7433666, + "num_input_tokens_seen": 161036985, + "step": 7448, + "time_per_iteration": 2.6259162425994873 + }, + { + "auxiliary_loss_clip": 0.01182239, + "auxiliary_loss_mlp": 0.0102209, + "balance_loss_clip": 1.04312086, + "balance_loss_mlp": 1.01574826, + "epoch": 0.8956892923705886, + "flos": 25447486498560.0, + "grad_norm": 1.9402430912514834, + "language_loss": 0.79542297, + "learning_rate": 1.1295115346739192e-07, + "loss": 0.81746626, + "num_input_tokens_seen": 161056985, + "step": 7449, + "time_per_iteration": 2.7314414978027344 + }, + { + "auxiliary_loss_clip": 0.01190616, + "auxiliary_loss_mlp": 0.01023522, + "balance_loss_clip": 1.04537737, + "balance_loss_mlp": 1.01620841, + "epoch": 0.8958095352612276, + "flos": 52661883939840.0, + "grad_norm": 2.24443434068816, + "language_loss": 0.73347205, + "learning_rate": 1.1269322097991629e-07, + "loss": 0.75561345, + "num_input_tokens_seen": 161080270, + "step": 7450, + "time_per_iteration": 2.934610366821289 + }, + { + "auxiliary_loss_clip": 0.01180549, + "auxiliary_loss_mlp": 0.01027285, + "balance_loss_clip": 1.04701567, + "balance_loss_mlp": 1.01957548, + "epoch": 0.8959297781518668, + "flos": 23186335392000.0, + "grad_norm": 2.047089722816764, + "language_loss": 0.67758149, + "learning_rate": 1.1243557479537846e-07, + "loss": 0.69965982, + "num_input_tokens_seen": 161100160, + "step": 7451, + "time_per_iteration": 2.634709596633911 + }, + { + "auxiliary_loss_clip": 0.01167673, + "auxiliary_loss_mlp": 0.01026024, + "balance_loss_clip": 1.04580045, + "balance_loss_mlp": 1.01838326, + "epoch": 0.8960500210425059, + "flos": 20334309557760.0, + "grad_norm": 2.1760950478332886, + "language_loss": 0.68686557, + "learning_rate": 1.121782149528634e-07, + "loss": 0.70880258, + "num_input_tokens_seen": 161117260, + "step": 7452, + "time_per_iteration": 2.5831198692321777 + }, + { + "auxiliary_loss_clip": 0.01190082, + "auxiliary_loss_mlp": 0.01022839, + "balance_loss_clip": 1.04609191, + "balance_loss_mlp": 1.0160861, + "epoch": 0.8961702639331449, + "flos": 19901694153600.0, + "grad_norm": 1.9928859824941243, + "language_loss": 0.789271, + "learning_rate": 1.1192114149141208e-07, + "loss": 0.81140018, + "num_input_tokens_seen": 161136895, + "step": 7453, + "time_per_iteration": 2.6633222103118896 + }, + { + "auxiliary_loss_clip": 0.01188008, + "auxiliary_loss_mlp": 0.0102554, + "balance_loss_clip": 1.04362643, + "balance_loss_mlp": 1.01761866, + "epoch": 0.8962905068237841, + "flos": 12896348567040.0, + "grad_norm": 2.310143908787794, + "language_loss": 0.65077281, + "learning_rate": 1.1166435445002197e-07, + "loss": 0.67290825, + "num_input_tokens_seen": 161154565, + "step": 7454, + "time_per_iteration": 2.6706089973449707 + }, + { + "auxiliary_loss_clip": 0.01178308, + "auxiliary_loss_mlp": 0.01027971, + "balance_loss_clip": 1.04594374, + "balance_loss_mlp": 1.02057147, + "epoch": 0.8964107497144231, + "flos": 23440331439360.0, + "grad_norm": 1.9906507729340626, + "language_loss": 0.68677408, + "learning_rate": 1.1140785386764818e-07, + "loss": 0.70883685, + "num_input_tokens_seen": 161173265, + "step": 7455, + "time_per_iteration": 2.6633384227752686 + }, + { + "auxiliary_loss_clip": 0.01170227, + "auxiliary_loss_mlp": 0.01023506, + "balance_loss_clip": 1.04534888, + "balance_loss_mlp": 1.01641893, + "epoch": 0.8965309926050622, + "flos": 19500176949120.0, + "grad_norm": 2.102630542021127, + "language_loss": 0.69381094, + "learning_rate": 1.1115163978320153e-07, + "loss": 0.71574831, + "num_input_tokens_seen": 161191995, + "step": 7456, + "time_per_iteration": 2.6693975925445557 + }, + { + "auxiliary_loss_clip": 0.01179742, + "auxiliary_loss_mlp": 0.01378606, + "balance_loss_clip": 1.04553223, + "balance_loss_mlp": 1.00016475, + "epoch": 0.8966512354957014, + "flos": 28658008022400.0, + "grad_norm": 3.2759548208690004, + "language_loss": 0.82091564, + "learning_rate": 1.1089571223554917e-07, + "loss": 0.84649909, + "num_input_tokens_seen": 161212880, + "step": 7457, + "time_per_iteration": 2.681952714920044 + }, + { + "auxiliary_loss_clip": 0.0117845, + "auxiliary_loss_mlp": 0.01022256, + "balance_loss_clip": 1.04519701, + "balance_loss_mlp": 1.01463544, + "epoch": 0.8967714783863404, + "flos": 23370916406400.0, + "grad_norm": 1.9188767404463598, + "language_loss": 0.85259521, + "learning_rate": 1.1064007126351537e-07, + "loss": 0.8746022, + "num_input_tokens_seen": 161233595, + "step": 7458, + "time_per_iteration": 2.669981002807617 + }, + { + "auxiliary_loss_clip": 0.01182946, + "auxiliary_loss_mlp": 0.01026492, + "balance_loss_clip": 1.04645896, + "balance_loss_mlp": 1.01926768, + "epoch": 0.8968917212769795, + "flos": 24535175938560.0, + "grad_norm": 1.9883198815749685, + "language_loss": 0.76269299, + "learning_rate": 1.1038471690588003e-07, + "loss": 0.78478736, + "num_input_tokens_seen": 161252740, + "step": 7459, + "time_per_iteration": 2.6492316722869873 + }, + { + "auxiliary_loss_clip": 0.01201309, + "auxiliary_loss_mlp": 0.01020949, + "balance_loss_clip": 1.04161334, + "balance_loss_mlp": 1.01434219, + "epoch": 0.8970119641676186, + "flos": 23475416048640.0, + "grad_norm": 1.970627910014143, + "language_loss": 0.79780602, + "learning_rate": 1.1012964920138145e-07, + "loss": 0.82002854, + "num_input_tokens_seen": 161272325, + "step": 7460, + "time_per_iteration": 2.7736923694610596 + }, + { + "auxiliary_loss_clip": 0.01176886, + "auxiliary_loss_mlp": 0.01023478, + "balance_loss_clip": 1.04121816, + "balance_loss_mlp": 1.01649785, + "epoch": 0.8971322070582577, + "flos": 24538192680960.0, + "grad_norm": 2.2287584397950164, + "language_loss": 0.75795519, + "learning_rate": 1.0987486818871205e-07, + "loss": 0.77995884, + "num_input_tokens_seen": 161295915, + "step": 7461, + "time_per_iteration": 2.737495183944702 + }, + { + "auxiliary_loss_clip": 0.01175425, + "auxiliary_loss_mlp": 0.0137851, + "balance_loss_clip": 1.04528165, + "balance_loss_mlp": 1.00015318, + "epoch": 0.8972524499488967, + "flos": 21797454159360.0, + "grad_norm": 2.373132974759038, + "language_loss": 0.72860718, + "learning_rate": 1.0962037390652245e-07, + "loss": 0.75414646, + "num_input_tokens_seen": 161314935, + "step": 7462, + "time_per_iteration": 3.582857608795166 + }, + { + "auxiliary_loss_clip": 0.01185594, + "auxiliary_loss_mlp": 0.01028178, + "balance_loss_clip": 1.04504585, + "balance_loss_mlp": 1.02066493, + "epoch": 0.8973726928395359, + "flos": 21726243446400.0, + "grad_norm": 1.7266057851547751, + "language_loss": 0.7184028, + "learning_rate": 1.0936616639341911e-07, + "loss": 0.7405405, + "num_input_tokens_seen": 161335225, + "step": 7463, + "time_per_iteration": 2.6895627975463867 + }, + { + "auxiliary_loss_clip": 0.01068491, + "auxiliary_loss_mlp": 0.01000545, + "balance_loss_clip": 1.01031232, + "balance_loss_mlp": 0.99944276, + "epoch": 0.897492935730175, + "flos": 53837100097920.0, + "grad_norm": 0.7347073949686159, + "language_loss": 0.54782438, + "learning_rate": 1.0911224568796473e-07, + "loss": 0.5685147, + "num_input_tokens_seen": 161393420, + "step": 7464, + "time_per_iteration": 3.2574825286865234 + }, + { + "auxiliary_loss_clip": 0.011758, + "auxiliary_loss_mlp": 0.01025661, + "balance_loss_clip": 1.04728746, + "balance_loss_mlp": 1.0186069, + "epoch": 0.897613178620814, + "flos": 18290346036480.0, + "grad_norm": 2.753960379887532, + "language_loss": 0.70993823, + "learning_rate": 1.0885861182867984e-07, + "loss": 0.73195279, + "num_input_tokens_seen": 161411525, + "step": 7465, + "time_per_iteration": 3.5700900554656982 + }, + { + "auxiliary_loss_clip": 0.01187886, + "auxiliary_loss_mlp": 0.01026485, + "balance_loss_clip": 1.04379106, + "balance_loss_mlp": 1.01935613, + "epoch": 0.8977334215114532, + "flos": 32993718059520.0, + "grad_norm": 2.029609308540845, + "language_loss": 0.71102285, + "learning_rate": 1.0860526485403942e-07, + "loss": 0.73316658, + "num_input_tokens_seen": 161432800, + "step": 7466, + "time_per_iteration": 2.7204155921936035 + }, + { + "auxiliary_loss_clip": 0.0116728, + "auxiliary_loss_mlp": 0.01026633, + "balance_loss_clip": 1.04666138, + "balance_loss_mlp": 1.01927149, + "epoch": 0.8978536644020922, + "flos": 15195636938880.0, + "grad_norm": 1.5748007539928197, + "language_loss": 0.77299231, + "learning_rate": 1.0835220480247675e-07, + "loss": 0.79493147, + "num_input_tokens_seen": 161451295, + "step": 7467, + "time_per_iteration": 3.5495851039886475 + }, + { + "auxiliary_loss_clip": 0.01184967, + "auxiliary_loss_mlp": 0.01031081, + "balance_loss_clip": 1.04607284, + "balance_loss_mlp": 1.02340996, + "epoch": 0.8979739072927313, + "flos": 18004389863040.0, + "grad_norm": 2.090204455666893, + "language_loss": 0.83573973, + "learning_rate": 1.0809943171238067e-07, + "loss": 0.85790026, + "num_input_tokens_seen": 161469220, + "step": 7468, + "time_per_iteration": 2.7024471759796143 + }, + { + "auxiliary_loss_clip": 0.0119114, + "auxiliary_loss_mlp": 0.01027043, + "balance_loss_clip": 1.04448581, + "balance_loss_mlp": 1.01867425, + "epoch": 0.8980941501833704, + "flos": 22271546793600.0, + "grad_norm": 1.863156869020839, + "language_loss": 0.62291807, + "learning_rate": 1.078469456220965e-07, + "loss": 0.64509988, + "num_input_tokens_seen": 161489375, + "step": 7469, + "time_per_iteration": 2.672053337097168 + }, + { + "auxiliary_loss_clip": 0.01175981, + "auxiliary_loss_mlp": 0.01021348, + "balance_loss_clip": 1.04392791, + "balance_loss_mlp": 1.01434469, + "epoch": 0.8982143930740095, + "flos": 37560729726720.0, + "grad_norm": 1.6189762053834733, + "language_loss": 0.69714534, + "learning_rate": 1.0759474656992606e-07, + "loss": 0.71911865, + "num_input_tokens_seen": 161512145, + "step": 7470, + "time_per_iteration": 2.7869186401367188 + }, + { + "auxiliary_loss_clip": 0.01188236, + "auxiliary_loss_mlp": 0.01032687, + "balance_loss_clip": 1.04237056, + "balance_loss_mlp": 1.02472425, + "epoch": 0.8983346359646486, + "flos": 18076893465600.0, + "grad_norm": 2.5606181161789148, + "language_loss": 0.78133637, + "learning_rate": 1.0734283459412785e-07, + "loss": 0.80354559, + "num_input_tokens_seen": 161528995, + "step": 7471, + "time_per_iteration": 2.6055779457092285 + }, + { + "auxiliary_loss_clip": 0.01206376, + "auxiliary_loss_mlp": 0.01027174, + "balance_loss_clip": 1.04108214, + "balance_loss_mlp": 1.01937747, + "epoch": 0.8984548788552876, + "flos": 20558895344640.0, + "grad_norm": 1.6980786143782771, + "language_loss": 0.80552316, + "learning_rate": 1.0709120973291707e-07, + "loss": 0.82785869, + "num_input_tokens_seen": 161548775, + "step": 7472, + "time_per_iteration": 2.703040838241577 + }, + { + "auxiliary_loss_clip": 0.0117008, + "auxiliary_loss_mlp": 0.01029175, + "balance_loss_clip": 1.04833949, + "balance_loss_mlp": 1.02184367, + "epoch": 0.8985751217459268, + "flos": 17785442511360.0, + "grad_norm": 2.010519162686434, + "language_loss": 0.77861929, + "learning_rate": 1.0683987202446475e-07, + "loss": 0.80061191, + "num_input_tokens_seen": 161566960, + "step": 7473, + "time_per_iteration": 3.5524089336395264 + }, + { + "auxiliary_loss_clip": 0.01181488, + "auxiliary_loss_mlp": 0.01022311, + "balance_loss_clip": 1.04698873, + "balance_loss_mlp": 1.0147711, + "epoch": 0.8986953646365659, + "flos": 21617003208960.0, + "grad_norm": 1.8659970004330326, + "language_loss": 0.69844562, + "learning_rate": 1.0658882150689862e-07, + "loss": 0.7204836, + "num_input_tokens_seen": 161585820, + "step": 7474, + "time_per_iteration": 2.6034116744995117 + }, + { + "auxiliary_loss_clip": 0.01196446, + "auxiliary_loss_mlp": 0.01030608, + "balance_loss_clip": 1.04234886, + "balance_loss_mlp": 1.02310395, + "epoch": 0.8988156075272049, + "flos": 14027355083520.0, + "grad_norm": 3.60246569783657, + "language_loss": 0.78494114, + "learning_rate": 1.0633805821830288e-07, + "loss": 0.8072117, + "num_input_tokens_seen": 161602505, + "step": 7475, + "time_per_iteration": 2.6457509994506836 + }, + { + "auxiliary_loss_clip": 0.0118806, + "auxiliary_loss_mlp": 0.01025045, + "balance_loss_clip": 1.04617131, + "balance_loss_mlp": 1.01779425, + "epoch": 0.8989358504178441, + "flos": 29059202004480.0, + "grad_norm": 2.655526778269924, + "language_loss": 0.82825965, + "learning_rate": 1.0608758219671753e-07, + "loss": 0.85039073, + "num_input_tokens_seen": 161621545, + "step": 7476, + "time_per_iteration": 2.6855006217956543 + }, + { + "auxiliary_loss_clip": 0.01189893, + "auxiliary_loss_mlp": 0.01023274, + "balance_loss_clip": 1.04441118, + "balance_loss_mlp": 1.01618099, + "epoch": 0.8990560933084831, + "flos": 20230420446720.0, + "grad_norm": 1.735102325139507, + "language_loss": 0.7073741, + "learning_rate": 1.0583739348014065e-07, + "loss": 0.72950578, + "num_input_tokens_seen": 161642630, + "step": 7477, + "time_per_iteration": 2.643251419067383 + }, + { + "auxiliary_loss_clip": 0.01169145, + "auxiliary_loss_mlp": 0.01020424, + "balance_loss_clip": 1.04859757, + "balance_loss_mlp": 1.01377785, + "epoch": 0.8991763361991222, + "flos": 25520672459520.0, + "grad_norm": 3.118074324405277, + "language_loss": 0.84522337, + "learning_rate": 1.0558749210652518e-07, + "loss": 0.86711913, + "num_input_tokens_seen": 161662560, + "step": 7478, + "time_per_iteration": 2.5579888820648193 + }, + { + "auxiliary_loss_clip": 0.0119652, + "auxiliary_loss_mlp": 0.01022508, + "balance_loss_clip": 1.04299557, + "balance_loss_mlp": 1.01524198, + "epoch": 0.8992965790897613, + "flos": 25119191168640.0, + "grad_norm": 1.6415183906507982, + "language_loss": 0.85747313, + "learning_rate": 1.053378781137808e-07, + "loss": 0.87966341, + "num_input_tokens_seen": 161683480, + "step": 7479, + "time_per_iteration": 2.719402313232422 + }, + { + "auxiliary_loss_clip": 0.01186426, + "auxiliary_loss_mlp": 0.01025368, + "balance_loss_clip": 1.04270792, + "balance_loss_mlp": 1.017488, + "epoch": 0.8994168219804004, + "flos": 16070815814400.0, + "grad_norm": 1.8032336699033353, + "language_loss": 0.77792966, + "learning_rate": 1.0508855153977392e-07, + "loss": 0.80004764, + "num_input_tokens_seen": 161699945, + "step": 7480, + "time_per_iteration": 2.5688161849975586 + }, + { + "auxiliary_loss_clip": 0.01173792, + "auxiliary_loss_mlp": 0.01027147, + "balance_loss_clip": 1.04205203, + "balance_loss_mlp": 1.01965499, + "epoch": 0.8995370648710395, + "flos": 24825764966400.0, + "grad_norm": 2.2310106499637437, + "language_loss": 0.66859555, + "learning_rate": 1.0483951242232669e-07, + "loss": 0.69060493, + "num_input_tokens_seen": 161720420, + "step": 7481, + "time_per_iteration": 2.654738664627075 + }, + { + "auxiliary_loss_clip": 0.01060506, + "auxiliary_loss_mlp": 0.01000566, + "balance_loss_clip": 1.00916827, + "balance_loss_mlp": 0.9994632, + "epoch": 0.8996573077616786, + "flos": 63116238378240.0, + "grad_norm": 0.9789878537947024, + "language_loss": 0.57745945, + "learning_rate": 1.0459076079921936e-07, + "loss": 0.59807014, + "num_input_tokens_seen": 161773080, + "step": 7482, + "time_per_iteration": 3.2098171710968018 + }, + { + "auxiliary_loss_clip": 0.01177837, + "auxiliary_loss_mlp": 0.0102579, + "balance_loss_clip": 1.0431906, + "balance_loss_mlp": 1.01869369, + "epoch": 0.8997775506523177, + "flos": 18219674027520.0, + "grad_norm": 3.34382244649808, + "language_loss": 0.85487688, + "learning_rate": 1.0434229670818618e-07, + "loss": 0.87691319, + "num_input_tokens_seen": 161789755, + "step": 7483, + "time_per_iteration": 2.7286550998687744 + }, + { + "auxiliary_loss_clip": 0.01176412, + "auxiliary_loss_mlp": 0.01025592, + "balance_loss_clip": 1.04294348, + "balance_loss_mlp": 1.01834369, + "epoch": 0.8998977935429567, + "flos": 24166768095360.0, + "grad_norm": 1.6296843731102268, + "language_loss": 0.80099326, + "learning_rate": 1.0409412018691944e-07, + "loss": 0.82301331, + "num_input_tokens_seen": 161810220, + "step": 7484, + "time_per_iteration": 2.6970012187957764 + }, + { + "auxiliary_loss_clip": 0.01180467, + "auxiliary_loss_mlp": 0.0102839, + "balance_loss_clip": 1.04449153, + "balance_loss_mlp": 1.02099311, + "epoch": 0.9000180364335959, + "flos": 20773030273920.0, + "grad_norm": 1.8842043928442447, + "language_loss": 0.74691606, + "learning_rate": 1.0384623127306724e-07, + "loss": 0.76900464, + "num_input_tokens_seen": 161827565, + "step": 7485, + "time_per_iteration": 2.6698803901672363 + }, + { + "auxiliary_loss_clip": 0.01190383, + "auxiliary_loss_mlp": 0.01024707, + "balance_loss_clip": 1.04139125, + "balance_loss_mlp": 1.0177722, + "epoch": 0.900138279324235, + "flos": 19205745166080.0, + "grad_norm": 1.7808711707110252, + "language_loss": 0.7927441, + "learning_rate": 1.0359863000423397e-07, + "loss": 0.81489503, + "num_input_tokens_seen": 161845700, + "step": 7486, + "time_per_iteration": 2.696153163909912 + }, + { + "auxiliary_loss_clip": 0.01170106, + "auxiliary_loss_mlp": 0.01026191, + "balance_loss_clip": 1.04840505, + "balance_loss_mlp": 1.01898432, + "epoch": 0.900258522214874, + "flos": 28731158069760.0, + "grad_norm": 1.5653336323020546, + "language_loss": 0.71777582, + "learning_rate": 1.0335131641798112e-07, + "loss": 0.73973876, + "num_input_tokens_seen": 161867660, + "step": 7487, + "time_per_iteration": 2.6452906131744385 + }, + { + "auxiliary_loss_clip": 0.01085051, + "auxiliary_loss_mlp": 0.00999749, + "balance_loss_clip": 1.00814199, + "balance_loss_mlp": 0.99873602, + "epoch": 0.9003787651055132, + "flos": 58280685655680.0, + "grad_norm": 0.8049341826240556, + "language_loss": 0.55604517, + "learning_rate": 1.0310429055182512e-07, + "loss": 0.57689315, + "num_input_tokens_seen": 161921980, + "step": 7488, + "time_per_iteration": 3.977370500564575 + }, + { + "auxiliary_loss_clip": 0.01197709, + "auxiliary_loss_mlp": 0.01030043, + "balance_loss_clip": 1.04517245, + "balance_loss_mlp": 1.0222224, + "epoch": 0.9004990079961522, + "flos": 25556475340800.0, + "grad_norm": 1.673028642399146, + "language_loss": 0.73931044, + "learning_rate": 1.0285755244324024e-07, + "loss": 0.76158798, + "num_input_tokens_seen": 161942725, + "step": 7489, + "time_per_iteration": 2.7437379360198975 + }, + { + "auxiliary_loss_clip": 0.01186954, + "auxiliary_loss_mlp": 0.01378277, + "balance_loss_clip": 1.04157948, + "balance_loss_mlp": 1.0001471, + "epoch": 0.9006192508867913, + "flos": 23335185352320.0, + "grad_norm": 1.7606926915898862, + "language_loss": 0.68638808, + "learning_rate": 1.0261110212965629e-07, + "loss": 0.71204042, + "num_input_tokens_seen": 161964520, + "step": 7490, + "time_per_iteration": 2.7569732666015625 + }, + { + "auxiliary_loss_clip": 0.01182782, + "auxiliary_loss_mlp": 0.01025075, + "balance_loss_clip": 1.0427525, + "balance_loss_mlp": 1.01800907, + "epoch": 0.9007394937774305, + "flos": 18040300485120.0, + "grad_norm": 3.3788811294949763, + "language_loss": 0.7905587, + "learning_rate": 1.023649396484596e-07, + "loss": 0.81263733, + "num_input_tokens_seen": 161983575, + "step": 7491, + "time_per_iteration": 3.617690086364746 + }, + { + "auxiliary_loss_clip": 0.01167427, + "auxiliary_loss_mlp": 0.01025143, + "balance_loss_clip": 1.04664218, + "balance_loss_mlp": 1.01811838, + "epoch": 0.9008597366680695, + "flos": 43068456633600.0, + "grad_norm": 3.4722689142032155, + "language_loss": 0.67840862, + "learning_rate": 1.0211906503699275e-07, + "loss": 0.70033431, + "num_input_tokens_seen": 162006550, + "step": 7492, + "time_per_iteration": 2.7503480911254883 + }, + { + "auxiliary_loss_clip": 0.01179358, + "auxiliary_loss_mlp": 0.01028303, + "balance_loss_clip": 1.04838562, + "balance_loss_mlp": 1.02083123, + "epoch": 0.9009799795587086, + "flos": 14939055112320.0, + "grad_norm": 2.3898699049090437, + "language_loss": 0.82653272, + "learning_rate": 1.0187347833255455e-07, + "loss": 0.84860933, + "num_input_tokens_seen": 162022455, + "step": 7493, + "time_per_iteration": 2.5890071392059326 + }, + { + "auxiliary_loss_clip": 0.01165978, + "auxiliary_loss_mlp": 0.0102602, + "balance_loss_clip": 1.04767871, + "balance_loss_mlp": 1.01887631, + "epoch": 0.9011002224493477, + "flos": 21579584215680.0, + "grad_norm": 1.8440271683154408, + "language_loss": 0.79590958, + "learning_rate": 1.0162817957240056e-07, + "loss": 0.81782961, + "num_input_tokens_seen": 162042350, + "step": 7494, + "time_per_iteration": 3.471806526184082 + }, + { + "auxiliary_loss_clip": 0.01074599, + "auxiliary_loss_mlp": 0.00998678, + "balance_loss_clip": 1.00988793, + "balance_loss_mlp": 0.99755716, + "epoch": 0.9012204653399868, + "flos": 71166367883520.0, + "grad_norm": 0.880163178035669, + "language_loss": 0.6301676, + "learning_rate": 1.0138316879374253e-07, + "loss": 0.65090036, + "num_input_tokens_seen": 162111640, + "step": 7495, + "time_per_iteration": 3.3351902961730957 + }, + { + "auxiliary_loss_clip": 0.01188381, + "auxiliary_loss_mlp": 0.01024236, + "balance_loss_clip": 1.04767656, + "balance_loss_mlp": 1.01688385, + "epoch": 0.9013407082306258, + "flos": 15594963413760.0, + "grad_norm": 2.329519531001152, + "language_loss": 0.74288797, + "learning_rate": 1.0113844603374833e-07, + "loss": 0.76501417, + "num_input_tokens_seen": 162128165, + "step": 7496, + "time_per_iteration": 2.769200563430786 + }, + { + "auxiliary_loss_clip": 0.01181686, + "auxiliary_loss_mlp": 0.01025147, + "balance_loss_clip": 1.04212749, + "balance_loss_mlp": 1.01757979, + "epoch": 0.901460951121265, + "flos": 15049157276160.0, + "grad_norm": 2.2056822546856125, + "language_loss": 0.72105092, + "learning_rate": 1.0089401132954178e-07, + "loss": 0.74311924, + "num_input_tokens_seen": 162146145, + "step": 7497, + "time_per_iteration": 2.6305058002471924 + }, + { + "auxiliary_loss_clip": 0.01187666, + "auxiliary_loss_mlp": 0.01022119, + "balance_loss_clip": 1.04740179, + "balance_loss_mlp": 1.01553559, + "epoch": 0.9015811940119041, + "flos": 22236857233920.0, + "grad_norm": 1.8380111829470138, + "language_loss": 0.72542864, + "learning_rate": 1.006498647182037e-07, + "loss": 0.74752653, + "num_input_tokens_seen": 162164800, + "step": 7498, + "time_per_iteration": 2.705444097518921 + }, + { + "auxiliary_loss_clip": 0.0120595, + "auxiliary_loss_mlp": 0.0102875, + "balance_loss_clip": 1.03807843, + "balance_loss_mlp": 1.02160335, + "epoch": 0.9017014369025431, + "flos": 24973824827520.0, + "grad_norm": 2.0510500896009343, + "language_loss": 0.71571529, + "learning_rate": 1.004060062367713e-07, + "loss": 0.73806226, + "num_input_tokens_seen": 162185895, + "step": 7499, + "time_per_iteration": 3.760711908340454 + }, + { + "auxiliary_loss_clip": 0.01176459, + "auxiliary_loss_mlp": 0.01037754, + "balance_loss_clip": 1.04388547, + "balance_loss_mlp": 1.03056586, + "epoch": 0.9018216797931822, + "flos": 18114168804480.0, + "grad_norm": 2.2404890588458044, + "language_loss": 0.69783938, + "learning_rate": 1.0016243592223728e-07, + "loss": 0.71998155, + "num_input_tokens_seen": 162206295, + "step": 7500, + "time_per_iteration": 2.886227607727051 + }, + { + "auxiliary_loss_clip": 0.0120967, + "auxiliary_loss_mlp": 0.01022692, + "balance_loss_clip": 1.0410887, + "balance_loss_mlp": 1.01568222, + "epoch": 0.9019419226838213, + "flos": 37268452759680.0, + "grad_norm": 8.93439090790845, + "language_loss": 0.65625685, + "learning_rate": 9.991915381155114e-08, + "loss": 0.67858046, + "num_input_tokens_seen": 162229275, + "step": 7501, + "time_per_iteration": 2.893756151199341 + }, + { + "auxiliary_loss_clip": 0.0117951, + "auxiliary_loss_mlp": 0.01035021, + "balance_loss_clip": 1.04603326, + "balance_loss_mlp": 1.02766025, + "epoch": 0.9020621655744604, + "flos": 23441121538560.0, + "grad_norm": 4.403933219825099, + "language_loss": 0.75427169, + "learning_rate": 9.967615994161871e-08, + "loss": 0.77641702, + "num_input_tokens_seen": 162248935, + "step": 7502, + "time_per_iteration": 2.6002585887908936 + }, + { + "auxiliary_loss_clip": 0.0116523, + "auxiliary_loss_mlp": 0.01024363, + "balance_loss_clip": 1.04625595, + "balance_loss_mlp": 1.01740408, + "epoch": 0.9021824084650995, + "flos": 22857465444480.0, + "grad_norm": 2.335439373324177, + "language_loss": 0.78271484, + "learning_rate": 9.943345434930161e-08, + "loss": 0.80461085, + "num_input_tokens_seen": 162269185, + "step": 7503, + "time_per_iteration": 2.584153175354004 + }, + { + "auxiliary_loss_clip": 0.01197369, + "auxiliary_loss_mlp": 0.01022051, + "balance_loss_clip": 1.04702067, + "balance_loss_mlp": 1.01532507, + "epoch": 0.9023026513557386, + "flos": 22127581082880.0, + "grad_norm": 2.0053151518659873, + "language_loss": 0.69253677, + "learning_rate": 9.919103707141885e-08, + "loss": 0.71473098, + "num_input_tokens_seen": 162288065, + "step": 7504, + "time_per_iteration": 2.633134603500366 + }, + { + "auxiliary_loss_clip": 0.01174852, + "auxiliary_loss_mlp": 0.01028323, + "balance_loss_clip": 1.04555142, + "balance_loss_mlp": 1.02021074, + "epoch": 0.9024228942463777, + "flos": 24199087357440.0, + "grad_norm": 1.8828108681588915, + "language_loss": 0.76374114, + "learning_rate": 9.89489081447441e-08, + "loss": 0.78577292, + "num_input_tokens_seen": 162305265, + "step": 7505, + "time_per_iteration": 2.6134743690490723 + }, + { + "auxiliary_loss_clip": 0.01181425, + "auxiliary_loss_mlp": 0.01024995, + "balance_loss_clip": 1.04237533, + "balance_loss_mlp": 1.0174464, + "epoch": 0.9025431371370167, + "flos": 25008262992000.0, + "grad_norm": 1.8527989939082767, + "language_loss": 0.83114159, + "learning_rate": 9.870706760600844e-08, + "loss": 0.8532058, + "num_input_tokens_seen": 162325215, + "step": 7506, + "time_per_iteration": 2.6387202739715576 + }, + { + "auxiliary_loss_clip": 0.01209375, + "auxiliary_loss_mlp": 0.01024336, + "balance_loss_clip": 1.04501367, + "balance_loss_mlp": 1.01721013, + "epoch": 0.9026633800276559, + "flos": 18952862440320.0, + "grad_norm": 1.8852521217962663, + "language_loss": 0.72590721, + "learning_rate": 9.846551549189918e-08, + "loss": 0.7482444, + "num_input_tokens_seen": 162344820, + "step": 7507, + "time_per_iteration": 2.684661865234375 + }, + { + "auxiliary_loss_clip": 0.0118216, + "auxiliary_loss_mlp": 0.01024908, + "balance_loss_clip": 1.04406011, + "balance_loss_mlp": 1.01683152, + "epoch": 0.902783622918295, + "flos": 32416059536640.0, + "grad_norm": 2.036300121295957, + "language_loss": 0.68607616, + "learning_rate": 9.822425183905902e-08, + "loss": 0.70814681, + "num_input_tokens_seen": 162365345, + "step": 7508, + "time_per_iteration": 2.7039647102355957 + }, + { + "auxiliary_loss_clip": 0.01099584, + "auxiliary_loss_mlp": 0.00999588, + "balance_loss_clip": 1.0091145, + "balance_loss_mlp": 0.99850315, + "epoch": 0.902903865808934, + "flos": 63717453244800.0, + "grad_norm": 0.9201433591254733, + "language_loss": 0.75182033, + "learning_rate": 9.798327668408823e-08, + "loss": 0.77281201, + "num_input_tokens_seen": 162426980, + "step": 7509, + "time_per_iteration": 3.3508803844451904 + }, + { + "auxiliary_loss_clip": 0.01171574, + "auxiliary_loss_mlp": 0.01021964, + "balance_loss_clip": 1.0478183, + "balance_loss_mlp": 1.01386929, + "epoch": 0.9030241086995732, + "flos": 23804034600960.0, + "grad_norm": 1.856583261228961, + "language_loss": 0.68841255, + "learning_rate": 9.774259006354158e-08, + "loss": 0.71034789, + "num_input_tokens_seen": 162447050, + "step": 7510, + "time_per_iteration": 2.634866952896118 + }, + { + "auxiliary_loss_clip": 0.01188329, + "auxiliary_loss_mlp": 0.0102633, + "balance_loss_clip": 1.04247975, + "balance_loss_mlp": 1.01901937, + "epoch": 0.9031443515902122, + "flos": 26395887248640.0, + "grad_norm": 1.8762552479417443, + "language_loss": 0.76420397, + "learning_rate": 9.750219201393184e-08, + "loss": 0.78635055, + "num_input_tokens_seen": 162467015, + "step": 7511, + "time_per_iteration": 2.728102922439575 + }, + { + "auxiliary_loss_clip": 0.01175937, + "auxiliary_loss_mlp": 0.01023834, + "balance_loss_clip": 1.04538071, + "balance_loss_mlp": 1.01663959, + "epoch": 0.9032645944808513, + "flos": 24939350749440.0, + "grad_norm": 2.259126740407888, + "language_loss": 0.77717137, + "learning_rate": 9.726208257172697e-08, + "loss": 0.79916906, + "num_input_tokens_seen": 162488710, + "step": 7512, + "time_per_iteration": 2.720759153366089 + }, + { + "auxiliary_loss_clip": 0.01167229, + "auxiliary_loss_mlp": 0.01024558, + "balance_loss_clip": 1.04761636, + "balance_loss_mlp": 1.01736331, + "epoch": 0.9033848373714904, + "flos": 21178821196800.0, + "grad_norm": 2.031488832055984, + "language_loss": 0.74801594, + "learning_rate": 9.702226177335115e-08, + "loss": 0.76993382, + "num_input_tokens_seen": 162507205, + "step": 7513, + "time_per_iteration": 2.535517692565918 + }, + { + "auxiliary_loss_clip": 0.01184927, + "auxiliary_loss_mlp": 0.01028941, + "balance_loss_clip": 1.0456593, + "balance_loss_mlp": 1.02146113, + "epoch": 0.9035050802621295, + "flos": 26286359702400.0, + "grad_norm": 1.6389587421253673, + "language_loss": 0.72500932, + "learning_rate": 9.67827296551853e-08, + "loss": 0.74714804, + "num_input_tokens_seen": 162528490, + "step": 7514, + "time_per_iteration": 3.5554039478302 + }, + { + "auxiliary_loss_clip": 0.01174457, + "auxiliary_loss_mlp": 0.01378545, + "balance_loss_clip": 1.04147458, + "balance_loss_mlp": 1.00016618, + "epoch": 0.9036253231527686, + "flos": 24204546224640.0, + "grad_norm": 1.866065458921758, + "language_loss": 0.68656552, + "learning_rate": 9.65434862535659e-08, + "loss": 0.7120955, + "num_input_tokens_seen": 162547860, + "step": 7515, + "time_per_iteration": 2.7525579929351807 + }, + { + "auxiliary_loss_clip": 0.01189025, + "auxiliary_loss_mlp": 0.01025945, + "balance_loss_clip": 1.04500508, + "balance_loss_mlp": 1.0188998, + "epoch": 0.9037455660434077, + "flos": 18072655660800.0, + "grad_norm": 3.70849214413223, + "language_loss": 0.64976585, + "learning_rate": 9.630453160478635e-08, + "loss": 0.67191553, + "num_input_tokens_seen": 162563215, + "step": 7516, + "time_per_iteration": 2.6192662715911865 + }, + { + "auxiliary_loss_clip": 0.01199501, + "auxiliary_loss_mlp": 0.01025384, + "balance_loss_clip": 1.04094744, + "balance_loss_mlp": 1.01891088, + "epoch": 0.9038658089340468, + "flos": 24060795995520.0, + "grad_norm": 1.611398428842578, + "language_loss": 0.82306588, + "learning_rate": 9.60658657450959e-08, + "loss": 0.84531474, + "num_input_tokens_seen": 162583515, + "step": 7517, + "time_per_iteration": 3.5710184574127197 + }, + { + "auxiliary_loss_clip": 0.01168258, + "auxiliary_loss_mlp": 0.01020363, + "balance_loss_clip": 1.03944147, + "balance_loss_mlp": 1.01323736, + "epoch": 0.9039860518246858, + "flos": 21834298535040.0, + "grad_norm": 1.9455105023814665, + "language_loss": 0.7946682, + "learning_rate": 9.582748871069979e-08, + "loss": 0.81655443, + "num_input_tokens_seen": 162602955, + "step": 7518, + "time_per_iteration": 2.605823516845703 + }, + { + "auxiliary_loss_clip": 0.01187761, + "auxiliary_loss_mlp": 0.01378399, + "balance_loss_clip": 1.04350162, + "balance_loss_mlp": 1.00012898, + "epoch": 0.904106294715325, + "flos": 26614870513920.0, + "grad_norm": 1.7894831126900532, + "language_loss": 0.83230066, + "learning_rate": 9.558940053775954e-08, + "loss": 0.85796225, + "num_input_tokens_seen": 162621595, + "step": 7519, + "time_per_iteration": 3.5652706623077393 + }, + { + "auxiliary_loss_clip": 0.01177074, + "auxiliary_loss_mlp": 0.010223, + "balance_loss_clip": 1.04629254, + "balance_loss_mlp": 1.01481378, + "epoch": 0.904226537605964, + "flos": 17785693906560.0, + "grad_norm": 2.0750362621959444, + "language_loss": 0.67769676, + "learning_rate": 9.535160126239294e-08, + "loss": 0.69969058, + "num_input_tokens_seen": 162638220, + "step": 7520, + "time_per_iteration": 2.639971971511841 + }, + { + "auxiliary_loss_clip": 0.01176896, + "auxiliary_loss_mlp": 0.01026622, + "balance_loss_clip": 1.04764915, + "balance_loss_mlp": 1.01923692, + "epoch": 0.9043467804966031, + "flos": 24790428961920.0, + "grad_norm": 2.1299100209625155, + "language_loss": 0.70821178, + "learning_rate": 9.511409092067424e-08, + "loss": 0.7302469, + "num_input_tokens_seen": 162658575, + "step": 7521, + "time_per_iteration": 2.667022466659546 + }, + { + "auxiliary_loss_clip": 0.01184607, + "auxiliary_loss_mlp": 0.0102622, + "balance_loss_clip": 1.04559231, + "balance_loss_mlp": 1.01877856, + "epoch": 0.9044670233872423, + "flos": 22632125472000.0, + "grad_norm": 1.9697808178096763, + "language_loss": 0.6758275, + "learning_rate": 9.487686954863327e-08, + "loss": 0.69793576, + "num_input_tokens_seen": 162678295, + "step": 7522, + "time_per_iteration": 2.6961302757263184 + }, + { + "auxiliary_loss_clip": 0.01176678, + "auxiliary_loss_mlp": 0.01022107, + "balance_loss_clip": 1.04761291, + "balance_loss_mlp": 1.01524019, + "epoch": 0.9045872662778813, + "flos": 23771320289280.0, + "grad_norm": 2.1058066908541577, + "language_loss": 0.77545971, + "learning_rate": 9.46399371822566e-08, + "loss": 0.79744756, + "num_input_tokens_seen": 162698070, + "step": 7523, + "time_per_iteration": 2.6690258979797363 + }, + { + "auxiliary_loss_clip": 0.0116851, + "auxiliary_loss_mlp": 0.01031719, + "balance_loss_clip": 1.04771399, + "balance_loss_mlp": 1.02425957, + "epoch": 0.9047075091685204, + "flos": 15191039998080.0, + "grad_norm": 2.086533018125026, + "language_loss": 0.72371364, + "learning_rate": 9.440329385748657e-08, + "loss": 0.74571598, + "num_input_tokens_seen": 162715140, + "step": 7524, + "time_per_iteration": 2.5603513717651367 + }, + { + "auxiliary_loss_clip": 0.01193828, + "auxiliary_loss_mlp": 0.0102368, + "balance_loss_clip": 1.04317987, + "balance_loss_mlp": 1.0173347, + "epoch": 0.9048277520591596, + "flos": 18003707504640.0, + "grad_norm": 1.765837515040909, + "language_loss": 0.70638633, + "learning_rate": 9.416693961022137e-08, + "loss": 0.7285614, + "num_input_tokens_seen": 162733390, + "step": 7525, + "time_per_iteration": 3.4955124855041504 + }, + { + "auxiliary_loss_clip": 0.01195122, + "auxiliary_loss_mlp": 0.0102613, + "balance_loss_clip": 1.0386008, + "balance_loss_mlp": 1.01900697, + "epoch": 0.9049479949497986, + "flos": 21872471713920.0, + "grad_norm": 1.723363412651695, + "language_loss": 0.76934481, + "learning_rate": 9.393087447631654e-08, + "loss": 0.79155731, + "num_input_tokens_seen": 162751670, + "step": 7526, + "time_per_iteration": 2.7604315280914307 + }, + { + "auxiliary_loss_clip": 0.01187709, + "auxiliary_loss_mlp": 0.01020413, + "balance_loss_clip": 1.04451919, + "balance_loss_mlp": 1.01383317, + "epoch": 0.9050682378404377, + "flos": 20773928113920.0, + "grad_norm": 1.848121471473049, + "language_loss": 0.72925246, + "learning_rate": 9.36950984915823e-08, + "loss": 0.75133371, + "num_input_tokens_seen": 162770025, + "step": 7527, + "time_per_iteration": 2.6612708568573 + }, + { + "auxiliary_loss_clip": 0.01170524, + "auxiliary_loss_mlp": 0.01022528, + "balance_loss_clip": 1.04929185, + "balance_loss_mlp": 1.01452303, + "epoch": 0.9051884807310768, + "flos": 21580015178880.0, + "grad_norm": 2.0733761376908575, + "language_loss": 0.69260132, + "learning_rate": 9.345961169178607e-08, + "loss": 0.71453178, + "num_input_tokens_seen": 162789710, + "step": 7528, + "time_per_iteration": 2.6359379291534424 + }, + { + "auxiliary_loss_clip": 0.01177826, + "auxiliary_loss_mlp": 0.01026637, + "balance_loss_clip": 1.04348779, + "balance_loss_mlp": 1.01901972, + "epoch": 0.9053087236217159, + "flos": 21908059113600.0, + "grad_norm": 1.4880874795702548, + "language_loss": 0.72623861, + "learning_rate": 9.322441411265081e-08, + "loss": 0.74828321, + "num_input_tokens_seen": 162810695, + "step": 7529, + "time_per_iteration": 2.677464485168457 + }, + { + "auxiliary_loss_clip": 0.01177759, + "auxiliary_loss_mlp": 0.0103023, + "balance_loss_clip": 1.04338658, + "balance_loss_mlp": 1.02297628, + "epoch": 0.9054289665123549, + "flos": 17055809544960.0, + "grad_norm": 1.816434028616128, + "language_loss": 0.73401296, + "learning_rate": 9.298950578985554e-08, + "loss": 0.75609291, + "num_input_tokens_seen": 162827770, + "step": 7530, + "time_per_iteration": 2.63997745513916 + }, + { + "auxiliary_loss_clip": 0.01172794, + "auxiliary_loss_mlp": 0.01378803, + "balance_loss_clip": 1.04605937, + "balance_loss_mlp": 1.00012779, + "epoch": 0.905549209402994, + "flos": 20777268078720.0, + "grad_norm": 1.994732607833962, + "language_loss": 0.71244973, + "learning_rate": 9.275488675903665e-08, + "loss": 0.7379657, + "num_input_tokens_seen": 162846715, + "step": 7531, + "time_per_iteration": 2.620920181274414 + }, + { + "auxiliary_loss_clip": 0.01201923, + "auxiliary_loss_mlp": 0.01024148, + "balance_loss_clip": 1.0422014, + "balance_loss_mlp": 1.01687956, + "epoch": 0.9056694522936332, + "flos": 21686813291520.0, + "grad_norm": 1.999374068697336, + "language_loss": 0.74266106, + "learning_rate": 9.252055705578454e-08, + "loss": 0.76492167, + "num_input_tokens_seen": 162866215, + "step": 7532, + "time_per_iteration": 2.716104030609131 + }, + { + "auxiliary_loss_clip": 0.01173825, + "auxiliary_loss_mlp": 0.01021458, + "balance_loss_clip": 1.04411566, + "balance_loss_mlp": 1.01452851, + "epoch": 0.9057896951842722, + "flos": 29569133433600.0, + "grad_norm": 1.8069734996854767, + "language_loss": 0.72151488, + "learning_rate": 9.228651671564747e-08, + "loss": 0.74346769, + "num_input_tokens_seen": 162888245, + "step": 7533, + "time_per_iteration": 2.6471152305603027 + }, + { + "auxiliary_loss_clip": 0.01195688, + "auxiliary_loss_mlp": 0.01025939, + "balance_loss_clip": 1.0418781, + "balance_loss_mlp": 1.01903391, + "epoch": 0.9059099380749113, + "flos": 27892248952320.0, + "grad_norm": 1.4638602658709543, + "language_loss": 0.78042769, + "learning_rate": 9.205276577412901e-08, + "loss": 0.80264401, + "num_input_tokens_seen": 162911025, + "step": 7534, + "time_per_iteration": 2.7776644229888916 + }, + { + "auxiliary_loss_clip": 0.01191287, + "auxiliary_loss_mlp": 0.01378765, + "balance_loss_clip": 1.0429405, + "balance_loss_mlp": 1.0001359, + "epoch": 0.9060301809655504, + "flos": 17748993185280.0, + "grad_norm": 2.777198113816532, + "language_loss": 0.77074981, + "learning_rate": 9.181930426668905e-08, + "loss": 0.79645032, + "num_input_tokens_seen": 162927820, + "step": 7535, + "time_per_iteration": 2.6688454151153564 + }, + { + "auxiliary_loss_clip": 0.01194951, + "auxiliary_loss_mlp": 0.01030255, + "balance_loss_clip": 1.03950763, + "balance_loss_mlp": 1.02357316, + "epoch": 0.9061504238561895, + "flos": 31759432963200.0, + "grad_norm": 1.7294764281663726, + "language_loss": 0.67870015, + "learning_rate": 9.158613222874346e-08, + "loss": 0.70095223, + "num_input_tokens_seen": 162949445, + "step": 7536, + "time_per_iteration": 2.7573442459106445 + }, + { + "auxiliary_loss_clip": 0.01184445, + "auxiliary_loss_mlp": 0.01026316, + "balance_loss_clip": 1.04372716, + "balance_loss_mlp": 1.0192771, + "epoch": 0.9062706667468285, + "flos": 20048066075520.0, + "grad_norm": 1.6043145084298565, + "language_loss": 0.81930828, + "learning_rate": 9.135324969566394e-08, + "loss": 0.84141588, + "num_input_tokens_seen": 162968945, + "step": 7537, + "time_per_iteration": 2.6777119636535645 + }, + { + "auxiliary_loss_clip": 0.01181215, + "auxiliary_loss_mlp": 0.01026732, + "balance_loss_clip": 1.04746222, + "balance_loss_mlp": 1.01938319, + "epoch": 0.9063909096374677, + "flos": 18437292576000.0, + "grad_norm": 1.94031533206056, + "language_loss": 0.75610614, + "learning_rate": 9.112065670277913e-08, + "loss": 0.77818555, + "num_input_tokens_seen": 162985310, + "step": 7538, + "time_per_iteration": 2.57731032371521 + }, + { + "auxiliary_loss_clip": 0.01181414, + "auxiliary_loss_mlp": 0.01022541, + "balance_loss_clip": 1.04300642, + "balance_loss_mlp": 1.01609731, + "epoch": 0.9065111525281068, + "flos": 33547353361920.0, + "grad_norm": 1.919711934683368, + "language_loss": 0.72902876, + "learning_rate": 9.088835328537303e-08, + "loss": 0.75106829, + "num_input_tokens_seen": 163006900, + "step": 7539, + "time_per_iteration": 2.7893927097320557 + }, + { + "auxiliary_loss_clip": 0.01188872, + "auxiliary_loss_mlp": 0.01023366, + "balance_loss_clip": 1.04535568, + "balance_loss_mlp": 1.01592791, + "epoch": 0.9066313954187458, + "flos": 23367863750400.0, + "grad_norm": 2.4872594754409745, + "language_loss": 0.71349013, + "learning_rate": 9.065633947868568e-08, + "loss": 0.73561251, + "num_input_tokens_seen": 163026505, + "step": 7540, + "time_per_iteration": 3.575864315032959 + }, + { + "auxiliary_loss_clip": 0.01194852, + "auxiliary_loss_mlp": 0.01378548, + "balance_loss_clip": 1.04686975, + "balance_loss_mlp": 1.00016141, + "epoch": 0.906751638309385, + "flos": 26249623067520.0, + "grad_norm": 2.2310546757804213, + "language_loss": 0.79994446, + "learning_rate": 9.042461531791379e-08, + "loss": 0.82567841, + "num_input_tokens_seen": 163044925, + "step": 7541, + "time_per_iteration": 2.7033817768096924 + }, + { + "auxiliary_loss_clip": 0.01162165, + "auxiliary_loss_mlp": 0.01022842, + "balance_loss_clip": 1.04439175, + "balance_loss_mlp": 1.01570153, + "epoch": 0.906871881200024, + "flos": 16544477485440.0, + "grad_norm": 1.920925443495151, + "language_loss": 0.78056121, + "learning_rate": 9.019318083820903e-08, + "loss": 0.80241126, + "num_input_tokens_seen": 163063505, + "step": 7542, + "time_per_iteration": 2.5718441009521484 + }, + { + "auxiliary_loss_clip": 0.01175875, + "auxiliary_loss_mlp": 0.01028978, + "balance_loss_clip": 1.04682279, + "balance_loss_mlp": 1.02150321, + "epoch": 0.9069921240906631, + "flos": 24605129675520.0, + "grad_norm": 3.2295697462419004, + "language_loss": 0.85218263, + "learning_rate": 8.996203607468045e-08, + "loss": 0.87423116, + "num_input_tokens_seen": 163082505, + "step": 7543, + "time_per_iteration": 3.5561976432800293 + }, + { + "auxiliary_loss_clip": 0.01169003, + "auxiliary_loss_mlp": 0.0102809, + "balance_loss_clip": 1.04154301, + "balance_loss_mlp": 1.02071989, + "epoch": 0.9071123669813023, + "flos": 25374731500800.0, + "grad_norm": 1.3859456974058848, + "language_loss": 0.75404745, + "learning_rate": 8.973118106239241e-08, + "loss": 0.77601838, + "num_input_tokens_seen": 163105110, + "step": 7544, + "time_per_iteration": 2.645620822906494 + }, + { + "auxiliary_loss_clip": 0.01204467, + "auxiliary_loss_mlp": 0.01025176, + "balance_loss_clip": 1.03578973, + "balance_loss_mlp": 1.01820517, + "epoch": 0.9072326098719413, + "flos": 26725798690560.0, + "grad_norm": 1.80454319589403, + "language_loss": 0.94576317, + "learning_rate": 8.95006158363656e-08, + "loss": 0.96805966, + "num_input_tokens_seen": 163125295, + "step": 7545, + "time_per_iteration": 3.7190914154052734 + }, + { + "auxiliary_loss_clip": 0.01178033, + "auxiliary_loss_mlp": 0.01027308, + "balance_loss_clip": 1.04733562, + "balance_loss_mlp": 1.01929712, + "epoch": 0.9073528527625804, + "flos": 23878800760320.0, + "grad_norm": 1.9132675640162147, + "language_loss": 0.77358377, + "learning_rate": 8.9270340431576e-08, + "loss": 0.79563713, + "num_input_tokens_seen": 163144385, + "step": 7546, + "time_per_iteration": 2.6117899417877197 + }, + { + "auxiliary_loss_clip": 0.01177549, + "auxiliary_loss_mlp": 0.01025587, + "balance_loss_clip": 1.04460263, + "balance_loss_mlp": 1.01860142, + "epoch": 0.9074730956532195, + "flos": 37852144767360.0, + "grad_norm": 1.8752621598557961, + "language_loss": 0.73428798, + "learning_rate": 8.904035488295658e-08, + "loss": 0.7563194, + "num_input_tokens_seen": 163163885, + "step": 7547, + "time_per_iteration": 2.782876968383789 + }, + { + "auxiliary_loss_clip": 0.01073888, + "auxiliary_loss_mlp": 0.01373552, + "balance_loss_clip": 1.00930929, + "balance_loss_mlp": 0.9997322, + "epoch": 0.9075933385438586, + "flos": 65173307385600.0, + "grad_norm": 0.6613274949136936, + "language_loss": 0.53229153, + "learning_rate": 8.881065922539632e-08, + "loss": 0.55676591, + "num_input_tokens_seen": 163224325, + "step": 7548, + "time_per_iteration": 3.1290652751922607 + }, + { + "auxiliary_loss_clip": 0.01187315, + "auxiliary_loss_mlp": 0.01025418, + "balance_loss_clip": 1.04354703, + "balance_loss_mlp": 1.01844716, + "epoch": 0.9077135814344977, + "flos": 19931571290880.0, + "grad_norm": 2.603434562959775, + "language_loss": 0.73568833, + "learning_rate": 8.85812534937389e-08, + "loss": 0.75781566, + "num_input_tokens_seen": 163242425, + "step": 7549, + "time_per_iteration": 2.7400050163269043 + }, + { + "auxiliary_loss_clip": 0.01184476, + "auxiliary_loss_mlp": 0.01025295, + "balance_loss_clip": 1.04814696, + "balance_loss_mlp": 1.01780879, + "epoch": 0.9078338243251368, + "flos": 17529650784000.0, + "grad_norm": 2.5954846067357, + "language_loss": 0.67344677, + "learning_rate": 8.835213772278583e-08, + "loss": 0.69554442, + "num_input_tokens_seen": 163259280, + "step": 7550, + "time_per_iteration": 3.4169626235961914 + }, + { + "auxiliary_loss_clip": 0.01186949, + "auxiliary_loss_mlp": 0.01022981, + "balance_loss_clip": 1.04368496, + "balance_loss_mlp": 1.01608443, + "epoch": 0.9079540672157759, + "flos": 28803410277120.0, + "grad_norm": 1.638688293250881, + "language_loss": 0.79064226, + "learning_rate": 8.812331194729373e-08, + "loss": 0.81274152, + "num_input_tokens_seen": 163278925, + "step": 7551, + "time_per_iteration": 2.6712772846221924 + }, + { + "auxiliary_loss_clip": 0.01172579, + "auxiliary_loss_mlp": 0.01028219, + "balance_loss_clip": 1.05031037, + "balance_loss_mlp": 1.02065504, + "epoch": 0.9080743101064149, + "flos": 23513840622720.0, + "grad_norm": 1.7589348248158867, + "language_loss": 0.72136188, + "learning_rate": 8.789477620197461e-08, + "loss": 0.74336988, + "num_input_tokens_seen": 163298450, + "step": 7552, + "time_per_iteration": 2.547487258911133 + }, + { + "auxiliary_loss_clip": 0.01181963, + "auxiliary_loss_mlp": 0.01027315, + "balance_loss_clip": 1.04347765, + "balance_loss_mlp": 1.01959682, + "epoch": 0.9081945529970541, + "flos": 22778102344320.0, + "grad_norm": 2.0157573403732716, + "language_loss": 0.79120421, + "learning_rate": 8.766653052149831e-08, + "loss": 0.81329691, + "num_input_tokens_seen": 163313635, + "step": 7553, + "time_per_iteration": 2.626213312149048 + }, + { + "auxiliary_loss_clip": 0.01185993, + "auxiliary_loss_mlp": 0.0102394, + "balance_loss_clip": 1.04449165, + "balance_loss_mlp": 1.01651037, + "epoch": 0.9083147958876931, + "flos": 18873714821760.0, + "grad_norm": 2.049634372574394, + "language_loss": 0.74618959, + "learning_rate": 8.743857494048823e-08, + "loss": 0.76828891, + "num_input_tokens_seen": 163330450, + "step": 7554, + "time_per_iteration": 2.6287925243377686 + }, + { + "auxiliary_loss_clip": 0.01195663, + "auxiliary_loss_mlp": 0.01021782, + "balance_loss_clip": 1.04526162, + "balance_loss_mlp": 1.01515126, + "epoch": 0.9084350387783322, + "flos": 18909374048640.0, + "grad_norm": 1.9059670139713698, + "language_loss": 0.62614667, + "learning_rate": 8.721090949352605e-08, + "loss": 0.64832115, + "num_input_tokens_seen": 163346690, + "step": 7555, + "time_per_iteration": 2.692394256591797 + }, + { + "auxiliary_loss_clip": 0.01187407, + "auxiliary_loss_mlp": 0.01024984, + "balance_loss_clip": 1.0485419, + "balance_loss_mlp": 1.0171876, + "epoch": 0.9085552816689714, + "flos": 20595488325120.0, + "grad_norm": 2.1281077006432425, + "language_loss": 0.73273051, + "learning_rate": 8.698353421514793e-08, + "loss": 0.75485444, + "num_input_tokens_seen": 163365065, + "step": 7556, + "time_per_iteration": 2.713810443878174 + }, + { + "auxiliary_loss_clip": 0.01174312, + "auxiliary_loss_mlp": 0.01028947, + "balance_loss_clip": 1.04644871, + "balance_loss_mlp": 1.02196383, + "epoch": 0.9086755245596104, + "flos": 18113163223680.0, + "grad_norm": 2.1220760509398864, + "language_loss": 0.80396748, + "learning_rate": 8.67564491398467e-08, + "loss": 0.82600009, + "num_input_tokens_seen": 163382070, + "step": 7557, + "time_per_iteration": 2.558793783187866 + }, + { + "auxiliary_loss_clip": 0.01179734, + "auxiliary_loss_mlp": 0.01027234, + "balance_loss_clip": 1.04552948, + "balance_loss_mlp": 1.02005124, + "epoch": 0.9087957674502495, + "flos": 19129793857920.0, + "grad_norm": 1.740021430366877, + "language_loss": 0.73906791, + "learning_rate": 8.652965430207104e-08, + "loss": 0.76113755, + "num_input_tokens_seen": 163399975, + "step": 7558, + "time_per_iteration": 2.6516034603118896 + }, + { + "auxiliary_loss_clip": 0.01180017, + "auxiliary_loss_mlp": 0.01029494, + "balance_loss_clip": 1.04579008, + "balance_loss_mlp": 1.02185917, + "epoch": 0.9089160103408886, + "flos": 18109930999680.0, + "grad_norm": 2.0710771370538272, + "language_loss": 0.65487176, + "learning_rate": 8.630314973622521e-08, + "loss": 0.67696685, + "num_input_tokens_seen": 163417520, + "step": 7559, + "time_per_iteration": 2.5550737380981445 + }, + { + "auxiliary_loss_clip": 0.01176189, + "auxiliary_loss_mlp": 0.01024351, + "balance_loss_clip": 1.04893148, + "balance_loss_mlp": 1.0174098, + "epoch": 0.9090362532315277, + "flos": 33364855336320.0, + "grad_norm": 1.8526635128722064, + "language_loss": 0.70926571, + "learning_rate": 8.607693547666995e-08, + "loss": 0.73127103, + "num_input_tokens_seen": 163440060, + "step": 7560, + "time_per_iteration": 2.7018625736236572 + }, + { + "auxiliary_loss_clip": 0.01098781, + "auxiliary_loss_mlp": 0.00999062, + "balance_loss_clip": 1.00916028, + "balance_loss_mlp": 0.99795324, + "epoch": 0.9091564961221668, + "flos": 71480585082240.0, + "grad_norm": 0.8900648790096518, + "language_loss": 0.5793305, + "learning_rate": 8.585101155772201e-08, + "loss": 0.60030895, + "num_input_tokens_seen": 163502180, + "step": 7561, + "time_per_iteration": 3.3317320346832275 + }, + { + "auxiliary_loss_clip": 0.01173038, + "auxiliary_loss_mlp": 0.01026622, + "balance_loss_clip": 1.03956318, + "balance_loss_mlp": 1.01949334, + "epoch": 0.9092767390128058, + "flos": 24712574232960.0, + "grad_norm": 1.7523994628692559, + "language_loss": 0.68874151, + "learning_rate": 8.562537801365377e-08, + "loss": 0.71073812, + "num_input_tokens_seen": 163521915, + "step": 7562, + "time_per_iteration": 2.6456046104431152 + }, + { + "auxiliary_loss_clip": 0.01167747, + "auxiliary_loss_mlp": 0.01025533, + "balance_loss_clip": 1.04688907, + "balance_loss_mlp": 1.01751077, + "epoch": 0.909396981903445, + "flos": 23586487879680.0, + "grad_norm": 1.725881088007337, + "language_loss": 0.70162749, + "learning_rate": 8.540003487869362e-08, + "loss": 0.72356027, + "num_input_tokens_seen": 163543585, + "step": 7563, + "time_per_iteration": 2.6980996131896973 + }, + { + "auxiliary_loss_clip": 0.01179033, + "auxiliary_loss_mlp": 0.01029104, + "balance_loss_clip": 1.04043794, + "balance_loss_mlp": 1.0216949, + "epoch": 0.909517224794084, + "flos": 23404169422080.0, + "grad_norm": 2.3860543179735947, + "language_loss": 0.79800594, + "learning_rate": 8.517498218702557e-08, + "loss": 0.82008731, + "num_input_tokens_seen": 163561515, + "step": 7564, + "time_per_iteration": 2.6648495197296143 + }, + { + "auxiliary_loss_clip": 0.01185446, + "auxiliary_loss_mlp": 0.01021802, + "balance_loss_clip": 1.04043472, + "balance_loss_mlp": 1.01462543, + "epoch": 0.9096374676847231, + "flos": 19208618254080.0, + "grad_norm": 1.6328773110875199, + "language_loss": 0.69703418, + "learning_rate": 8.49502199727905e-08, + "loss": 0.71910667, + "num_input_tokens_seen": 163579540, + "step": 7565, + "time_per_iteration": 2.6490581035614014 + }, + { + "auxiliary_loss_clip": 0.01168512, + "auxiliary_loss_mlp": 0.01030082, + "balance_loss_clip": 1.04109573, + "balance_loss_mlp": 1.02220213, + "epoch": 0.9097577105753623, + "flos": 33292495388160.0, + "grad_norm": 2.444107435388698, + "language_loss": 0.65925193, + "learning_rate": 8.472574827008428e-08, + "loss": 0.68123794, + "num_input_tokens_seen": 163600425, + "step": 7566, + "time_per_iteration": 3.64900279045105 + }, + { + "auxiliary_loss_clip": 0.01178717, + "auxiliary_loss_mlp": 0.01024351, + "balance_loss_clip": 1.04584825, + "balance_loss_mlp": 1.01719534, + "epoch": 0.9098779534660013, + "flos": 21906443001600.0, + "grad_norm": 1.6318193180462692, + "language_loss": 0.84067166, + "learning_rate": 8.450156711295942e-08, + "loss": 0.86270225, + "num_input_tokens_seen": 163620595, + "step": 7567, + "time_per_iteration": 2.6893296241760254 + }, + { + "auxiliary_loss_clip": 0.0118378, + "auxiliary_loss_mlp": 0.01028563, + "balance_loss_clip": 1.04646039, + "balance_loss_mlp": 1.02158642, + "epoch": 0.9099981963566404, + "flos": 25730354102400.0, + "grad_norm": 1.930160805931789, + "language_loss": 0.8681227, + "learning_rate": 8.427767653542383e-08, + "loss": 0.89024615, + "num_input_tokens_seen": 163635765, + "step": 7568, + "time_per_iteration": 2.6325042247772217 + }, + { + "auxiliary_loss_clip": 0.01196849, + "auxiliary_loss_mlp": 0.01025471, + "balance_loss_clip": 1.04006469, + "balance_loss_mlp": 1.01860476, + "epoch": 0.9101184392472795, + "flos": 21069437304960.0, + "grad_norm": 1.740888936769195, + "language_loss": 0.70053625, + "learning_rate": 8.405407657144125e-08, + "loss": 0.72275943, + "num_input_tokens_seen": 163654925, + "step": 7569, + "time_per_iteration": 2.7721943855285645 + }, + { + "auxiliary_loss_clip": 0.01177456, + "auxiliary_loss_mlp": 0.01027273, + "balance_loss_clip": 1.04310799, + "balance_loss_mlp": 1.01998341, + "epoch": 0.9102386821379186, + "flos": 24752614919040.0, + "grad_norm": 1.8050131802233684, + "language_loss": 0.72513235, + "learning_rate": 8.383076725493232e-08, + "loss": 0.74717963, + "num_input_tokens_seen": 163672245, + "step": 7570, + "time_per_iteration": 3.566356658935547 + }, + { + "auxiliary_loss_clip": 0.01175719, + "auxiliary_loss_mlp": 0.01022317, + "balance_loss_clip": 1.04468179, + "balance_loss_mlp": 1.01528692, + "epoch": 0.9103589250285576, + "flos": 22562818179840.0, + "grad_norm": 2.0432104491595884, + "language_loss": 0.67941582, + "learning_rate": 8.360774861977216e-08, + "loss": 0.70139623, + "num_input_tokens_seen": 163691365, + "step": 7571, + "time_per_iteration": 3.4600698947906494 + }, + { + "auxiliary_loss_clip": 0.01180495, + "auxiliary_loss_mlp": 0.01023123, + "balance_loss_clip": 1.03953815, + "balance_loss_mlp": 1.01568401, + "epoch": 0.9104791679191968, + "flos": 25373474524800.0, + "grad_norm": 2.003997180253617, + "language_loss": 0.7440486, + "learning_rate": 8.338502069979281e-08, + "loss": 0.76608479, + "num_input_tokens_seen": 163711675, + "step": 7572, + "time_per_iteration": 2.6929636001586914 + }, + { + "auxiliary_loss_clip": 0.01175263, + "auxiliary_loss_mlp": 0.01027973, + "balance_loss_clip": 1.04235303, + "balance_loss_mlp": 1.0205698, + "epoch": 0.9105994108098359, + "flos": 14426681558400.0, + "grad_norm": 3.156125523669606, + "language_loss": 0.80854678, + "learning_rate": 8.316258352878214e-08, + "loss": 0.83057916, + "num_input_tokens_seen": 163728095, + "step": 7573, + "time_per_iteration": 2.544158935546875 + }, + { + "auxiliary_loss_clip": 0.01180019, + "auxiliary_loss_mlp": 0.01021275, + "balance_loss_clip": 1.04569817, + "balance_loss_mlp": 1.01424754, + "epoch": 0.9107196537004749, + "flos": 26718292748160.0, + "grad_norm": 1.7809485400531129, + "language_loss": 0.70886236, + "learning_rate": 8.294043714048338e-08, + "loss": 0.73087525, + "num_input_tokens_seen": 163747175, + "step": 7574, + "time_per_iteration": 2.720402717590332 + }, + { + "auxiliary_loss_clip": 0.01086, + "auxiliary_loss_mlp": 0.01002293, + "balance_loss_clip": 1.00908232, + "balance_loss_mlp": 1.00113702, + "epoch": 0.9108398965911141, + "flos": 66532634703360.0, + "grad_norm": 0.7494417691554045, + "language_loss": 0.60437191, + "learning_rate": 8.271858156859624e-08, + "loss": 0.62525487, + "num_input_tokens_seen": 163812545, + "step": 7575, + "time_per_iteration": 3.31634259223938 + }, + { + "auxiliary_loss_clip": 0.01164106, + "auxiliary_loss_mlp": 0.01026016, + "balance_loss_clip": 1.04549932, + "balance_loss_mlp": 1.01906347, + "epoch": 0.9109601394817531, + "flos": 25411073086080.0, + "grad_norm": 1.6249152493924315, + "language_loss": 0.73925269, + "learning_rate": 8.249701684677557e-08, + "loss": 0.76115388, + "num_input_tokens_seen": 163833870, + "step": 7576, + "time_per_iteration": 2.5734169483184814 + }, + { + "auxiliary_loss_clip": 0.01178328, + "auxiliary_loss_mlp": 0.01023627, + "balance_loss_clip": 1.04990363, + "balance_loss_mlp": 1.01608384, + "epoch": 0.9110803823723922, + "flos": 22747794243840.0, + "grad_norm": 1.909107378960764, + "language_loss": 0.81030345, + "learning_rate": 8.227574300863294e-08, + "loss": 0.83232301, + "num_input_tokens_seen": 163854040, + "step": 7577, + "time_per_iteration": 3.6243858337402344 + }, + { + "auxiliary_loss_clip": 0.01189961, + "auxiliary_loss_mlp": 0.01021875, + "balance_loss_clip": 1.04635382, + "balance_loss_mlp": 1.01442194, + "epoch": 0.9112006252630314, + "flos": 48469924131840.0, + "grad_norm": 1.6818632121610129, + "language_loss": 0.69769609, + "learning_rate": 8.205476008773548e-08, + "loss": 0.71981442, + "num_input_tokens_seen": 163878040, + "step": 7578, + "time_per_iteration": 2.84867525100708 + }, + { + "auxiliary_loss_clip": 0.01188833, + "auxiliary_loss_mlp": 0.01027246, + "balance_loss_clip": 1.04450643, + "balance_loss_mlp": 1.01958084, + "epoch": 0.9113208681536704, + "flos": 30009649829760.0, + "grad_norm": 2.009826471171645, + "language_loss": 0.82904303, + "learning_rate": 8.183406811760596e-08, + "loss": 0.8512038, + "num_input_tokens_seen": 163897770, + "step": 7579, + "time_per_iteration": 2.7619709968566895 + }, + { + "auxiliary_loss_clip": 0.01179665, + "auxiliary_loss_mlp": 0.01027123, + "balance_loss_clip": 1.04188287, + "balance_loss_mlp": 1.02019095, + "epoch": 0.9114411110443095, + "flos": 25594971742080.0, + "grad_norm": 1.6245575400791936, + "language_loss": 0.73915178, + "learning_rate": 8.161366713172313e-08, + "loss": 0.76121968, + "num_input_tokens_seen": 163920160, + "step": 7580, + "time_per_iteration": 2.6986281871795654 + }, + { + "auxiliary_loss_clip": 0.01196404, + "auxiliary_loss_mlp": 0.01026735, + "balance_loss_clip": 1.04135728, + "balance_loss_mlp": 1.01907611, + "epoch": 0.9115613539349486, + "flos": 18399729928320.0, + "grad_norm": 2.349394812055369, + "language_loss": 0.8402406, + "learning_rate": 8.139355716352137e-08, + "loss": 0.86247206, + "num_input_tokens_seen": 163935000, + "step": 7581, + "time_per_iteration": 2.6584115028381348 + }, + { + "auxiliary_loss_clip": 0.01186489, + "auxiliary_loss_mlp": 0.0102637, + "balance_loss_clip": 1.04299045, + "balance_loss_mlp": 1.01880288, + "epoch": 0.9116815968255877, + "flos": 21726171619200.0, + "grad_norm": 1.6066692538603609, + "language_loss": 0.7003817, + "learning_rate": 8.117373824639196e-08, + "loss": 0.72251034, + "num_input_tokens_seen": 163955265, + "step": 7582, + "time_per_iteration": 2.6353492736816406 + }, + { + "auxiliary_loss_clip": 0.01059819, + "auxiliary_loss_mlp": 0.0100217, + "balance_loss_clip": 1.00880623, + "balance_loss_mlp": 1.00108492, + "epoch": 0.9118018397162267, + "flos": 65363526835200.0, + "grad_norm": 0.7213409140852138, + "language_loss": 0.59224498, + "learning_rate": 8.095421041368067e-08, + "loss": 0.61286485, + "num_input_tokens_seen": 164014680, + "step": 7583, + "time_per_iteration": 3.1141934394836426 + }, + { + "auxiliary_loss_clip": 0.01182479, + "auxiliary_loss_mlp": 0.01378149, + "balance_loss_clip": 1.0452441, + "balance_loss_mlp": 1.00013709, + "epoch": 0.9119220826068659, + "flos": 20922885815040.0, + "grad_norm": 2.122074076544407, + "language_loss": 0.70366037, + "learning_rate": 8.073497369868999e-08, + "loss": 0.72926664, + "num_input_tokens_seen": 164033140, + "step": 7584, + "time_per_iteration": 2.662602186203003 + }, + { + "auxiliary_loss_clip": 0.01191234, + "auxiliary_loss_mlp": 0.01021839, + "balance_loss_clip": 1.0442965, + "balance_loss_mlp": 1.01414466, + "epoch": 0.912042325497505, + "flos": 28366449327360.0, + "grad_norm": 1.5987158698763753, + "language_loss": 0.75559944, + "learning_rate": 8.051602813467772e-08, + "loss": 0.77773017, + "num_input_tokens_seen": 164054995, + "step": 7585, + "time_per_iteration": 2.6948559284210205 + }, + { + "auxiliary_loss_clip": 0.01178721, + "auxiliary_loss_mlp": 0.0102474, + "balance_loss_clip": 1.04663777, + "balance_loss_mlp": 1.01766849, + "epoch": 0.912162568388144, + "flos": 17566782468480.0, + "grad_norm": 1.785342714970948, + "language_loss": 0.71195912, + "learning_rate": 8.029737375485756e-08, + "loss": 0.73399377, + "num_input_tokens_seen": 164074225, + "step": 7586, + "time_per_iteration": 2.6458895206451416 + }, + { + "auxiliary_loss_clip": 0.01168646, + "auxiliary_loss_mlp": 0.01021602, + "balance_loss_clip": 1.04770398, + "balance_loss_mlp": 1.01400256, + "epoch": 0.9122828112787832, + "flos": 19827897661440.0, + "grad_norm": 1.7599387487745644, + "language_loss": 0.721892, + "learning_rate": 8.007901059239986e-08, + "loss": 0.7437945, + "num_input_tokens_seen": 164093505, + "step": 7587, + "time_per_iteration": 2.5693204402923584 + }, + { + "auxiliary_loss_clip": 0.01182573, + "auxiliary_loss_mlp": 0.01024428, + "balance_loss_clip": 1.04161036, + "balance_loss_mlp": 1.01723385, + "epoch": 0.9124030541694222, + "flos": 20813789232000.0, + "grad_norm": 1.6555199959687823, + "language_loss": 0.79626441, + "learning_rate": 7.986093868042964e-08, + "loss": 0.8183344, + "num_input_tokens_seen": 164113750, + "step": 7588, + "time_per_iteration": 2.69376540184021 + }, + { + "auxiliary_loss_clip": 0.01174898, + "auxiliary_loss_mlp": 0.01026505, + "balance_loss_clip": 1.04578733, + "balance_loss_mlp": 1.02012992, + "epoch": 0.9125232970600613, + "flos": 25192305302400.0, + "grad_norm": 1.770396095399832, + "language_loss": 0.67838025, + "learning_rate": 7.964315805202826e-08, + "loss": 0.70039433, + "num_input_tokens_seen": 164134330, + "step": 7589, + "time_per_iteration": 2.6720666885375977 + }, + { + "auxiliary_loss_clip": 0.01185768, + "auxiliary_loss_mlp": 0.01026185, + "balance_loss_clip": 1.04446316, + "balance_loss_mlp": 1.01885653, + "epoch": 0.9126435399507005, + "flos": 19719591177600.0, + "grad_norm": 1.7821864886057974, + "language_loss": 0.73009229, + "learning_rate": 7.942566874023304e-08, + "loss": 0.75221181, + "num_input_tokens_seen": 164153515, + "step": 7590, + "time_per_iteration": 2.6359989643096924 + }, + { + "auxiliary_loss_clip": 0.0117932, + "auxiliary_loss_mlp": 0.01019809, + "balance_loss_clip": 1.04191697, + "balance_loss_mlp": 1.01268911, + "epoch": 0.9127637828413395, + "flos": 19573614305280.0, + "grad_norm": 2.3549860471458994, + "language_loss": 0.70088214, + "learning_rate": 7.920847077803649e-08, + "loss": 0.72287345, + "num_input_tokens_seen": 164171305, + "step": 7591, + "time_per_iteration": 2.5678529739379883 + }, + { + "auxiliary_loss_clip": 0.01183882, + "auxiliary_loss_mlp": 0.01030745, + "balance_loss_clip": 1.03519583, + "balance_loss_mlp": 1.0237956, + "epoch": 0.9128840257319786, + "flos": 20230635928320.0, + "grad_norm": 1.826435280088008, + "language_loss": 0.81989932, + "learning_rate": 7.899156419838826e-08, + "loss": 0.84204555, + "num_input_tokens_seen": 164190275, + "step": 7592, + "time_per_iteration": 3.657386541366577 + }, + { + "auxiliary_loss_clip": 0.0119044, + "auxiliary_loss_mlp": 0.0102791, + "balance_loss_clip": 1.04193187, + "balance_loss_mlp": 1.02096558, + "epoch": 0.9130042686226177, + "flos": 24858658846080.0, + "grad_norm": 1.7933298894395047, + "language_loss": 0.65910459, + "learning_rate": 7.87749490341918e-08, + "loss": 0.68128812, + "num_input_tokens_seen": 164210550, + "step": 7593, + "time_per_iteration": 2.7077548503875732 + }, + { + "auxiliary_loss_clip": 0.01171664, + "auxiliary_loss_mlp": 0.01026717, + "balance_loss_clip": 1.04889619, + "balance_loss_mlp": 1.01924539, + "epoch": 0.9131245115132568, + "flos": 23581747284480.0, + "grad_norm": 1.9528776397657162, + "language_loss": 0.83380967, + "learning_rate": 7.855862531830836e-08, + "loss": 0.85579348, + "num_input_tokens_seen": 164226660, + "step": 7594, + "time_per_iteration": 2.581590414047241 + }, + { + "auxiliary_loss_clip": 0.01170819, + "auxiliary_loss_mlp": 0.01025749, + "balance_loss_clip": 1.04303348, + "balance_loss_mlp": 1.01834583, + "epoch": 0.9132447544038959, + "flos": 19931607204480.0, + "grad_norm": 1.7574401606583503, + "language_loss": 0.72797918, + "learning_rate": 7.834259308355373e-08, + "loss": 0.74994481, + "num_input_tokens_seen": 164245425, + "step": 7595, + "time_per_iteration": 3.510378122329712 + }, + { + "auxiliary_loss_clip": 0.01191416, + "auxiliary_loss_mlp": 0.01026776, + "balance_loss_clip": 1.03847182, + "balance_loss_mlp": 1.01945686, + "epoch": 0.9133649972945349, + "flos": 21981747864960.0, + "grad_norm": 7.387302881285032, + "language_loss": 0.74860656, + "learning_rate": 7.812685236269989e-08, + "loss": 0.77078849, + "num_input_tokens_seen": 164264085, + "step": 7596, + "time_per_iteration": 2.736605644226074 + }, + { + "auxiliary_loss_clip": 0.01093978, + "auxiliary_loss_mlp": 0.01001005, + "balance_loss_clip": 1.00987113, + "balance_loss_mlp": 0.99988407, + "epoch": 0.9134852401851741, + "flos": 71240523511680.0, + "grad_norm": 0.7924383527087546, + "language_loss": 0.58673722, + "learning_rate": 7.791140318847445e-08, + "loss": 0.60768706, + "num_input_tokens_seen": 164322220, + "step": 7597, + "time_per_iteration": 4.13280725479126 + }, + { + "auxiliary_loss_clip": 0.01181427, + "auxiliary_loss_mlp": 0.01028201, + "balance_loss_clip": 1.04605448, + "balance_loss_mlp": 1.02160549, + "epoch": 0.9136054830758131, + "flos": 23626923615360.0, + "grad_norm": 1.6552329886663624, + "language_loss": 0.80395776, + "learning_rate": 7.769624559356081e-08, + "loss": 0.8260541, + "num_input_tokens_seen": 164345615, + "step": 7598, + "time_per_iteration": 2.7800025939941406 + }, + { + "auxiliary_loss_clip": 0.01174238, + "auxiliary_loss_mlp": 0.01025172, + "balance_loss_clip": 1.04453635, + "balance_loss_mlp": 1.01745629, + "epoch": 0.9137257259664522, + "flos": 23438858981760.0, + "grad_norm": 3.1135486074152614, + "language_loss": 0.74956989, + "learning_rate": 7.748137961059842e-08, + "loss": 0.77156401, + "num_input_tokens_seen": 164359595, + "step": 7599, + "time_per_iteration": 2.6657750606536865 + }, + { + "auxiliary_loss_clip": 0.01166208, + "auxiliary_loss_mlp": 0.01025068, + "balance_loss_clip": 1.0481559, + "balance_loss_mlp": 1.01828504, + "epoch": 0.9138459688570914, + "flos": 19127854523520.0, + "grad_norm": 2.136078321431987, + "language_loss": 0.65264726, + "learning_rate": 7.726680527218211e-08, + "loss": 0.67456007, + "num_input_tokens_seen": 164376635, + "step": 7600, + "time_per_iteration": 2.646888256072998 + }, + { + "auxiliary_loss_clip": 0.01167908, + "auxiliary_loss_mlp": 0.01025696, + "balance_loss_clip": 1.04521585, + "balance_loss_mlp": 1.01790822, + "epoch": 0.9139662117477304, + "flos": 46281240714240.0, + "grad_norm": 1.8386386314895873, + "language_loss": 0.75580394, + "learning_rate": 7.70525226108627e-08, + "loss": 0.77773994, + "num_input_tokens_seen": 164400305, + "step": 7601, + "time_per_iteration": 2.8000574111938477 + }, + { + "auxiliary_loss_clip": 0.0118011, + "auxiliary_loss_mlp": 0.01026202, + "balance_loss_clip": 1.04908061, + "balance_loss_mlp": 1.01904988, + "epoch": 0.9140864546383695, + "flos": 22273198819200.0, + "grad_norm": 2.0320750971959893, + "language_loss": 0.79713035, + "learning_rate": 7.683853165914666e-08, + "loss": 0.81919348, + "num_input_tokens_seen": 164418075, + "step": 7602, + "time_per_iteration": 3.563960552215576 + }, + { + "auxiliary_loss_clip": 0.01202612, + "auxiliary_loss_mlp": 0.01027625, + "balance_loss_clip": 1.04162264, + "balance_loss_mlp": 1.02015638, + "epoch": 0.9142066975290086, + "flos": 17530009920000.0, + "grad_norm": 2.408229685194719, + "language_loss": 0.77170682, + "learning_rate": 7.662483244949602e-08, + "loss": 0.79400915, + "num_input_tokens_seen": 164435335, + "step": 7603, + "time_per_iteration": 2.765648603439331 + }, + { + "auxiliary_loss_clip": 0.01186201, + "auxiliary_loss_mlp": 0.01023991, + "balance_loss_clip": 1.04184055, + "balance_loss_mlp": 1.01715159, + "epoch": 0.9143269404196477, + "flos": 17712148809600.0, + "grad_norm": 2.3092717370937543, + "language_loss": 0.80412686, + "learning_rate": 7.641142501432951e-08, + "loss": 0.82622874, + "num_input_tokens_seen": 164451530, + "step": 7604, + "time_per_iteration": 2.6561129093170166 + }, + { + "auxiliary_loss_clip": 0.01179932, + "auxiliary_loss_mlp": 0.01031226, + "balance_loss_clip": 1.04213428, + "balance_loss_mlp": 1.02386427, + "epoch": 0.9144471833102867, + "flos": 33323414019840.0, + "grad_norm": 2.3511741028615405, + "language_loss": 0.73713005, + "learning_rate": 7.619830938602013e-08, + "loss": 0.75924164, + "num_input_tokens_seen": 164472755, + "step": 7605, + "time_per_iteration": 2.7594525814056396 + }, + { + "auxiliary_loss_clip": 0.01168375, + "auxiliary_loss_mlp": 0.01024272, + "balance_loss_clip": 1.04473138, + "balance_loss_mlp": 1.01694012, + "epoch": 0.9145674262009259, + "flos": 21068970428160.0, + "grad_norm": 6.653798930516699, + "language_loss": 0.82328582, + "learning_rate": 7.598548559689777e-08, + "loss": 0.84521234, + "num_input_tokens_seen": 164491155, + "step": 7606, + "time_per_iteration": 2.595553398132324 + }, + { + "auxiliary_loss_clip": 0.01184303, + "auxiliary_loss_mlp": 0.01024005, + "balance_loss_clip": 1.04105103, + "balance_loss_mlp": 1.01689434, + "epoch": 0.914687669091565, + "flos": 16800269212800.0, + "grad_norm": 2.2894450147109167, + "language_loss": 0.81213999, + "learning_rate": 7.577295367924751e-08, + "loss": 0.83422309, + "num_input_tokens_seen": 164507555, + "step": 7607, + "time_per_iteration": 2.6750619411468506 + }, + { + "auxiliary_loss_clip": 0.01191962, + "auxiliary_loss_mlp": 0.01024347, + "balance_loss_clip": 1.04700494, + "balance_loss_mlp": 1.01702714, + "epoch": 0.914807911982204, + "flos": 25773627012480.0, + "grad_norm": 2.311815280086121, + "language_loss": 0.82082736, + "learning_rate": 7.556071366531002e-08, + "loss": 0.84299052, + "num_input_tokens_seen": 164528525, + "step": 7608, + "time_per_iteration": 2.6764302253723145 + }, + { + "auxiliary_loss_clip": 0.01176023, + "auxiliary_loss_mlp": 0.0103086, + "balance_loss_clip": 1.04589415, + "balance_loss_mlp": 1.02342176, + "epoch": 0.9149281548728432, + "flos": 19208043636480.0, + "grad_norm": 4.027985119354868, + "language_loss": 0.79069477, + "learning_rate": 7.53487655872822e-08, + "loss": 0.81276357, + "num_input_tokens_seen": 164547695, + "step": 7609, + "time_per_iteration": 2.6549649238586426 + }, + { + "auxiliary_loss_clip": 0.01203907, + "auxiliary_loss_mlp": 0.01023471, + "balance_loss_clip": 1.03943932, + "balance_loss_mlp": 1.01653576, + "epoch": 0.9150483977634822, + "flos": 26870554500480.0, + "grad_norm": 2.0223752814736167, + "language_loss": 0.73725462, + "learning_rate": 7.513710947731656e-08, + "loss": 0.7595284, + "num_input_tokens_seen": 164568905, + "step": 7610, + "time_per_iteration": 2.7899904251098633 + }, + { + "auxiliary_loss_clip": 0.01174327, + "auxiliary_loss_mlp": 0.01026423, + "balance_loss_clip": 1.04283643, + "balance_loss_mlp": 1.01961923, + "epoch": 0.9151686406541213, + "flos": 21908956953600.0, + "grad_norm": 2.200684749209769, + "language_loss": 0.85162568, + "learning_rate": 7.492574536752095e-08, + "loss": 0.87363321, + "num_input_tokens_seen": 164588895, + "step": 7611, + "time_per_iteration": 2.6686530113220215 + }, + { + "auxiliary_loss_clip": 0.01174399, + "auxiliary_loss_mlp": 0.01021699, + "balance_loss_clip": 1.04648805, + "balance_loss_mlp": 1.01509237, + "epoch": 0.9152888835447605, + "flos": 27308556944640.0, + "grad_norm": 1.9237483262776582, + "language_loss": 0.78099233, + "learning_rate": 7.471467328995907e-08, + "loss": 0.80295324, + "num_input_tokens_seen": 164607705, + "step": 7612, + "time_per_iteration": 2.723728656768799 + }, + { + "auxiliary_loss_clip": 0.01235081, + "auxiliary_loss_mlp": 0.01023464, + "balance_loss_clip": 1.03702974, + "balance_loss_mlp": 1.01598358, + "epoch": 0.9154091264353995, + "flos": 13370728510080.0, + "grad_norm": 2.4383751461649013, + "language_loss": 0.60786998, + "learning_rate": 7.450389327665018e-08, + "loss": 0.63045543, + "num_input_tokens_seen": 164625540, + "step": 7613, + "time_per_iteration": 3.1337428092956543 + }, + { + "auxiliary_loss_clip": 0.01196236, + "auxiliary_loss_mlp": 0.01023284, + "balance_loss_clip": 1.04584396, + "balance_loss_mlp": 1.01579189, + "epoch": 0.9155293693260386, + "flos": 20193037367040.0, + "grad_norm": 2.403811149476224, + "language_loss": 0.6795519, + "learning_rate": 7.429340535957029e-08, + "loss": 0.70174706, + "num_input_tokens_seen": 164640735, + "step": 7614, + "time_per_iteration": 3.045193910598755 + }, + { + "auxiliary_loss_clip": 0.01187739, + "auxiliary_loss_mlp": 0.01022599, + "balance_loss_clip": 1.04512382, + "balance_loss_mlp": 1.01538706, + "epoch": 0.9156496122166777, + "flos": 19354990176000.0, + "grad_norm": 2.6929445124277485, + "language_loss": 0.70418817, + "learning_rate": 7.40832095706494e-08, + "loss": 0.72629154, + "num_input_tokens_seen": 164657430, + "step": 7615, + "time_per_iteration": 2.617128849029541 + }, + { + "auxiliary_loss_clip": 0.01201557, + "auxiliary_loss_mlp": 0.01020443, + "balance_loss_clip": 1.04510474, + "balance_loss_mlp": 1.01349878, + "epoch": 0.9157698551073168, + "flos": 21107287261440.0, + "grad_norm": 1.8115430165904447, + "language_loss": 0.80103171, + "learning_rate": 7.387330594177443e-08, + "loss": 0.82325172, + "num_input_tokens_seen": 164679505, + "step": 7616, + "time_per_iteration": 2.7083215713500977 + }, + { + "auxiliary_loss_clip": 0.01189541, + "auxiliary_loss_mlp": 0.01025099, + "balance_loss_clip": 1.04437101, + "balance_loss_mlp": 1.01769614, + "epoch": 0.9158900979979558, + "flos": 25193167228800.0, + "grad_norm": 2.2818230829149533, + "language_loss": 0.78858614, + "learning_rate": 7.366369450478749e-08, + "loss": 0.81073248, + "num_input_tokens_seen": 164700615, + "step": 7617, + "time_per_iteration": 3.626487970352173 + }, + { + "auxiliary_loss_clip": 0.01187252, + "auxiliary_loss_mlp": 0.01023101, + "balance_loss_clip": 1.0422318, + "balance_loss_mlp": 1.01618052, + "epoch": 0.916010340888595, + "flos": 30146648302080.0, + "grad_norm": 1.7036797478402306, + "language_loss": 0.66451824, + "learning_rate": 7.345437529148646e-08, + "loss": 0.68662179, + "num_input_tokens_seen": 164719625, + "step": 7618, + "time_per_iteration": 2.731440544128418 + }, + { + "auxiliary_loss_clip": 0.0119264, + "auxiliary_loss_mlp": 0.01025481, + "balance_loss_clip": 1.04261148, + "balance_loss_mlp": 1.01864779, + "epoch": 0.9161305837792341, + "flos": 17091827907840.0, + "grad_norm": 1.9159254236010428, + "language_loss": 0.72796553, + "learning_rate": 7.324534833362483e-08, + "loss": 0.75014675, + "num_input_tokens_seen": 164737200, + "step": 7619, + "time_per_iteration": 2.6449859142303467 + }, + { + "auxiliary_loss_clip": 0.01185204, + "auxiliary_loss_mlp": 0.01028573, + "balance_loss_clip": 1.04614592, + "balance_loss_mlp": 1.02159381, + "epoch": 0.9162508266698731, + "flos": 22893699288960.0, + "grad_norm": 1.8004178594729294, + "language_loss": 0.6829887, + "learning_rate": 7.303661366291192e-08, + "loss": 0.7051264, + "num_input_tokens_seen": 164757870, + "step": 7620, + "time_per_iteration": 2.660576820373535 + }, + { + "auxiliary_loss_clip": 0.01197632, + "auxiliary_loss_mlp": 0.01025107, + "balance_loss_clip": 1.0399878, + "balance_loss_mlp": 1.01822519, + "epoch": 0.9163710695605123, + "flos": 19974808287360.0, + "grad_norm": 2.0604482402028954, + "language_loss": 0.81807172, + "learning_rate": 7.28281713110126e-08, + "loss": 0.84029913, + "num_input_tokens_seen": 164775945, + "step": 7621, + "time_per_iteration": 3.6666712760925293 + }, + { + "auxiliary_loss_clip": 0.0117907, + "auxiliary_loss_mlp": 0.01021236, + "balance_loss_clip": 1.04403996, + "balance_loss_mlp": 1.01430678, + "epoch": 0.9164913124511513, + "flos": 22783812606720.0, + "grad_norm": 1.8404589094003678, + "language_loss": 0.77135885, + "learning_rate": 7.262002130954759e-08, + "loss": 0.7933619, + "num_input_tokens_seen": 164794400, + "step": 7622, + "time_per_iteration": 3.613988161087036 + }, + { + "auxiliary_loss_clip": 0.01203451, + "auxiliary_loss_mlp": 0.01027855, + "balance_loss_clip": 1.03997397, + "balance_loss_mlp": 1.02053285, + "epoch": 0.9166115553417904, + "flos": 24900854348160.0, + "grad_norm": 1.8433396288163268, + "language_loss": 0.78748578, + "learning_rate": 7.241216369009296e-08, + "loss": 0.80979884, + "num_input_tokens_seen": 164814585, + "step": 7623, + "time_per_iteration": 2.7563016414642334 + }, + { + "auxiliary_loss_clip": 0.01168424, + "auxiliary_loss_mlp": 0.01025678, + "balance_loss_clip": 1.04670262, + "balance_loss_mlp": 1.01875198, + "epoch": 0.9167317982324296, + "flos": 25702919089920.0, + "grad_norm": 1.9894835084000808, + "language_loss": 0.66781497, + "learning_rate": 7.220459848418037e-08, + "loss": 0.68975604, + "num_input_tokens_seen": 164834660, + "step": 7624, + "time_per_iteration": 2.658539295196533 + }, + { + "auxiliary_loss_clip": 0.01166517, + "auxiliary_loss_mlp": 0.01026248, + "balance_loss_clip": 1.0485971, + "balance_loss_mlp": 1.01957512, + "epoch": 0.9168520411230686, + "flos": 15632813370240.0, + "grad_norm": 1.8442056059534566, + "language_loss": 0.79646373, + "learning_rate": 7.199732572329708e-08, + "loss": 0.81839132, + "num_input_tokens_seen": 164852560, + "step": 7625, + "time_per_iteration": 2.5389339923858643 + }, + { + "auxiliary_loss_clip": 0.01195951, + "auxiliary_loss_mlp": 0.01026662, + "balance_loss_clip": 1.04255176, + "balance_loss_mlp": 1.01965833, + "epoch": 0.9169722840137077, + "flos": 30258151096320.0, + "grad_norm": 2.5104371681564954, + "language_loss": 0.75936788, + "learning_rate": 7.179034543888684e-08, + "loss": 0.78159398, + "num_input_tokens_seen": 164872065, + "step": 7626, + "time_per_iteration": 2.7492763996124268 + }, + { + "auxiliary_loss_clip": 0.01176682, + "auxiliary_loss_mlp": 0.01021592, + "balance_loss_clip": 1.04369736, + "balance_loss_mlp": 1.01476157, + "epoch": 0.9170925269043467, + "flos": 22491643380480.0, + "grad_norm": 7.776888002302232, + "language_loss": 0.77499193, + "learning_rate": 7.158365766234808e-08, + "loss": 0.79697466, + "num_input_tokens_seen": 164890915, + "step": 7627, + "time_per_iteration": 2.638888120651245 + }, + { + "auxiliary_loss_clip": 0.01180247, + "auxiliary_loss_mlp": 0.01025323, + "balance_loss_clip": 1.03804541, + "balance_loss_mlp": 1.01762772, + "epoch": 0.9172127697949859, + "flos": 22893914770560.0, + "grad_norm": 1.7829833128632673, + "language_loss": 0.72326702, + "learning_rate": 7.137726242503527e-08, + "loss": 0.7453227, + "num_input_tokens_seen": 164909835, + "step": 7628, + "time_per_iteration": 3.585360288619995 + }, + { + "auxiliary_loss_clip": 0.01176318, + "auxiliary_loss_mlp": 0.01378678, + "balance_loss_clip": 1.04697216, + "balance_loss_mlp": 1.00016642, + "epoch": 0.917333012685625, + "flos": 17451867882240.0, + "grad_norm": 2.1843010119215793, + "language_loss": 0.77780735, + "learning_rate": 7.11711597582585e-08, + "loss": 0.80335736, + "num_input_tokens_seen": 164927195, + "step": 7629, + "time_per_iteration": 2.5966873168945312 + }, + { + "auxiliary_loss_clip": 0.01192471, + "auxiliary_loss_mlp": 0.01026777, + "balance_loss_clip": 1.0404793, + "balance_loss_mlp": 1.01994371, + "epoch": 0.917453255576264, + "flos": 14318949692160.0, + "grad_norm": 1.7424240705100842, + "language_loss": 0.80047178, + "learning_rate": 7.096534969328271e-08, + "loss": 0.82266426, + "num_input_tokens_seen": 164944640, + "step": 7630, + "time_per_iteration": 2.693420648574829 + }, + { + "auxiliary_loss_clip": 0.01185773, + "auxiliary_loss_mlp": 0.01024904, + "balance_loss_clip": 1.04142046, + "balance_loss_mlp": 1.01781678, + "epoch": 0.9175734984669032, + "flos": 20741177888640.0, + "grad_norm": 2.307011642584449, + "language_loss": 0.84167469, + "learning_rate": 7.075983226132987e-08, + "loss": 0.86378145, + "num_input_tokens_seen": 164963570, + "step": 7631, + "time_per_iteration": 2.6943302154541016 + }, + { + "auxiliary_loss_clip": 0.01189651, + "auxiliary_loss_mlp": 0.01379135, + "balance_loss_clip": 1.0440104, + "balance_loss_mlp": 1.00013113, + "epoch": 0.9176937413575422, + "flos": 14830497233280.0, + "grad_norm": 2.4147015268129315, + "language_loss": 0.790892, + "learning_rate": 7.055460749357656e-08, + "loss": 0.81657988, + "num_input_tokens_seen": 164979850, + "step": 7632, + "time_per_iteration": 2.633988618850708 + }, + { + "auxiliary_loss_clip": 0.01183499, + "auxiliary_loss_mlp": 0.01025326, + "balance_loss_clip": 1.04405439, + "balance_loss_mlp": 1.01812291, + "epoch": 0.9178139842481813, + "flos": 18474603828480.0, + "grad_norm": 1.633577770332044, + "language_loss": 0.70505285, + "learning_rate": 7.034967542115521e-08, + "loss": 0.72714108, + "num_input_tokens_seen": 164998115, + "step": 7633, + "time_per_iteration": 2.71527099609375 + }, + { + "auxiliary_loss_clip": 0.01163536, + "auxiliary_loss_mlp": 0.01378667, + "balance_loss_clip": 1.04293084, + "balance_loss_mlp": 1.00012016, + "epoch": 0.9179342271388204, + "flos": 20047455544320.0, + "grad_norm": 2.0472566824023537, + "language_loss": 0.75517672, + "learning_rate": 7.014503607515388e-08, + "loss": 0.78059876, + "num_input_tokens_seen": 165017420, + "step": 7634, + "time_per_iteration": 2.627797842025757 + }, + { + "auxiliary_loss_clip": 0.0118514, + "auxiliary_loss_mlp": 0.01026736, + "balance_loss_clip": 1.04704869, + "balance_loss_mlp": 1.01963139, + "epoch": 0.9180544700294595, + "flos": 24676232647680.0, + "grad_norm": 2.448042624905063, + "language_loss": 0.68216991, + "learning_rate": 6.994068948661592e-08, + "loss": 0.70428872, + "num_input_tokens_seen": 165035575, + "step": 7635, + "time_per_iteration": 2.717848777770996 + }, + { + "auxiliary_loss_clip": 0.01178387, + "auxiliary_loss_mlp": 0.01024517, + "balance_loss_clip": 1.04797459, + "balance_loss_mlp": 1.01622045, + "epoch": 0.9181747129200986, + "flos": 16727478301440.0, + "grad_norm": 1.9736718989535351, + "language_loss": 0.76875544, + "learning_rate": 6.973663568654142e-08, + "loss": 0.79078448, + "num_input_tokens_seen": 165053280, + "step": 7636, + "time_per_iteration": 2.5694196224212646 + }, + { + "auxiliary_loss_clip": 0.01166383, + "auxiliary_loss_mlp": 0.01025368, + "balance_loss_clip": 1.04687726, + "balance_loss_mlp": 1.01828051, + "epoch": 0.9182949558107377, + "flos": 24271626873600.0, + "grad_norm": 2.6824889317539395, + "language_loss": 0.65460968, + "learning_rate": 6.953287470588386e-08, + "loss": 0.6765272, + "num_input_tokens_seen": 165071235, + "step": 7637, + "time_per_iteration": 2.653327465057373 + }, + { + "auxiliary_loss_clip": 0.01179191, + "auxiliary_loss_mlp": 0.0102652, + "balance_loss_clip": 1.04435265, + "balance_loss_mlp": 1.01870561, + "epoch": 0.9184151987013768, + "flos": 22082117443200.0, + "grad_norm": 2.1623726469375755, + "language_loss": 0.85588366, + "learning_rate": 6.932940657555452e-08, + "loss": 0.87794077, + "num_input_tokens_seen": 165087365, + "step": 7638, + "time_per_iteration": 2.5825610160827637 + }, + { + "auxiliary_loss_clip": 0.01161674, + "auxiliary_loss_mlp": 0.01023743, + "balance_loss_clip": 1.04539073, + "balance_loss_mlp": 1.01721644, + "epoch": 0.9185354415920158, + "flos": 32166732257280.0, + "grad_norm": 1.4951725563036313, + "language_loss": 0.76348734, + "learning_rate": 6.912623132641938e-08, + "loss": 0.78534156, + "num_input_tokens_seen": 165112455, + "step": 7639, + "time_per_iteration": 2.7463901042938232 + }, + { + "auxiliary_loss_clip": 0.01185554, + "auxiliary_loss_mlp": 0.01028088, + "balance_loss_clip": 1.04391122, + "balance_loss_mlp": 1.02021384, + "epoch": 0.918655684482655, + "flos": 20997831542400.0, + "grad_norm": 1.9963680235136336, + "language_loss": 0.76933783, + "learning_rate": 6.892334898929952e-08, + "loss": 0.79147422, + "num_input_tokens_seen": 165132700, + "step": 7640, + "time_per_iteration": 2.6399028301239014 + }, + { + "auxiliary_loss_clip": 0.01167196, + "auxiliary_loss_mlp": 0.01026475, + "balance_loss_clip": 1.04289401, + "balance_loss_mlp": 1.01916742, + "epoch": 0.918775927373294, + "flos": 15560704817280.0, + "grad_norm": 3.2101658818412413, + "language_loss": 0.84704006, + "learning_rate": 6.872075959497236e-08, + "loss": 0.86897677, + "num_input_tokens_seen": 165151475, + "step": 7641, + "time_per_iteration": 2.6764779090881348 + }, + { + "auxiliary_loss_clip": 0.01177039, + "auxiliary_loss_mlp": 0.01022008, + "balance_loss_clip": 1.04380214, + "balance_loss_mlp": 1.01493919, + "epoch": 0.9188961702639331, + "flos": 29934057657600.0, + "grad_norm": 1.7792698315007114, + "language_loss": 0.82608986, + "learning_rate": 6.85184631741702e-08, + "loss": 0.84808028, + "num_input_tokens_seen": 165172040, + "step": 7642, + "time_per_iteration": 2.6268813610076904 + }, + { + "auxiliary_loss_clip": 0.01174889, + "auxiliary_loss_mlp": 0.01026186, + "balance_loss_clip": 1.04559755, + "balance_loss_mlp": 1.01872659, + "epoch": 0.9190164131545723, + "flos": 20701244943360.0, + "grad_norm": 1.815818113124842, + "language_loss": 0.77205485, + "learning_rate": 6.831645975758161e-08, + "loss": 0.79406559, + "num_input_tokens_seen": 165189980, + "step": 7643, + "time_per_iteration": 2.590512990951538 + }, + { + "auxiliary_loss_clip": 0.0117761, + "auxiliary_loss_mlp": 0.01024339, + "balance_loss_clip": 1.04586279, + "balance_loss_mlp": 1.01662064, + "epoch": 0.9191366560452113, + "flos": 25629912696960.0, + "grad_norm": 1.7522653936788855, + "language_loss": 0.6752423, + "learning_rate": 6.811474937585026e-08, + "loss": 0.69726175, + "num_input_tokens_seen": 165209770, + "step": 7644, + "time_per_iteration": 3.6651031970977783 + }, + { + "auxiliary_loss_clip": 0.01189704, + "auxiliary_loss_mlp": 0.01023128, + "balance_loss_clip": 1.04224849, + "balance_loss_mlp": 1.01648164, + "epoch": 0.9192568989358504, + "flos": 21434325615360.0, + "grad_norm": 1.6976534981068192, + "language_loss": 0.79248405, + "learning_rate": 6.79133320595755e-08, + "loss": 0.81461239, + "num_input_tokens_seen": 165229690, + "step": 7645, + "time_per_iteration": 2.718583583831787 + }, + { + "auxiliary_loss_clip": 0.01189157, + "auxiliary_loss_mlp": 0.01025514, + "balance_loss_clip": 1.04694307, + "balance_loss_mlp": 1.01839733, + "epoch": 0.9193771418264896, + "flos": 23185078416000.0, + "grad_norm": 1.6781816319082703, + "language_loss": 0.75271654, + "learning_rate": 6.771220783931198e-08, + "loss": 0.77486324, + "num_input_tokens_seen": 165249850, + "step": 7646, + "time_per_iteration": 2.653042793273926 + }, + { + "auxiliary_loss_clip": 0.01156855, + "auxiliary_loss_mlp": 0.01373574, + "balance_loss_clip": 1.02167749, + "balance_loss_mlp": 0.99974674, + "epoch": 0.9194973847171286, + "flos": 70582963184640.0, + "grad_norm": 0.8291854618155998, + "language_loss": 0.64588392, + "learning_rate": 6.751137674556994e-08, + "loss": 0.67118812, + "num_input_tokens_seen": 165310235, + "step": 7647, + "time_per_iteration": 4.573877334594727 + }, + { + "auxiliary_loss_clip": 0.01178405, + "auxiliary_loss_mlp": 0.01024701, + "balance_loss_clip": 1.04369342, + "balance_loss_mlp": 1.017802, + "epoch": 0.9196176276077677, + "flos": 14720682378240.0, + "grad_norm": 1.9950618405868943, + "language_loss": 0.77709055, + "learning_rate": 6.731083880881572e-08, + "loss": 0.79912162, + "num_input_tokens_seen": 165326455, + "step": 7648, + "time_per_iteration": 4.053877353668213 + }, + { + "auxiliary_loss_clip": 0.01185543, + "auxiliary_loss_mlp": 0.01029804, + "balance_loss_clip": 1.04534626, + "balance_loss_mlp": 1.02325654, + "epoch": 0.9197378704984068, + "flos": 23294893271040.0, + "grad_norm": 6.4723927851160745, + "language_loss": 0.81224155, + "learning_rate": 6.711059405947072e-08, + "loss": 0.83439493, + "num_input_tokens_seen": 165344645, + "step": 7649, + "time_per_iteration": 3.634269952774048 + }, + { + "auxiliary_loss_clip": 0.0118975, + "auxiliary_loss_mlp": 0.01020914, + "balance_loss_clip": 1.04298007, + "balance_loss_mlp": 1.01394284, + "epoch": 0.9198581133890459, + "flos": 20302564913280.0, + "grad_norm": 2.8877738849871544, + "language_loss": 0.76896846, + "learning_rate": 6.691064252791156e-08, + "loss": 0.79107511, + "num_input_tokens_seen": 165364120, + "step": 7650, + "time_per_iteration": 2.7106704711914062 + }, + { + "auxiliary_loss_clip": 0.01197622, + "auxiliary_loss_mlp": 0.01023972, + "balance_loss_clip": 1.04243445, + "balance_loss_mlp": 1.01625907, + "epoch": 0.9199783562796849, + "flos": 17675663569920.0, + "grad_norm": 1.5515824261706093, + "language_loss": 0.77851599, + "learning_rate": 6.67109842444713e-08, + "loss": 0.80073196, + "num_input_tokens_seen": 165383050, + "step": 7651, + "time_per_iteration": 2.65402889251709 + }, + { + "auxiliary_loss_clip": 0.01175342, + "auxiliary_loss_mlp": 0.01378517, + "balance_loss_clip": 1.04667401, + "balance_loss_mlp": 1.00012743, + "epoch": 0.9200985991703241, + "flos": 17676022705920.0, + "grad_norm": 1.968944858262557, + "language_loss": 0.76692903, + "learning_rate": 6.651161923943704e-08, + "loss": 0.79246759, + "num_input_tokens_seen": 165400955, + "step": 7652, + "time_per_iteration": 2.620016574859619 + }, + { + "auxiliary_loss_clip": 0.01167718, + "auxiliary_loss_mlp": 0.01023063, + "balance_loss_clip": 1.04372084, + "balance_loss_mlp": 1.01579666, + "epoch": 0.9202188420609632, + "flos": 20996574566400.0, + "grad_norm": 2.2751526622841545, + "language_loss": 0.76567471, + "learning_rate": 6.631254754305326e-08, + "loss": 0.78758252, + "num_input_tokens_seen": 165420415, + "step": 7653, + "time_per_iteration": 2.60598087310791 + }, + { + "auxiliary_loss_clip": 0.01167683, + "auxiliary_loss_mlp": 0.01023785, + "balance_loss_clip": 1.04592764, + "balance_loss_mlp": 1.01644468, + "epoch": 0.9203390849516022, + "flos": 13918222586880.0, + "grad_norm": 2.120863091498633, + "language_loss": 0.78027451, + "learning_rate": 6.611376918551848e-08, + "loss": 0.80218923, + "num_input_tokens_seen": 165439200, + "step": 7654, + "time_per_iteration": 2.6487441062927246 + }, + { + "auxiliary_loss_clip": 0.01191437, + "auxiliary_loss_mlp": 0.01378674, + "balance_loss_clip": 1.04269671, + "balance_loss_mlp": 1.00011647, + "epoch": 0.9204593278422414, + "flos": 21175912195200.0, + "grad_norm": 1.9872779689450901, + "language_loss": 0.79460686, + "learning_rate": 6.591528419698744e-08, + "loss": 0.82030797, + "num_input_tokens_seen": 165458985, + "step": 7655, + "time_per_iteration": 3.6161162853240967 + }, + { + "auxiliary_loss_clip": 0.01182735, + "auxiliary_loss_mlp": 0.01025523, + "balance_loss_clip": 1.04079413, + "balance_loss_mlp": 1.01861775, + "epoch": 0.9205795707328804, + "flos": 14501375890560.0, + "grad_norm": 2.6494101954492266, + "language_loss": 0.83514285, + "learning_rate": 6.571709260756986e-08, + "loss": 0.85722542, + "num_input_tokens_seen": 165475630, + "step": 7656, + "time_per_iteration": 2.615906000137329 + }, + { + "auxiliary_loss_clip": 0.01177469, + "auxiliary_loss_mlp": 0.01023522, + "balance_loss_clip": 1.04717779, + "balance_loss_mlp": 1.01651859, + "epoch": 0.9206998136235195, + "flos": 22417559579520.0, + "grad_norm": 2.155564871242244, + "language_loss": 0.76358938, + "learning_rate": 6.551919444733122e-08, + "loss": 0.78559935, + "num_input_tokens_seen": 165493445, + "step": 7657, + "time_per_iteration": 2.649564504623413 + }, + { + "auxiliary_loss_clip": 0.01186121, + "auxiliary_loss_mlp": 0.01026127, + "balance_loss_clip": 1.04653251, + "balance_loss_mlp": 1.01859951, + "epoch": 0.9208200565141585, + "flos": 53358407544960.0, + "grad_norm": 2.150301923906477, + "language_loss": 0.65891325, + "learning_rate": 6.53215897462931e-08, + "loss": 0.6810357, + "num_input_tokens_seen": 165517200, + "step": 7658, + "time_per_iteration": 2.892683506011963 + }, + { + "auxiliary_loss_clip": 0.01172057, + "auxiliary_loss_mlp": 0.01025512, + "balance_loss_clip": 1.04444242, + "balance_loss_mlp": 1.01842833, + "epoch": 0.9209402994047977, + "flos": 30589139946240.0, + "grad_norm": 3.2574581241948204, + "language_loss": 0.75046092, + "learning_rate": 6.512427853443103e-08, + "loss": 0.77243662, + "num_input_tokens_seen": 165539280, + "step": 7659, + "time_per_iteration": 2.7148525714874268 + }, + { + "auxiliary_loss_clip": 0.01177181, + "auxiliary_loss_mlp": 0.01024929, + "balance_loss_clip": 1.04526472, + "balance_loss_mlp": 1.01774669, + "epoch": 0.9210605422954368, + "flos": 29132711187840.0, + "grad_norm": 1.487771169476471, + "language_loss": 0.75862914, + "learning_rate": 6.492726084167799e-08, + "loss": 0.7806502, + "num_input_tokens_seen": 165561395, + "step": 7660, + "time_per_iteration": 2.6635515689849854 + }, + { + "auxiliary_loss_clip": 0.0106002, + "auxiliary_loss_mlp": 0.01002512, + "balance_loss_clip": 1.00897741, + "balance_loss_mlp": 1.00149238, + "epoch": 0.9211807851860758, + "flos": 54853838472960.0, + "grad_norm": 0.7764271375059278, + "language_loss": 0.57535434, + "learning_rate": 6.473053669792072e-08, + "loss": 0.59597969, + "num_input_tokens_seen": 165616085, + "step": 7661, + "time_per_iteration": 3.1392672061920166 + }, + { + "auxiliary_loss_clip": 0.01173054, + "auxiliary_loss_mlp": 0.01025677, + "balance_loss_clip": 1.0438379, + "balance_loss_mlp": 1.01840568, + "epoch": 0.921301028076715, + "flos": 19201974238080.0, + "grad_norm": 4.42536018255282, + "language_loss": 0.72561103, + "learning_rate": 6.453410613300248e-08, + "loss": 0.74759829, + "num_input_tokens_seen": 165634015, + "step": 7662, + "time_per_iteration": 2.6199100017547607 + }, + { + "auxiliary_loss_clip": 0.01212793, + "auxiliary_loss_mlp": 0.01022888, + "balance_loss_clip": 1.04209757, + "balance_loss_mlp": 1.01571798, + "epoch": 0.921421270967354, + "flos": 27526893765120.0, + "grad_norm": 1.7936314859156268, + "language_loss": 0.58115268, + "learning_rate": 6.43379691767214e-08, + "loss": 0.60350955, + "num_input_tokens_seen": 165653220, + "step": 7663, + "time_per_iteration": 2.987335681915283 + }, + { + "auxiliary_loss_clip": 0.01110841, + "auxiliary_loss_mlp": 0.01000457, + "balance_loss_clip": 1.0086658, + "balance_loss_mlp": 0.99927104, + "epoch": 0.9215415138579931, + "flos": 70209311955840.0, + "grad_norm": 0.7211815051765912, + "language_loss": 0.55128974, + "learning_rate": 6.414212585883105e-08, + "loss": 0.57240272, + "num_input_tokens_seen": 165715850, + "step": 7664, + "time_per_iteration": 3.7366700172424316 + }, + { + "auxiliary_loss_clip": 0.01188535, + "auxiliary_loss_mlp": 0.01030534, + "balance_loss_clip": 1.04441047, + "balance_loss_mlp": 1.02351284, + "epoch": 0.9216617567486323, + "flos": 35553107790720.0, + "grad_norm": 1.4609687889538947, + "language_loss": 0.69610399, + "learning_rate": 6.394657620904143e-08, + "loss": 0.71829462, + "num_input_tokens_seen": 165738960, + "step": 7665, + "time_per_iteration": 2.756194829940796 + }, + { + "auxiliary_loss_clip": 0.0117272, + "auxiliary_loss_mlp": 0.01025899, + "balance_loss_clip": 1.04837286, + "balance_loss_mlp": 1.01788807, + "epoch": 0.9217819996392713, + "flos": 29533330552320.0, + "grad_norm": 1.7776632030146027, + "language_loss": 0.71466064, + "learning_rate": 6.375132025701657e-08, + "loss": 0.73664683, + "num_input_tokens_seen": 165761260, + "step": 7666, + "time_per_iteration": 2.74495530128479 + }, + { + "auxiliary_loss_clip": 0.01170376, + "auxiliary_loss_mlp": 0.0102431, + "balance_loss_clip": 1.04864454, + "balance_loss_mlp": 1.01680613, + "epoch": 0.9219022425299104, + "flos": 14574669592320.0, + "grad_norm": 2.4073193239029727, + "language_loss": 0.6910947, + "learning_rate": 6.355635803237724e-08, + "loss": 0.71304154, + "num_input_tokens_seen": 165776960, + "step": 7667, + "time_per_iteration": 2.6223180294036865 + }, + { + "auxiliary_loss_clip": 0.01173758, + "auxiliary_loss_mlp": 0.0102545, + "balance_loss_clip": 1.04528916, + "balance_loss_mlp": 1.01846147, + "epoch": 0.9220224854205495, + "flos": 18077503996800.0, + "grad_norm": 3.6885590744371792, + "language_loss": 0.80262566, + "learning_rate": 6.336168956469867e-08, + "loss": 0.82461774, + "num_input_tokens_seen": 165795435, + "step": 7668, + "time_per_iteration": 2.6044728755950928 + }, + { + "auxiliary_loss_clip": 0.01176908, + "auxiliary_loss_mlp": 0.01024465, + "balance_loss_clip": 1.04262602, + "balance_loss_mlp": 1.01746464, + "epoch": 0.9221427283111886, + "flos": 24790464875520.0, + "grad_norm": 1.6423901863418509, + "language_loss": 0.72085875, + "learning_rate": 6.316731488351168e-08, + "loss": 0.74287248, + "num_input_tokens_seen": 165816625, + "step": 7669, + "time_per_iteration": 3.6488425731658936 + }, + { + "auxiliary_loss_clip": 0.01176226, + "auxiliary_loss_mlp": 0.01028258, + "balance_loss_clip": 1.0456053, + "balance_loss_mlp": 1.02106118, + "epoch": 0.9222629712018277, + "flos": 13845036625920.0, + "grad_norm": 1.9243572781106366, + "language_loss": 0.63064432, + "learning_rate": 6.297323401830334e-08, + "loss": 0.6526891, + "num_input_tokens_seen": 165835410, + "step": 7670, + "time_per_iteration": 2.545971155166626 + }, + { + "auxiliary_loss_clip": 0.01177987, + "auxiliary_loss_mlp": 0.01026286, + "balance_loss_clip": 1.04486692, + "balance_loss_mlp": 1.01945496, + "epoch": 0.9223832140924668, + "flos": 21616177196160.0, + "grad_norm": 1.849112503992362, + "language_loss": 0.69166207, + "learning_rate": 6.277944699851523e-08, + "loss": 0.71370476, + "num_input_tokens_seen": 165854930, + "step": 7671, + "time_per_iteration": 2.6314375400543213 + }, + { + "auxiliary_loss_clip": 0.01168616, + "auxiliary_loss_mlp": 0.01027357, + "balance_loss_clip": 1.04739904, + "balance_loss_mlp": 1.0201509, + "epoch": 0.9225034569831059, + "flos": 21142084561920.0, + "grad_norm": 1.7867240911212325, + "language_loss": 0.73294008, + "learning_rate": 6.25859538535447e-08, + "loss": 0.7548998, + "num_input_tokens_seen": 165875725, + "step": 7672, + "time_per_iteration": 2.565683126449585 + }, + { + "auxiliary_loss_clip": 0.01182356, + "auxiliary_loss_mlp": 0.0102623, + "balance_loss_clip": 1.04389834, + "balance_loss_mlp": 1.01883924, + "epoch": 0.9226236998737449, + "flos": 12495046844160.0, + "grad_norm": 2.8641864881281256, + "language_loss": 0.78413033, + "learning_rate": 6.239275461274474e-08, + "loss": 0.80621618, + "num_input_tokens_seen": 165892100, + "step": 7673, + "time_per_iteration": 3.537278652191162 + }, + { + "auxiliary_loss_clip": 0.01177088, + "auxiliary_loss_mlp": 0.01027261, + "balance_loss_clip": 1.04667294, + "balance_loss_mlp": 1.02010202, + "epoch": 0.9227439427643841, + "flos": 26214071581440.0, + "grad_norm": 1.6442020384691833, + "language_loss": 0.85953295, + "learning_rate": 6.219984930542299e-08, + "loss": 0.88157642, + "num_input_tokens_seen": 165912840, + "step": 7674, + "time_per_iteration": 2.649054765701294 + }, + { + "auxiliary_loss_clip": 0.01178353, + "auxiliary_loss_mlp": 0.01021615, + "balance_loss_clip": 1.04624391, + "balance_loss_mlp": 1.01453125, + "epoch": 0.9228641856550232, + "flos": 17967581400960.0, + "grad_norm": 2.0272248166418043, + "language_loss": 0.76079547, + "learning_rate": 6.200723796084383e-08, + "loss": 0.78279519, + "num_input_tokens_seen": 165930935, + "step": 7675, + "time_per_iteration": 3.6455888748168945 + }, + { + "auxiliary_loss_clip": 0.01101234, + "auxiliary_loss_mlp": 0.01000931, + "balance_loss_clip": 1.00914168, + "balance_loss_mlp": 0.99982196, + "epoch": 0.9229844285456622, + "flos": 70420609710720.0, + "grad_norm": 0.7644146054501015, + "language_loss": 0.62995791, + "learning_rate": 6.181492060822546e-08, + "loss": 0.65097952, + "num_input_tokens_seen": 165991110, + "step": 7676, + "time_per_iteration": 3.253247022628784 + }, + { + "auxiliary_loss_clip": 0.01198505, + "auxiliary_loss_mlp": 0.01024281, + "balance_loss_clip": 1.04044402, + "balance_loss_mlp": 1.01703334, + "epoch": 0.9231046714363014, + "flos": 17967832796160.0, + "grad_norm": 2.0747918759853516, + "language_loss": 0.81729424, + "learning_rate": 6.162289727674274e-08, + "loss": 0.83952212, + "num_input_tokens_seen": 166008790, + "step": 7677, + "time_per_iteration": 2.7503082752227783 + }, + { + "auxiliary_loss_clip": 0.0119434, + "auxiliary_loss_mlp": 0.01024515, + "balance_loss_clip": 1.04209423, + "balance_loss_mlp": 1.01778591, + "epoch": 0.9232249143269404, + "flos": 17858233422720.0, + "grad_norm": 3.171409048800749, + "language_loss": 0.88040006, + "learning_rate": 6.143116799552527e-08, + "loss": 0.90258861, + "num_input_tokens_seen": 166025035, + "step": 7678, + "time_per_iteration": 2.6987507343292236 + }, + { + "auxiliary_loss_clip": 0.01181606, + "auxiliary_loss_mlp": 0.01029766, + "balance_loss_clip": 1.04845548, + "balance_loss_mlp": 1.02242279, + "epoch": 0.9233451572175795, + "flos": 23404384903680.0, + "grad_norm": 2.406544775956482, + "language_loss": 0.55989015, + "learning_rate": 6.123973279365802e-08, + "loss": 0.58200383, + "num_input_tokens_seen": 166044010, + "step": 7679, + "time_per_iteration": 2.666292667388916 + }, + { + "auxiliary_loss_clip": 0.01181007, + "auxiliary_loss_mlp": 0.01027369, + "balance_loss_clip": 1.04778647, + "balance_loss_mlp": 1.02066612, + "epoch": 0.9234654001082186, + "flos": 17999326045440.0, + "grad_norm": 2.740011276319156, + "language_loss": 0.77909732, + "learning_rate": 6.10485917001824e-08, + "loss": 0.80118108, + "num_input_tokens_seen": 166061865, + "step": 7680, + "time_per_iteration": 3.5075716972351074 + }, + { + "auxiliary_loss_clip": 0.01187974, + "auxiliary_loss_mlp": 0.01021873, + "balance_loss_clip": 1.04346251, + "balance_loss_mlp": 1.01461339, + "epoch": 0.9235856429988577, + "flos": 24750747411840.0, + "grad_norm": 1.4975340937261987, + "language_loss": 0.81104207, + "learning_rate": 6.085774474409322e-08, + "loss": 0.83314049, + "num_input_tokens_seen": 166082425, + "step": 7681, + "time_per_iteration": 2.7795047760009766 + }, + { + "auxiliary_loss_clip": 0.01185538, + "auxiliary_loss_mlp": 0.01028216, + "balance_loss_clip": 1.04746616, + "balance_loss_mlp": 1.02124834, + "epoch": 0.9237058858894968, + "flos": 14099894599680.0, + "grad_norm": 1.9467823659646608, + "language_loss": 0.70577973, + "learning_rate": 6.066719195434267e-08, + "loss": 0.72791725, + "num_input_tokens_seen": 166100225, + "step": 7682, + "time_per_iteration": 2.5578393936157227 + }, + { + "auxiliary_loss_clip": 0.01180075, + "auxiliary_loss_mlp": 0.01028172, + "balance_loss_clip": 1.04680061, + "balance_loss_mlp": 1.02024126, + "epoch": 0.9238261287801359, + "flos": 28694529175680.0, + "grad_norm": 2.0540086574115937, + "language_loss": 0.6663233, + "learning_rate": 6.047693335983717e-08, + "loss": 0.68840575, + "num_input_tokens_seen": 166122570, + "step": 7683, + "time_per_iteration": 2.7147469520568848 + }, + { + "auxiliary_loss_clip": 0.01178191, + "auxiliary_loss_mlp": 0.01024017, + "balance_loss_clip": 1.04417396, + "balance_loss_mlp": 1.01710308, + "epoch": 0.923946371670775, + "flos": 23111856541440.0, + "grad_norm": 3.719094461814953, + "language_loss": 0.82437426, + "learning_rate": 6.028696898943853e-08, + "loss": 0.84639639, + "num_input_tokens_seen": 166141630, + "step": 7684, + "time_per_iteration": 2.662731170654297 + }, + { + "auxiliary_loss_clip": 0.01182806, + "auxiliary_loss_mlp": 0.01378881, + "balance_loss_clip": 1.04194558, + "balance_loss_mlp": 1.00014126, + "epoch": 0.924066614561414, + "flos": 21867120587520.0, + "grad_norm": 1.9478755892772863, + "language_loss": 0.70595187, + "learning_rate": 6.00972988719648e-08, + "loss": 0.73156875, + "num_input_tokens_seen": 166159865, + "step": 7685, + "time_per_iteration": 2.6689705848693848 + }, + { + "auxiliary_loss_clip": 0.0119575, + "auxiliary_loss_mlp": 0.01378849, + "balance_loss_clip": 1.04321074, + "balance_loss_mlp": 1.00017476, + "epoch": 0.9241868574520532, + "flos": 28511887495680.0, + "grad_norm": 2.5227790456789423, + "language_loss": 0.70612997, + "learning_rate": 5.990792303618807e-08, + "loss": 0.7318759, + "num_input_tokens_seen": 166179445, + "step": 7686, + "time_per_iteration": 2.740675210952759 + }, + { + "auxiliary_loss_clip": 0.01191351, + "auxiliary_loss_mlp": 0.01026372, + "balance_loss_clip": 1.04137814, + "balance_loss_mlp": 1.01959217, + "epoch": 0.9243071003426923, + "flos": 30518324282880.0, + "grad_norm": 1.7000744052033627, + "language_loss": 0.69360518, + "learning_rate": 5.971884151083695e-08, + "loss": 0.71578246, + "num_input_tokens_seen": 166201855, + "step": 7687, + "time_per_iteration": 2.7787866592407227 + }, + { + "auxiliary_loss_clip": 0.01185056, + "auxiliary_loss_mlp": 0.01024657, + "balance_loss_clip": 1.04370308, + "balance_loss_mlp": 1.01788044, + "epoch": 0.9244273432333313, + "flos": 28658331244800.0, + "grad_norm": 1.983781193230455, + "language_loss": 0.74187797, + "learning_rate": 5.9530054324595124e-08, + "loss": 0.76397514, + "num_input_tokens_seen": 166221970, + "step": 7688, + "time_per_iteration": 2.714290142059326 + }, + { + "auxiliary_loss_clip": 0.01068578, + "auxiliary_loss_mlp": 0.01373733, + "balance_loss_clip": 1.00921297, + "balance_loss_mlp": 0.99979287, + "epoch": 0.9245475861239704, + "flos": 66230589237120.0, + "grad_norm": 0.873955429003935, + "language_loss": 0.5750761, + "learning_rate": 5.934156150610103e-08, + "loss": 0.59949923, + "num_input_tokens_seen": 166279335, + "step": 7689, + "time_per_iteration": 3.2330780029296875 + }, + { + "auxiliary_loss_clip": 0.01181058, + "auxiliary_loss_mlp": 0.0103093, + "balance_loss_clip": 1.04362595, + "balance_loss_mlp": 1.0236578, + "epoch": 0.9246678290146095, + "flos": 24239918142720.0, + "grad_norm": 2.270082752709778, + "language_loss": 0.79216605, + "learning_rate": 5.915336308394914e-08, + "loss": 0.81428587, + "num_input_tokens_seen": 166298170, + "step": 7690, + "time_per_iteration": 2.613739013671875 + }, + { + "auxiliary_loss_clip": 0.01170833, + "auxiliary_loss_mlp": 0.01025221, + "balance_loss_clip": 1.04431796, + "balance_loss_mlp": 1.01850367, + "epoch": 0.9247880719052486, + "flos": 18988808976000.0, + "grad_norm": 1.682175295223415, + "language_loss": 0.76942503, + "learning_rate": 5.89654590866886e-08, + "loss": 0.79138553, + "num_input_tokens_seen": 166317670, + "step": 7691, + "time_per_iteration": 2.53904390335083 + }, + { + "auxiliary_loss_clip": 0.01217694, + "auxiliary_loss_mlp": 0.01027025, + "balance_loss_clip": 1.04430556, + "balance_loss_mlp": 1.01900864, + "epoch": 0.9249083147958876, + "flos": 24024095274240.0, + "grad_norm": 2.407655237901151, + "language_loss": 0.88606131, + "learning_rate": 5.877784954282483e-08, + "loss": 0.90850848, + "num_input_tokens_seen": 166337010, + "step": 7692, + "time_per_iteration": 2.8267343044281006 + }, + { + "auxiliary_loss_clip": 0.01179585, + "auxiliary_loss_mlp": 0.01029647, + "balance_loss_clip": 1.04558527, + "balance_loss_mlp": 1.02190423, + "epoch": 0.9250285576865268, + "flos": 30773972355840.0, + "grad_norm": 2.1037495065219627, + "language_loss": 0.72253656, + "learning_rate": 5.8590534480817963e-08, + "loss": 0.74462891, + "num_input_tokens_seen": 166358735, + "step": 7693, + "time_per_iteration": 2.6698756217956543 + }, + { + "auxiliary_loss_clip": 0.01167962, + "auxiliary_loss_mlp": 0.01024705, + "balance_loss_clip": 1.04836643, + "balance_loss_mlp": 1.01749861, + "epoch": 0.9251488005771659, + "flos": 10633581348480.0, + "grad_norm": 2.1772638928433916, + "language_loss": 0.72409284, + "learning_rate": 5.840351392908349e-08, + "loss": 0.74601948, + "num_input_tokens_seen": 166374455, + "step": 7694, + "time_per_iteration": 2.5729236602783203 + }, + { + "auxiliary_loss_clip": 0.01191128, + "auxiliary_loss_mlp": 0.01378559, + "balance_loss_clip": 1.04468584, + "balance_loss_mlp": 1.00018346, + "epoch": 0.9252690434678049, + "flos": 23586416052480.0, + "grad_norm": 2.498990277537258, + "language_loss": 0.70782667, + "learning_rate": 5.821678791599205e-08, + "loss": 0.73352355, + "num_input_tokens_seen": 166393900, + "step": 7695, + "time_per_iteration": 3.6639485359191895 + }, + { + "auxiliary_loss_clip": 0.01186108, + "auxiliary_loss_mlp": 0.01022171, + "balance_loss_clip": 1.04637229, + "balance_loss_mlp": 1.01500988, + "epoch": 0.9253892863584441, + "flos": 21469158829440.0, + "grad_norm": 2.979054351588214, + "language_loss": 0.8070696, + "learning_rate": 5.803035646986965e-08, + "loss": 0.82915235, + "num_input_tokens_seen": 166413235, + "step": 7696, + "time_per_iteration": 2.663562297821045 + }, + { + "auxiliary_loss_clip": 0.01169663, + "auxiliary_loss_mlp": 0.01024309, + "balance_loss_clip": 1.0468874, + "balance_loss_mlp": 1.01678717, + "epoch": 0.9255095292490831, + "flos": 17456680304640.0, + "grad_norm": 3.0281031801484697, + "language_loss": 0.67634082, + "learning_rate": 5.7844219618998766e-08, + "loss": 0.69828057, + "num_input_tokens_seen": 166427560, + "step": 7697, + "time_per_iteration": 2.5029969215393066 + }, + { + "auxiliary_loss_clip": 0.01177573, + "auxiliary_loss_mlp": 0.01021493, + "balance_loss_clip": 1.03970385, + "balance_loss_mlp": 1.0144062, + "epoch": 0.9256297721397222, + "flos": 24750675584640.0, + "grad_norm": 1.7432972857348334, + "language_loss": 0.71493614, + "learning_rate": 5.765837739161505e-08, + "loss": 0.73692685, + "num_input_tokens_seen": 166446680, + "step": 7698, + "time_per_iteration": 2.7021374702453613 + }, + { + "auxiliary_loss_clip": 0.01195888, + "auxiliary_loss_mlp": 0.01023417, + "balance_loss_clip": 1.04209256, + "balance_loss_mlp": 1.01625824, + "epoch": 0.9257500150303614, + "flos": 23112215677440.0, + "grad_norm": 1.5852229053247051, + "language_loss": 0.74280834, + "learning_rate": 5.7472829815911504e-08, + "loss": 0.7650013, + "num_input_tokens_seen": 166465505, + "step": 7699, + "time_per_iteration": 3.758185386657715 + }, + { + "auxiliary_loss_clip": 0.01177198, + "auxiliary_loss_mlp": 0.01031482, + "balance_loss_clip": 1.04225802, + "balance_loss_mlp": 1.02391219, + "epoch": 0.9258702579210004, + "flos": 22564685687040.0, + "grad_norm": 2.2337231373180244, + "language_loss": 0.81568778, + "learning_rate": 5.7287576920035164e-08, + "loss": 0.83777457, + "num_input_tokens_seen": 166484520, + "step": 7700, + "time_per_iteration": 2.637949228286743 + }, + { + "auxiliary_loss_clip": 0.0119022, + "auxiliary_loss_mlp": 0.01022954, + "balance_loss_clip": 1.04264855, + "balance_loss_mlp": 1.01600432, + "epoch": 0.9259905008116395, + "flos": 30004298703360.0, + "grad_norm": 1.7435176646382122, + "language_loss": 0.76689994, + "learning_rate": 5.7102618732088435e-08, + "loss": 0.78903168, + "num_input_tokens_seen": 166503850, + "step": 7701, + "time_per_iteration": 3.574610948562622 + }, + { + "auxiliary_loss_clip": 0.01188857, + "auxiliary_loss_mlp": 0.01019035, + "balance_loss_clip": 1.04481065, + "balance_loss_mlp": 1.01207304, + "epoch": 0.9261107437022786, + "flos": 24572128055040.0, + "grad_norm": 1.5150337981344593, + "language_loss": 0.74768424, + "learning_rate": 5.6917955280130216e-08, + "loss": 0.76976311, + "num_input_tokens_seen": 166525330, + "step": 7702, + "time_per_iteration": 2.65103816986084 + }, + { + "auxiliary_loss_clip": 0.01176715, + "auxiliary_loss_mlp": 0.01025239, + "balance_loss_clip": 1.0480597, + "balance_loss_mlp": 1.01750851, + "epoch": 0.9262309865929177, + "flos": 22018448586240.0, + "grad_norm": 5.626523180990815, + "language_loss": 0.71969098, + "learning_rate": 5.6733586592172755e-08, + "loss": 0.74171054, + "num_input_tokens_seen": 166544825, + "step": 7703, + "time_per_iteration": 2.584062099456787 + }, + { + "auxiliary_loss_clip": 0.01176818, + "auxiliary_loss_mlp": 0.01378185, + "balance_loss_clip": 1.04060733, + "balance_loss_mlp": 1.00010514, + "epoch": 0.9263512294835567, + "flos": 20339481116160.0, + "grad_norm": 1.95491690161956, + "language_loss": 0.79927486, + "learning_rate": 5.6549512696185244e-08, + "loss": 0.82482493, + "num_input_tokens_seen": 166563325, + "step": 7704, + "time_per_iteration": 2.682770013809204 + }, + { + "auxiliary_loss_clip": 0.01165399, + "auxiliary_loss_mlp": 0.01025062, + "balance_loss_clip": 1.04754066, + "balance_loss_mlp": 1.01796937, + "epoch": 0.9264714723741959, + "flos": 21215378263680.0, + "grad_norm": 1.6030387356103857, + "language_loss": 0.68077433, + "learning_rate": 5.636573362009156e-08, + "loss": 0.70267892, + "num_input_tokens_seen": 166583385, + "step": 7705, + "time_per_iteration": 2.551318883895874 + }, + { + "auxiliary_loss_clip": 0.01169453, + "auxiliary_loss_mlp": 0.01020009, + "balance_loss_clip": 1.04685724, + "balance_loss_mlp": 1.01249266, + "epoch": 0.926591715264835, + "flos": 18004964480640.0, + "grad_norm": 1.954262922582468, + "language_loss": 0.7693612, + "learning_rate": 5.618224939177074e-08, + "loss": 0.79125583, + "num_input_tokens_seen": 166601290, + "step": 7706, + "time_per_iteration": 3.5156848430633545 + }, + { + "auxiliary_loss_clip": 0.01171852, + "auxiliary_loss_mlp": 0.01027833, + "balance_loss_clip": 1.04282522, + "balance_loss_mlp": 1.02075756, + "epoch": 0.926711958155474, + "flos": 36167969825280.0, + "grad_norm": 1.726943564433637, + "language_loss": 0.7030288, + "learning_rate": 5.599906003905719e-08, + "loss": 0.72502565, + "num_input_tokens_seen": 166623835, + "step": 7707, + "time_per_iteration": 2.75311279296875 + }, + { + "auxiliary_loss_clip": 0.01171258, + "auxiliary_loss_mlp": 0.01029095, + "balance_loss_clip": 1.04675531, + "balance_loss_mlp": 1.02101862, + "epoch": 0.9268322010461132, + "flos": 21032736583680.0, + "grad_norm": 2.1765077738720016, + "language_loss": 0.81789136, + "learning_rate": 5.581616558974023e-08, + "loss": 0.83989489, + "num_input_tokens_seen": 166642400, + "step": 7708, + "time_per_iteration": 2.6756787300109863 + }, + { + "auxiliary_loss_clip": 0.01179395, + "auxiliary_loss_mlp": 0.01378997, + "balance_loss_clip": 1.04518962, + "balance_loss_mlp": 1.0001682, + "epoch": 0.9269524439367522, + "flos": 22964838174720.0, + "grad_norm": 2.0760942955845603, + "language_loss": 0.79088783, + "learning_rate": 5.5633566071565444e-08, + "loss": 0.81647176, + "num_input_tokens_seen": 166661640, + "step": 7709, + "time_per_iteration": 2.6174354553222656 + }, + { + "auxiliary_loss_clip": 0.01216256, + "auxiliary_loss_mlp": 0.01023813, + "balance_loss_clip": 1.04077864, + "balance_loss_mlp": 1.01624322, + "epoch": 0.9270726868273913, + "flos": 41975551468800.0, + "grad_norm": 1.9876357383591106, + "language_loss": 0.70604694, + "learning_rate": 5.5451261512232896e-08, + "loss": 0.72844768, + "num_input_tokens_seen": 166684320, + "step": 7710, + "time_per_iteration": 2.9070885181427 + }, + { + "auxiliary_loss_clip": 0.01179418, + "auxiliary_loss_mlp": 0.01028446, + "balance_loss_clip": 1.04463983, + "balance_loss_mlp": 1.0207088, + "epoch": 0.9271929297180305, + "flos": 19791771557760.0, + "grad_norm": 1.8351666535555773, + "language_loss": 0.62704945, + "learning_rate": 5.5269251939397576e-08, + "loss": 0.64912814, + "num_input_tokens_seen": 166703835, + "step": 7711, + "time_per_iteration": 2.5822901725769043 + }, + { + "auxiliary_loss_clip": 0.01192431, + "auxiliary_loss_mlp": 0.01024914, + "balance_loss_clip": 1.04016805, + "balance_loss_mlp": 1.0177319, + "epoch": 0.9273131726086695, + "flos": 19968343839360.0, + "grad_norm": 2.0736446064438505, + "language_loss": 0.76577711, + "learning_rate": 5.508753738067073e-08, + "loss": 0.78795052, + "num_input_tokens_seen": 166723375, + "step": 7712, + "time_per_iteration": 2.661391019821167 + }, + { + "auxiliary_loss_clip": 0.01178939, + "auxiliary_loss_mlp": 0.010192, + "balance_loss_clip": 1.04527342, + "balance_loss_mlp": 1.0119586, + "epoch": 0.9274334154993086, + "flos": 23258587599360.0, + "grad_norm": 1.9439029464696138, + "language_loss": 0.79250395, + "learning_rate": 5.4906117863617875e-08, + "loss": 0.81448543, + "num_input_tokens_seen": 166742760, + "step": 7713, + "time_per_iteration": 2.5700273513793945 + }, + { + "auxiliary_loss_clip": 0.01186224, + "auxiliary_loss_mlp": 0.01022452, + "balance_loss_clip": 1.04020703, + "balance_loss_mlp": 1.01592255, + "epoch": 0.9275536583899477, + "flos": 31795343585280.0, + "grad_norm": 2.0318917546833277, + "language_loss": 0.78094882, + "learning_rate": 5.4724993415760533e-08, + "loss": 0.80303562, + "num_input_tokens_seen": 166761115, + "step": 7714, + "time_per_iteration": 2.696192502975464 + }, + { + "auxiliary_loss_clip": 0.01199742, + "auxiliary_loss_mlp": 0.01378585, + "balance_loss_clip": 1.04154599, + "balance_loss_mlp": 1.00017643, + "epoch": 0.9276739012805868, + "flos": 18696998885760.0, + "grad_norm": 2.457053331108182, + "language_loss": 0.74746764, + "learning_rate": 5.454416406457496e-08, + "loss": 0.77325094, + "num_input_tokens_seen": 166780210, + "step": 7715, + "time_per_iteration": 2.657360315322876 + }, + { + "auxiliary_loss_clip": 0.01173254, + "auxiliary_loss_mlp": 0.01021987, + "balance_loss_clip": 1.04342782, + "balance_loss_mlp": 1.01521611, + "epoch": 0.9277941441712259, + "flos": 13879079740800.0, + "grad_norm": 2.750235371518242, + "language_loss": 0.73590064, + "learning_rate": 5.436362983749299e-08, + "loss": 0.75785309, + "num_input_tokens_seen": 166795380, + "step": 7716, + "time_per_iteration": 2.548280715942383 + }, + { + "auxiliary_loss_clip": 0.01188851, + "auxiliary_loss_mlp": 0.01024633, + "balance_loss_clip": 1.04580522, + "balance_loss_mlp": 1.01823413, + "epoch": 0.927914387061865, + "flos": 23258659426560.0, + "grad_norm": 1.9356208156641352, + "language_loss": 0.64109504, + "learning_rate": 5.418339076190137e-08, + "loss": 0.66322988, + "num_input_tokens_seen": 166814890, + "step": 7717, + "time_per_iteration": 2.666433095932007 + }, + { + "auxiliary_loss_clip": 0.01175568, + "auxiliary_loss_mlp": 0.01028032, + "balance_loss_clip": 1.04384232, + "balance_loss_mlp": 1.02077866, + "epoch": 0.9280346299525041, + "flos": 18073733068800.0, + "grad_norm": 1.8397622367512985, + "language_loss": 0.88491297, + "learning_rate": 5.400344686514202e-08, + "loss": 0.90694898, + "num_input_tokens_seen": 166832475, + "step": 7718, + "time_per_iteration": 2.549764394760132 + }, + { + "auxiliary_loss_clip": 0.01177233, + "auxiliary_loss_mlp": 0.0102652, + "balance_loss_clip": 1.04711103, + "balance_loss_mlp": 1.019418, + "epoch": 0.9281548728431431, + "flos": 22342901160960.0, + "grad_norm": 2.041997555516203, + "language_loss": 0.66457242, + "learning_rate": 5.38237981745131e-08, + "loss": 0.68660998, + "num_input_tokens_seen": 166850590, + "step": 7719, + "time_per_iteration": 2.6074352264404297 + }, + { + "auxiliary_loss_clip": 0.0117804, + "auxiliary_loss_mlp": 0.01378313, + "balance_loss_clip": 1.04553151, + "balance_loss_mlp": 1.00014997, + "epoch": 0.9282751157337822, + "flos": 18843765857280.0, + "grad_norm": 1.6243687683351775, + "language_loss": 0.81288177, + "learning_rate": 5.364444471726592e-08, + "loss": 0.83844531, + "num_input_tokens_seen": 166869795, + "step": 7720, + "time_per_iteration": 2.577554941177368 + }, + { + "auxiliary_loss_clip": 0.0117442, + "auxiliary_loss_mlp": 0.01022911, + "balance_loss_clip": 1.04421067, + "balance_loss_mlp": 1.01584744, + "epoch": 0.9283953586244214, + "flos": 25556834476800.0, + "grad_norm": 2.146061040690408, + "language_loss": 0.80006468, + "learning_rate": 5.346538652060939e-08, + "loss": 0.82203805, + "num_input_tokens_seen": 166891150, + "step": 7721, + "time_per_iteration": 3.5751163959503174 + }, + { + "auxiliary_loss_clip": 0.01183206, + "auxiliary_loss_mlp": 0.01027357, + "balance_loss_clip": 1.04684639, + "balance_loss_mlp": 1.01993585, + "epoch": 0.9285156015150604, + "flos": 18223480869120.0, + "grad_norm": 1.89829398468031, + "language_loss": 0.70112908, + "learning_rate": 5.3286623611705994e-08, + "loss": 0.72323465, + "num_input_tokens_seen": 166909195, + "step": 7722, + "time_per_iteration": 2.71737003326416 + }, + { + "auxiliary_loss_clip": 0.01059937, + "auxiliary_loss_mlp": 0.01000612, + "balance_loss_clip": 1.00891638, + "balance_loss_mlp": 0.99955696, + "epoch": 0.9286358444056995, + "flos": 66400017690240.0, + "grad_norm": 0.8189718829122191, + "language_loss": 0.60640383, + "learning_rate": 5.3108156017673824e-08, + "loss": 0.62700939, + "num_input_tokens_seen": 166970955, + "step": 7723, + "time_per_iteration": 3.2331368923187256 + }, + { + "auxiliary_loss_clip": 0.01192123, + "auxiliary_loss_mlp": 0.01022939, + "balance_loss_clip": 1.0456692, + "balance_loss_mlp": 1.01572704, + "epoch": 0.9287560872963386, + "flos": 22345630594560.0, + "grad_norm": 1.6230210666836087, + "language_loss": 0.71570414, + "learning_rate": 5.2929983765586775e-08, + "loss": 0.73785472, + "num_input_tokens_seen": 166989735, + "step": 7724, + "time_per_iteration": 2.6482787132263184 + }, + { + "auxiliary_loss_clip": 0.0116639, + "auxiliary_loss_mlp": 0.01021051, + "balance_loss_clip": 1.04743791, + "balance_loss_mlp": 1.01376772, + "epoch": 0.9288763301869777, + "flos": 25700225569920.0, + "grad_norm": 2.5993830622088088, + "language_loss": 0.62579197, + "learning_rate": 5.275210688247278e-08, + "loss": 0.64766639, + "num_input_tokens_seen": 167010060, + "step": 7725, + "time_per_iteration": 3.5197906494140625 + }, + { + "auxiliary_loss_clip": 0.01203982, + "auxiliary_loss_mlp": 0.01023052, + "balance_loss_clip": 1.04396868, + "balance_loss_mlp": 1.0161382, + "epoch": 0.9289965730776167, + "flos": 12312046028160.0, + "grad_norm": 2.472917254948725, + "language_loss": 0.85364187, + "learning_rate": 5.257452539531604e-08, + "loss": 0.87591219, + "num_input_tokens_seen": 167027130, + "step": 7726, + "time_per_iteration": 2.6786510944366455 + }, + { + "auxiliary_loss_clip": 0.01175622, + "auxiliary_loss_mlp": 0.01019758, + "balance_loss_clip": 1.04399002, + "balance_loss_mlp": 1.01286733, + "epoch": 0.9291168159682559, + "flos": 26685973486080.0, + "grad_norm": 1.4978134361758402, + "language_loss": 0.68473351, + "learning_rate": 5.2397239331055445e-08, + "loss": 0.70668727, + "num_input_tokens_seen": 167049130, + "step": 7727, + "time_per_iteration": 3.6154468059539795 + }, + { + "auxiliary_loss_clip": 0.0118069, + "auxiliary_loss_mlp": 0.01022321, + "balance_loss_clip": 1.04514956, + "balance_loss_mlp": 1.01463807, + "epoch": 0.929237058858895, + "flos": 14538256179840.0, + "grad_norm": 2.117684270166136, + "language_loss": 0.8118369, + "learning_rate": 5.2220248716585036e-08, + "loss": 0.83386701, + "num_input_tokens_seen": 167066810, + "step": 7728, + "time_per_iteration": 2.615736722946167 + }, + { + "auxiliary_loss_clip": 0.01164497, + "auxiliary_loss_mlp": 0.01031346, + "balance_loss_clip": 1.04207587, + "balance_loss_mlp": 1.02405941, + "epoch": 0.929357301749534, + "flos": 23835456023040.0, + "grad_norm": 2.265199114294253, + "language_loss": 0.75432837, + "learning_rate": 5.204355357875445e-08, + "loss": 0.77628678, + "num_input_tokens_seen": 167085155, + "step": 7729, + "time_per_iteration": 2.6278128623962402 + }, + { + "auxiliary_loss_clip": 0.01181154, + "auxiliary_loss_mlp": 0.01023739, + "balance_loss_clip": 1.04321337, + "balance_loss_mlp": 1.0168488, + "epoch": 0.9294775446401732, + "flos": 12969319046400.0, + "grad_norm": 2.640820518868172, + "language_loss": 0.70550931, + "learning_rate": 5.1867153944367584e-08, + "loss": 0.72755826, + "num_input_tokens_seen": 167101545, + "step": 7730, + "time_per_iteration": 2.5854170322418213 + }, + { + "auxiliary_loss_clip": 0.01197227, + "auxiliary_loss_mlp": 0.01023461, + "balance_loss_clip": 1.04284739, + "balance_loss_mlp": 1.01635051, + "epoch": 0.9295977875308122, + "flos": 26211809024640.0, + "grad_norm": 1.5520177258686938, + "language_loss": 0.73609471, + "learning_rate": 5.16910498401848e-08, + "loss": 0.75830162, + "num_input_tokens_seen": 167120995, + "step": 7731, + "time_per_iteration": 2.7541136741638184 + }, + { + "auxiliary_loss_clip": 0.01165643, + "auxiliary_loss_mlp": 0.01027176, + "balance_loss_clip": 1.0476737, + "balance_loss_mlp": 1.02025318, + "epoch": 0.9297180304214513, + "flos": 16472297105280.0, + "grad_norm": 1.9089810534128586, + "language_loss": 0.83322948, + "learning_rate": 5.151524129292073e-08, + "loss": 0.85515767, + "num_input_tokens_seen": 167138890, + "step": 7732, + "time_per_iteration": 3.373880624771118 + }, + { + "auxiliary_loss_clip": 0.01176955, + "auxiliary_loss_mlp": 0.0102675, + "balance_loss_clip": 1.04689503, + "balance_loss_mlp": 1.019678, + "epoch": 0.9298382733120905, + "flos": 24060436859520.0, + "grad_norm": 1.9580543804342088, + "language_loss": 0.66307104, + "learning_rate": 5.1339728329245155e-08, + "loss": 0.68510807, + "num_input_tokens_seen": 167159455, + "step": 7733, + "time_per_iteration": 2.6227240562438965 + }, + { + "auxiliary_loss_clip": 0.01169265, + "auxiliary_loss_mlp": 0.01025284, + "balance_loss_clip": 1.04612303, + "balance_loss_mlp": 1.01818538, + "epoch": 0.9299585162027295, + "flos": 22127652910080.0, + "grad_norm": 2.092131103859893, + "language_loss": 0.7948631, + "learning_rate": 5.116451097578367e-08, + "loss": 0.81680858, + "num_input_tokens_seen": 167178495, + "step": 7734, + "time_per_iteration": 2.567070245742798 + }, + { + "auxiliary_loss_clip": 0.01189758, + "auxiliary_loss_mlp": 0.0102067, + "balance_loss_clip": 1.04116714, + "balance_loss_mlp": 1.01394367, + "epoch": 0.9300787590933686, + "flos": 21471780522240.0, + "grad_norm": 1.6554260859612695, + "language_loss": 0.74299914, + "learning_rate": 5.0989589259115895e-08, + "loss": 0.76510346, + "num_input_tokens_seen": 167199380, + "step": 7735, + "time_per_iteration": 2.7052767276763916 + }, + { + "auxiliary_loss_clip": 0.01172927, + "auxiliary_loss_mlp": 0.0102409, + "balance_loss_clip": 1.04358327, + "balance_loss_mlp": 1.01616263, + "epoch": 0.9301990019840077, + "flos": 17779588594560.0, + "grad_norm": 1.8647468479127345, + "language_loss": 0.71428424, + "learning_rate": 5.081496320577816e-08, + "loss": 0.73625439, + "num_input_tokens_seen": 167216500, + "step": 7736, + "time_per_iteration": 2.596349000930786 + }, + { + "auxiliary_loss_clip": 0.0108876, + "auxiliary_loss_mlp": 0.01000137, + "balance_loss_clip": 1.01572633, + "balance_loss_mlp": 0.99912351, + "epoch": 0.9303192448746468, + "flos": 58896122307840.0, + "grad_norm": 0.9176887874911803, + "language_loss": 0.61244059, + "learning_rate": 5.0640632842260835e-08, + "loss": 0.63332963, + "num_input_tokens_seen": 167276760, + "step": 7737, + "time_per_iteration": 3.3437211513519287 + }, + { + "auxiliary_loss_clip": 0.01193591, + "auxiliary_loss_mlp": 0.01378472, + "balance_loss_clip": 1.04624796, + "balance_loss_mlp": 1.00013399, + "epoch": 0.9304394877652858, + "flos": 57663522172800.0, + "grad_norm": 1.4349720565474386, + "language_loss": 0.72426462, + "learning_rate": 5.0466598195009426e-08, + "loss": 0.74998522, + "num_input_tokens_seen": 167303630, + "step": 7738, + "time_per_iteration": 3.0098605155944824 + }, + { + "auxiliary_loss_clip": 0.01196113, + "auxiliary_loss_mlp": 0.01025815, + "balance_loss_clip": 1.04414713, + "balance_loss_mlp": 1.01892757, + "epoch": 0.930559730655925, + "flos": 20996143603200.0, + "grad_norm": 1.8462735862302495, + "language_loss": 0.70206422, + "learning_rate": 5.0292859290425036e-08, + "loss": 0.72428346, + "num_input_tokens_seen": 167321500, + "step": 7739, + "time_per_iteration": 2.653343677520752 + }, + { + "auxiliary_loss_clip": 0.01166511, + "auxiliary_loss_mlp": 0.01023883, + "balance_loss_clip": 1.04755366, + "balance_loss_mlp": 1.0174576, + "epoch": 0.9306799735465641, + "flos": 23258264376960.0, + "grad_norm": 1.8499084502199332, + "language_loss": 0.77676666, + "learning_rate": 5.011941615486348e-08, + "loss": 0.79867065, + "num_input_tokens_seen": 167340615, + "step": 7740, + "time_per_iteration": 2.5917210578918457 + }, + { + "auxiliary_loss_clip": 0.01167586, + "auxiliary_loss_mlp": 0.01026369, + "balance_loss_clip": 1.04756975, + "balance_loss_mlp": 1.01939249, + "epoch": 0.9308002164372031, + "flos": 15231547560960.0, + "grad_norm": 1.9558434051784355, + "language_loss": 0.84475476, + "learning_rate": 4.994626881463659e-08, + "loss": 0.86669433, + "num_input_tokens_seen": 167356870, + "step": 7741, + "time_per_iteration": 2.515688419342041 + }, + { + "auxiliary_loss_clip": 0.01185572, + "auxiliary_loss_mlp": 0.01025189, + "balance_loss_clip": 1.03931773, + "balance_loss_mlp": 1.01791739, + "epoch": 0.9309204593278423, + "flos": 30847481539200.0, + "grad_norm": 1.7152072772996674, + "language_loss": 0.70838916, + "learning_rate": 4.9773417296009814e-08, + "loss": 0.73049676, + "num_input_tokens_seen": 167378390, + "step": 7742, + "time_per_iteration": 2.7832376956939697 + }, + { + "auxiliary_loss_clip": 0.01180577, + "auxiliary_loss_mlp": 0.01022379, + "balance_loss_clip": 1.04760754, + "balance_loss_mlp": 1.01450515, + "epoch": 0.9310407022184813, + "flos": 23037269950080.0, + "grad_norm": 1.8789217010907049, + "language_loss": 0.65254378, + "learning_rate": 4.960086162520527e-08, + "loss": 0.67457336, + "num_input_tokens_seen": 167398480, + "step": 7743, + "time_per_iteration": 2.63679575920105 + }, + { + "auxiliary_loss_clip": 0.012108, + "auxiliary_loss_mlp": 0.01020197, + "balance_loss_clip": 1.04426897, + "balance_loss_mlp": 1.01334584, + "epoch": 0.9311609451091204, + "flos": 22127976132480.0, + "grad_norm": 1.9059094215605652, + "language_loss": 0.82385671, + "learning_rate": 4.942860182839936e-08, + "loss": 0.84616673, + "num_input_tokens_seen": 167416825, + "step": 7744, + "time_per_iteration": 2.76179838180542 + }, + { + "auxiliary_loss_clip": 0.01182315, + "auxiliary_loss_mlp": 0.01029998, + "balance_loss_clip": 1.0437932, + "balance_loss_mlp": 1.02238083, + "epoch": 0.9312811879997596, + "flos": 21099206701440.0, + "grad_norm": 2.233657312251956, + "language_loss": 0.79680371, + "learning_rate": 4.925663793172341e-08, + "loss": 0.81892681, + "num_input_tokens_seen": 167434785, + "step": 7745, + "time_per_iteration": 2.644158124923706 + }, + { + "auxiliary_loss_clip": 0.01082569, + "auxiliary_loss_mlp": 0.01373673, + "balance_loss_clip": 1.01019502, + "balance_loss_mlp": 0.99982339, + "epoch": 0.9314014308903986, + "flos": 67148179096320.0, + "grad_norm": 0.7840882582309235, + "language_loss": 0.56510955, + "learning_rate": 4.908496996126477e-08, + "loss": 0.58967197, + "num_input_tokens_seen": 167498245, + "step": 7746, + "time_per_iteration": 3.282951831817627 + }, + { + "auxiliary_loss_clip": 0.01176988, + "auxiliary_loss_mlp": 0.01026594, + "balance_loss_clip": 1.04982412, + "balance_loss_mlp": 1.01893771, + "epoch": 0.9315216737810377, + "flos": 22565583527040.0, + "grad_norm": 1.468563504632021, + "language_loss": 0.76166737, + "learning_rate": 4.89135979430646e-08, + "loss": 0.78370321, + "num_input_tokens_seen": 167518290, + "step": 7747, + "time_per_iteration": 3.56777286529541 + }, + { + "auxiliary_loss_clip": 0.01167948, + "auxiliary_loss_mlp": 0.01024882, + "balance_loss_clip": 1.04844379, + "balance_loss_mlp": 1.01720762, + "epoch": 0.9316419166716768, + "flos": 23984054588160.0, + "grad_norm": 1.7113281867237826, + "language_loss": 0.85404062, + "learning_rate": 4.874252190312078e-08, + "loss": 0.87596893, + "num_input_tokens_seen": 167538675, + "step": 7748, + "time_per_iteration": 2.5617198944091797 + }, + { + "auxiliary_loss_clip": 0.01178563, + "auxiliary_loss_mlp": 0.01031094, + "balance_loss_clip": 1.04521906, + "balance_loss_mlp": 1.02277362, + "epoch": 0.9317621595623159, + "flos": 30230464688640.0, + "grad_norm": 2.017901559023009, + "language_loss": 0.65061224, + "learning_rate": 4.857174186738477e-08, + "loss": 0.67270881, + "num_input_tokens_seen": 167562025, + "step": 7749, + "time_per_iteration": 2.6953608989715576 + }, + { + "auxiliary_loss_clip": 0.01169641, + "auxiliary_loss_mlp": 0.01026909, + "balance_loss_clip": 1.04977536, + "balance_loss_mlp": 1.01980972, + "epoch": 0.931882402452955, + "flos": 15742735966080.0, + "grad_norm": 2.1307267185636163, + "language_loss": 0.7313441, + "learning_rate": 4.840125786176408e-08, + "loss": 0.75330961, + "num_input_tokens_seen": 167578230, + "step": 7750, + "time_per_iteration": 2.5581443309783936 + }, + { + "auxiliary_loss_clip": 0.0117942, + "auxiliary_loss_mlp": 0.01019899, + "balance_loss_clip": 1.04279447, + "balance_loss_mlp": 1.01299691, + "epoch": 0.932002645343594, + "flos": 28366521154560.0, + "grad_norm": 1.7582790267526067, + "language_loss": 0.77526677, + "learning_rate": 4.823106991212067e-08, + "loss": 0.79725993, + "num_input_tokens_seen": 167597470, + "step": 7751, + "time_per_iteration": 3.584836721420288 + }, + { + "auxiliary_loss_clip": 0.01178001, + "auxiliary_loss_mlp": 0.01024983, + "balance_loss_clip": 1.04544735, + "balance_loss_mlp": 1.0183996, + "epoch": 0.9321228882342332, + "flos": 15341146934400.0, + "grad_norm": 1.8997253742973028, + "language_loss": 0.82900161, + "learning_rate": 4.806117804427212e-08, + "loss": 0.85103142, + "num_input_tokens_seen": 167615405, + "step": 7752, + "time_per_iteration": 2.6271846294403076 + }, + { + "auxiliary_loss_clip": 0.01169921, + "auxiliary_loss_mlp": 0.01023198, + "balance_loss_clip": 1.04414749, + "balance_loss_mlp": 1.01583147, + "epoch": 0.9322431311248722, + "flos": 17895365107200.0, + "grad_norm": 2.248324433480775, + "language_loss": 0.64405727, + "learning_rate": 4.7891582283990926e-08, + "loss": 0.66598856, + "num_input_tokens_seen": 167634130, + "step": 7753, + "time_per_iteration": 3.515371799468994 + }, + { + "auxiliary_loss_clip": 0.0119195, + "auxiliary_loss_mlp": 0.0103036, + "balance_loss_clip": 1.04210877, + "balance_loss_mlp": 1.02331471, + "epoch": 0.9323633740155113, + "flos": 24169713010560.0, + "grad_norm": 1.722748482018521, + "language_loss": 0.72668099, + "learning_rate": 4.772228265700473e-08, + "loss": 0.74890411, + "num_input_tokens_seen": 167654990, + "step": 7754, + "time_per_iteration": 2.7152297496795654 + }, + { + "auxiliary_loss_clip": 0.0117826, + "auxiliary_loss_mlp": 0.01028155, + "balance_loss_clip": 1.04541707, + "balance_loss_mlp": 1.02021265, + "epoch": 0.9324836169061504, + "flos": 15043482927360.0, + "grad_norm": 2.0626155377399806, + "language_loss": 0.75859207, + "learning_rate": 4.75532791889961e-08, + "loss": 0.78065622, + "num_input_tokens_seen": 167671690, + "step": 7755, + "time_per_iteration": 2.6027402877807617 + }, + { + "auxiliary_loss_clip": 0.01172179, + "auxiliary_loss_mlp": 0.01025215, + "balance_loss_clip": 1.04311919, + "balance_loss_mlp": 1.0179373, + "epoch": 0.9326038597967895, + "flos": 18624890332800.0, + "grad_norm": 1.957239479196228, + "language_loss": 0.65767437, + "learning_rate": 4.738457190560252e-08, + "loss": 0.67964828, + "num_input_tokens_seen": 167690800, + "step": 7756, + "time_per_iteration": 2.6303696632385254 + }, + { + "auxiliary_loss_clip": 0.01200887, + "auxiliary_loss_mlp": 0.0102419, + "balance_loss_clip": 1.04351163, + "balance_loss_mlp": 1.01746941, + "epoch": 0.9327241026874286, + "flos": 18952646958720.0, + "grad_norm": 2.155936107569324, + "language_loss": 0.78889161, + "learning_rate": 4.721616083241664e-08, + "loss": 0.81114239, + "num_input_tokens_seen": 167709055, + "step": 7757, + "time_per_iteration": 2.701907157897949 + }, + { + "auxiliary_loss_clip": 0.01170686, + "auxiliary_loss_mlp": 0.01029836, + "balance_loss_clip": 1.04523599, + "balance_loss_mlp": 1.02229023, + "epoch": 0.9328443455780677, + "flos": 29570282668800.0, + "grad_norm": 1.724588796472838, + "language_loss": 0.77740788, + "learning_rate": 4.7048045994986684e-08, + "loss": 0.79941314, + "num_input_tokens_seen": 167729915, + "step": 7758, + "time_per_iteration": 3.5772109031677246 + }, + { + "auxiliary_loss_clip": 0.01182178, + "auxiliary_loss_mlp": 0.01023906, + "balance_loss_clip": 1.04733968, + "balance_loss_mlp": 1.01674736, + "epoch": 0.9329645884687068, + "flos": 30081722469120.0, + "grad_norm": 2.2406519839477914, + "language_loss": 0.91125745, + "learning_rate": 4.688022741881559e-08, + "loss": 0.93331826, + "num_input_tokens_seen": 167750440, + "step": 7759, + "time_per_iteration": 2.689298391342163 + }, + { + "auxiliary_loss_clip": 0.01171791, + "auxiliary_loss_mlp": 0.01023955, + "balance_loss_clip": 1.04465151, + "balance_loss_mlp": 1.01738656, + "epoch": 0.9330848313593458, + "flos": 21867982513920.0, + "grad_norm": 1.8332045752816242, + "language_loss": 0.75074673, + "learning_rate": 4.671270512936076e-08, + "loss": 0.77270424, + "num_input_tokens_seen": 167769600, + "step": 7760, + "time_per_iteration": 2.6059141159057617 + }, + { + "auxiliary_loss_clip": 0.01185651, + "auxiliary_loss_mlp": 0.01029169, + "balance_loss_clip": 1.04129243, + "balance_loss_mlp": 1.02243662, + "epoch": 0.933205074249985, + "flos": 22127221946880.0, + "grad_norm": 2.7882332313149867, + "language_loss": 0.82990772, + "learning_rate": 4.6545479152035884e-08, + "loss": 0.85205591, + "num_input_tokens_seen": 167788770, + "step": 7761, + "time_per_iteration": 2.696234703063965 + }, + { + "auxiliary_loss_clip": 0.01180029, + "auxiliary_loss_mlp": 0.01023885, + "balance_loss_clip": 1.04746509, + "balance_loss_mlp": 1.01703644, + "epoch": 0.9333253171406241, + "flos": 15341254675200.0, + "grad_norm": 2.1005158519710334, + "language_loss": 0.76059246, + "learning_rate": 4.637854951220821e-08, + "loss": 0.78263152, + "num_input_tokens_seen": 167805555, + "step": 7762, + "time_per_iteration": 2.5514473915100098 + }, + { + "auxiliary_loss_clip": 0.01187328, + "auxiliary_loss_mlp": 0.01024664, + "balance_loss_clip": 1.04209185, + "balance_loss_mlp": 1.01721978, + "epoch": 0.9334455600312631, + "flos": 15706142985600.0, + "grad_norm": 2.035762143329696, + "language_loss": 0.75210375, + "learning_rate": 4.621191623520171e-08, + "loss": 0.77422369, + "num_input_tokens_seen": 167823985, + "step": 7763, + "time_per_iteration": 2.754770278930664 + }, + { + "auxiliary_loss_clip": 0.0122602, + "auxiliary_loss_mlp": 0.01024596, + "balance_loss_clip": 1.04255939, + "balance_loss_mlp": 1.01732457, + "epoch": 0.9335658029219023, + "flos": 22163563532160.0, + "grad_norm": 2.5412907397197118, + "language_loss": 0.84432667, + "learning_rate": 4.604557934629372e-08, + "loss": 0.86683285, + "num_input_tokens_seen": 167843060, + "step": 7764, + "time_per_iteration": 2.7558391094207764 + }, + { + "auxiliary_loss_clip": 0.01178692, + "auxiliary_loss_mlp": 0.01028462, + "balance_loss_clip": 1.04382932, + "balance_loss_mlp": 1.0213989, + "epoch": 0.9336860458125413, + "flos": 20266833859200.0, + "grad_norm": 1.912224865488971, + "language_loss": 0.80305141, + "learning_rate": 4.587953887071805e-08, + "loss": 0.82512295, + "num_input_tokens_seen": 167862880, + "step": 7765, + "time_per_iteration": 2.674241542816162 + }, + { + "auxiliary_loss_clip": 0.01177367, + "auxiliary_loss_mlp": 0.01027496, + "balance_loss_clip": 1.04110897, + "balance_loss_mlp": 1.02040279, + "epoch": 0.9338062887031804, + "flos": 20919689504640.0, + "grad_norm": 1.7912256634861887, + "language_loss": 0.85710418, + "learning_rate": 4.5713794833662554e-08, + "loss": 0.87915283, + "num_input_tokens_seen": 167882095, + "step": 7766, + "time_per_iteration": 2.662278652191162 + }, + { + "auxiliary_loss_clip": 0.01169111, + "auxiliary_loss_mlp": 0.01025077, + "balance_loss_clip": 1.04730678, + "balance_loss_mlp": 1.0173161, + "epoch": 0.9339265315938196, + "flos": 23221635482880.0, + "grad_norm": 1.8544237184575554, + "language_loss": 0.63127422, + "learning_rate": 4.5548347260270236e-08, + "loss": 0.65321612, + "num_input_tokens_seen": 167901385, + "step": 7767, + "time_per_iteration": 2.63283109664917 + }, + { + "auxiliary_loss_clip": 0.01191579, + "auxiliary_loss_mlp": 0.01025307, + "balance_loss_clip": 1.04386151, + "balance_loss_mlp": 1.01844049, + "epoch": 0.9340467744844586, + "flos": 22820261932800.0, + "grad_norm": 1.8196192161742994, + "language_loss": 0.69270957, + "learning_rate": 4.538319617564012e-08, + "loss": 0.71487844, + "num_input_tokens_seen": 167920405, + "step": 7768, + "time_per_iteration": 2.7043559551239014 + }, + { + "auxiliary_loss_clip": 0.0118162, + "auxiliary_loss_mlp": 0.01027346, + "balance_loss_clip": 1.04283845, + "balance_loss_mlp": 1.01981831, + "epoch": 0.9341670173750977, + "flos": 23660428026240.0, + "grad_norm": 2.043851074557788, + "language_loss": 0.74577248, + "learning_rate": 4.521834160482485e-08, + "loss": 0.76786208, + "num_input_tokens_seen": 167939145, + "step": 7769, + "time_per_iteration": 2.664968967437744 + }, + { + "auxiliary_loss_clip": 0.01179537, + "auxiliary_loss_mlp": 0.01025607, + "balance_loss_clip": 1.04688728, + "balance_loss_mlp": 1.01854134, + "epoch": 0.9342872602657368, + "flos": 24824256595200.0, + "grad_norm": 1.6125694887665936, + "language_loss": 0.81700349, + "learning_rate": 4.5053783572832846e-08, + "loss": 0.83905488, + "num_input_tokens_seen": 167959325, + "step": 7770, + "time_per_iteration": 2.6430249214172363 + }, + { + "auxiliary_loss_clip": 0.0117459, + "auxiliary_loss_mlp": 0.0102794, + "balance_loss_clip": 1.04605544, + "balance_loss_mlp": 1.02058172, + "epoch": 0.9344075031563759, + "flos": 25771831332480.0, + "grad_norm": 1.6418389329162506, + "language_loss": 0.76341903, + "learning_rate": 4.488952210462771e-08, + "loss": 0.78544426, + "num_input_tokens_seen": 167979530, + "step": 7771, + "time_per_iteration": 2.6204662322998047 + }, + { + "auxiliary_loss_clip": 0.01167086, + "auxiliary_loss_mlp": 0.01028206, + "balance_loss_clip": 1.04795325, + "balance_loss_mlp": 1.02122068, + "epoch": 0.9345277460470149, + "flos": 25551303782400.0, + "grad_norm": 1.8090590636152735, + "language_loss": 0.85611975, + "learning_rate": 4.4725557225127495e-08, + "loss": 0.87807262, + "num_input_tokens_seen": 167997870, + "step": 7772, + "time_per_iteration": 2.62441349029541 + }, + { + "auxiliary_loss_clip": 0.01178578, + "auxiliary_loss_mlp": 0.01025011, + "balance_loss_clip": 1.04666483, + "balance_loss_mlp": 1.01828754, + "epoch": 0.9346479889376541, + "flos": 34313112432000.0, + "grad_norm": 1.5798412179172765, + "language_loss": 0.79463351, + "learning_rate": 4.456188895920565e-08, + "loss": 0.81666946, + "num_input_tokens_seen": 168019625, + "step": 7773, + "time_per_iteration": 3.5899994373321533 + }, + { + "auxiliary_loss_clip": 0.01168771, + "auxiliary_loss_mlp": 0.01023283, + "balance_loss_clip": 1.04771566, + "balance_loss_mlp": 1.0158143, + "epoch": 0.9347682318282932, + "flos": 19093739581440.0, + "grad_norm": 2.052552415056983, + "language_loss": 0.85510504, + "learning_rate": 4.439851733169031e-08, + "loss": 0.8770256, + "num_input_tokens_seen": 168037415, + "step": 7774, + "time_per_iteration": 2.5415220260620117 + }, + { + "auxiliary_loss_clip": 0.01192723, + "auxiliary_loss_mlp": 0.0102383, + "balance_loss_clip": 1.04217196, + "balance_loss_mlp": 1.0169313, + "epoch": 0.9348884747189322, + "flos": 26249587153920.0, + "grad_norm": 3.3266690447433214, + "language_loss": 0.69148797, + "learning_rate": 4.4235442367365204e-08, + "loss": 0.7136535, + "num_input_tokens_seen": 168057725, + "step": 7775, + "time_per_iteration": 2.7863287925720215 + }, + { + "auxiliary_loss_clip": 0.01176664, + "auxiliary_loss_mlp": 0.01025345, + "balance_loss_clip": 1.0407002, + "balance_loss_mlp": 1.01813841, + "epoch": 0.9350087176095714, + "flos": 18333080242560.0, + "grad_norm": 1.933404453800415, + "language_loss": 0.79255641, + "learning_rate": 4.4072664090968545e-08, + "loss": 0.81457651, + "num_input_tokens_seen": 168076110, + "step": 7776, + "time_per_iteration": 2.635148763656616 + }, + { + "auxiliary_loss_clip": 0.01182578, + "auxiliary_loss_mlp": 0.010234, + "balance_loss_clip": 1.04124999, + "balance_loss_mlp": 1.01623869, + "epoch": 0.9351289605002104, + "flos": 19318253541120.0, + "grad_norm": 1.7796379282127908, + "language_loss": 0.84687281, + "learning_rate": 4.391018252719347e-08, + "loss": 0.8689326, + "num_input_tokens_seen": 168095905, + "step": 7777, + "time_per_iteration": 3.5820493698120117 + }, + { + "auxiliary_loss_clip": 0.01185064, + "auxiliary_loss_mlp": 0.01028756, + "balance_loss_clip": 1.04285288, + "balance_loss_mlp": 1.02141881, + "epoch": 0.9352492033908495, + "flos": 18799990156800.0, + "grad_norm": 1.7479053931319843, + "language_loss": 0.69432288, + "learning_rate": 4.374799770068849e-08, + "loss": 0.71646106, + "num_input_tokens_seen": 168112580, + "step": 7778, + "time_per_iteration": 2.6149957180023193 + }, + { + "auxiliary_loss_clip": 0.01174543, + "auxiliary_loss_mlp": 0.0102512, + "balance_loss_clip": 1.04619122, + "balance_loss_mlp": 1.0173471, + "epoch": 0.9353694462814887, + "flos": 29530134241920.0, + "grad_norm": 1.9961886540456177, + "language_loss": 0.74762487, + "learning_rate": 4.358610963605658e-08, + "loss": 0.76962149, + "num_input_tokens_seen": 168133030, + "step": 7779, + "time_per_iteration": 3.596085548400879 + }, + { + "auxiliary_loss_clip": 0.01170858, + "auxiliary_loss_mlp": 0.01022759, + "balance_loss_clip": 1.0481689, + "balance_loss_mlp": 1.01526356, + "epoch": 0.9354896891721277, + "flos": 30665450390400.0, + "grad_norm": 2.503748262725146, + "language_loss": 0.68884194, + "learning_rate": 4.342451835785677e-08, + "loss": 0.71077812, + "num_input_tokens_seen": 168153940, + "step": 7780, + "time_per_iteration": 2.643629550933838 + }, + { + "auxiliary_loss_clip": 0.01181589, + "auxiliary_loss_mlp": 0.0102391, + "balance_loss_clip": 1.04437435, + "balance_loss_mlp": 1.01672149, + "epoch": 0.9356099320627668, + "flos": 19463907191040.0, + "grad_norm": 1.6934176498445064, + "language_loss": 0.75158226, + "learning_rate": 4.3263223890601665e-08, + "loss": 0.77363724, + "num_input_tokens_seen": 168172650, + "step": 7781, + "time_per_iteration": 2.685864210128784 + }, + { + "auxiliary_loss_clip": 0.01170353, + "auxiliary_loss_mlp": 0.01378357, + "balance_loss_clip": 1.04627514, + "balance_loss_mlp": 1.00021088, + "epoch": 0.9357301749534058, + "flos": 19098156954240.0, + "grad_norm": 1.756255285801717, + "language_loss": 0.7936542, + "learning_rate": 4.31022262587597e-08, + "loss": 0.81914127, + "num_input_tokens_seen": 168191325, + "step": 7782, + "time_per_iteration": 2.6016461849212646 + }, + { + "auxiliary_loss_clip": 0.01177415, + "auxiliary_loss_mlp": 0.01027931, + "balance_loss_clip": 1.04772019, + "balance_loss_mlp": 1.01995003, + "epoch": 0.935850417844045, + "flos": 23550361776000.0, + "grad_norm": 1.7356395571383632, + "language_loss": 0.65937603, + "learning_rate": 4.2941525486754225e-08, + "loss": 0.68142951, + "num_input_tokens_seen": 168211645, + "step": 7783, + "time_per_iteration": 2.598583936691284 + }, + { + "auxiliary_loss_clip": 0.01189381, + "auxiliary_loss_mlp": 0.01023747, + "balance_loss_clip": 1.04408503, + "balance_loss_mlp": 1.01727104, + "epoch": 0.935970660734684, + "flos": 18588333265920.0, + "grad_norm": 1.8535449363642638, + "language_loss": 0.79432744, + "learning_rate": 4.278112159896286e-08, + "loss": 0.81645864, + "num_input_tokens_seen": 168229485, + "step": 7784, + "time_per_iteration": 3.516096830368042 + }, + { + "auxiliary_loss_clip": 0.01175099, + "auxiliary_loss_mlp": 0.01018883, + "balance_loss_clip": 1.04046869, + "balance_loss_mlp": 1.01242435, + "epoch": 0.9360909036253231, + "flos": 20631255292800.0, + "grad_norm": 2.2213790646184157, + "language_loss": 0.67723441, + "learning_rate": 4.2621014619719896e-08, + "loss": 0.69917423, + "num_input_tokens_seen": 168247250, + "step": 7785, + "time_per_iteration": 2.6217610836029053 + }, + { + "auxiliary_loss_clip": 0.01084289, + "auxiliary_loss_mlp": 0.0100297, + "balance_loss_clip": 1.01022339, + "balance_loss_mlp": 1.00195646, + "epoch": 0.9362111465159623, + "flos": 61791421052160.0, + "grad_norm": 0.718696518542313, + "language_loss": 0.58604312, + "learning_rate": 4.246120457331215e-08, + "loss": 0.60691571, + "num_input_tokens_seen": 168309425, + "step": 7786, + "time_per_iteration": 3.2606430053710938 + }, + { + "auxiliary_loss_clip": 0.0118031, + "auxiliary_loss_mlp": 0.01029621, + "balance_loss_clip": 1.04711509, + "balance_loss_mlp": 1.02212811, + "epoch": 0.9363313894066013, + "flos": 24170395368960.0, + "grad_norm": 2.0469867316278303, + "language_loss": 0.71962106, + "learning_rate": 4.2301691483983325e-08, + "loss": 0.74172032, + "num_input_tokens_seen": 168329545, + "step": 7787, + "time_per_iteration": 2.6304144859313965 + }, + { + "auxiliary_loss_clip": 0.01180872, + "auxiliary_loss_mlp": 0.01027071, + "balance_loss_clip": 1.04651237, + "balance_loss_mlp": 1.02063966, + "epoch": 0.9364516322972404, + "flos": 20120354196480.0, + "grad_norm": 2.065555403343176, + "language_loss": 0.75793493, + "learning_rate": 4.214247537593163e-08, + "loss": 0.7800144, + "num_input_tokens_seen": 168348795, + "step": 7788, + "time_per_iteration": 2.588043689727783 + }, + { + "auxiliary_loss_clip": 0.01182717, + "auxiliary_loss_mlp": 0.01026392, + "balance_loss_clip": 1.04212236, + "balance_loss_mlp": 1.01872969, + "epoch": 0.9365718751878795, + "flos": 20703758895360.0, + "grad_norm": 2.0433805395196374, + "language_loss": 0.80592191, + "learning_rate": 4.1983556273309293e-08, + "loss": 0.82801294, + "num_input_tokens_seen": 168367545, + "step": 7789, + "time_per_iteration": 2.6374399662017822 + }, + { + "auxiliary_loss_clip": 0.01170983, + "auxiliary_loss_mlp": 0.0102866, + "balance_loss_clip": 1.04832625, + "balance_loss_mlp": 1.02083993, + "epoch": 0.9366921180785186, + "flos": 18655270260480.0, + "grad_norm": 2.9585011187172685, + "language_loss": 0.6929068, + "learning_rate": 4.182493420022526e-08, + "loss": 0.71490324, + "num_input_tokens_seen": 168383215, + "step": 7790, + "time_per_iteration": 2.5790600776672363 + }, + { + "auxiliary_loss_clip": 0.01194292, + "auxiliary_loss_mlp": 0.01025185, + "balance_loss_clip": 1.04243684, + "balance_loss_mlp": 1.01853609, + "epoch": 0.9368123609691577, + "flos": 25774955815680.0, + "grad_norm": 2.4192796487055226, + "language_loss": 0.78351915, + "learning_rate": 4.166660918074139e-08, + "loss": 0.80571389, + "num_input_tokens_seen": 168403120, + "step": 7791, + "time_per_iteration": 2.6883041858673096 + }, + { + "auxiliary_loss_clip": 0.01188702, + "auxiliary_loss_mlp": 0.01025457, + "balance_loss_clip": 1.04206228, + "balance_loss_mlp": 1.0185008, + "epoch": 0.9369326038597968, + "flos": 25553386771200.0, + "grad_norm": 1.4852892472545582, + "language_loss": 0.73407972, + "learning_rate": 4.15085812388758e-08, + "loss": 0.75622129, + "num_input_tokens_seen": 168425340, + "step": 7792, + "time_per_iteration": 2.68436598777771 + }, + { + "auxiliary_loss_clip": 0.01182977, + "auxiliary_loss_mlp": 0.01027731, + "balance_loss_clip": 1.04460549, + "balance_loss_mlp": 1.02040839, + "epoch": 0.9370528467504359, + "flos": 23220019370880.0, + "grad_norm": 1.6591853853716512, + "language_loss": 0.78666353, + "learning_rate": 4.135085039860153e-08, + "loss": 0.80877066, + "num_input_tokens_seen": 168444740, + "step": 7793, + "time_per_iteration": 2.7614285945892334 + }, + { + "auxiliary_loss_clip": 0.01184021, + "auxiliary_loss_mlp": 0.01026055, + "balance_loss_clip": 1.0479517, + "balance_loss_mlp": 1.01853859, + "epoch": 0.9371730896410749, + "flos": 24967468120320.0, + "grad_norm": 2.0982200927413266, + "language_loss": 0.7824645, + "learning_rate": 4.1193416683845906e-08, + "loss": 0.80456525, + "num_input_tokens_seen": 168463670, + "step": 7794, + "time_per_iteration": 2.6550183296203613 + }, + { + "auxiliary_loss_clip": 0.01195052, + "auxiliary_loss_mlp": 0.01022863, + "balance_loss_clip": 1.04457188, + "balance_loss_mlp": 1.01679802, + "epoch": 0.9372933325317141, + "flos": 15553091134080.0, + "grad_norm": 2.514372932226363, + "language_loss": 0.83519292, + "learning_rate": 4.103628011849136e-08, + "loss": 0.85737211, + "num_input_tokens_seen": 168479030, + "step": 7795, + "time_per_iteration": 2.6553494930267334 + }, + { + "auxiliary_loss_clip": 0.01189402, + "auxiliary_loss_mlp": 0.01024669, + "balance_loss_clip": 1.04644096, + "balance_loss_mlp": 1.01740885, + "epoch": 0.9374135754223532, + "flos": 21871861182720.0, + "grad_norm": 1.9090935711584602, + "language_loss": 0.75868082, + "learning_rate": 4.0879440726375506e-08, + "loss": 0.78082156, + "num_input_tokens_seen": 168496815, + "step": 7796, + "time_per_iteration": 2.616743326187134 + }, + { + "auxiliary_loss_clip": 0.01180511, + "auxiliary_loss_mlp": 0.01023924, + "balance_loss_clip": 1.04185653, + "balance_loss_mlp": 1.01644647, + "epoch": 0.9375338183129922, + "flos": 22631048064000.0, + "grad_norm": 3.412105688637198, + "language_loss": 0.56560236, + "learning_rate": 4.0722898531291074e-08, + "loss": 0.58764672, + "num_input_tokens_seen": 168514055, + "step": 7797, + "time_per_iteration": 2.6676807403564453 + }, + { + "auxiliary_loss_clip": 0.01191194, + "auxiliary_loss_mlp": 0.01024752, + "balance_loss_clip": 1.04455233, + "balance_loss_mlp": 1.01748347, + "epoch": 0.9376540612036314, + "flos": 26104292640000.0, + "grad_norm": 1.7926690708662874, + "language_loss": 0.76458901, + "learning_rate": 4.0566653556985295e-08, + "loss": 0.78674847, + "num_input_tokens_seen": 168534600, + "step": 7798, + "time_per_iteration": 2.708671808242798 + }, + { + "auxiliary_loss_clip": 0.01223435, + "auxiliary_loss_mlp": 0.01027087, + "balance_loss_clip": 1.04014432, + "balance_loss_mlp": 1.01936257, + "epoch": 0.9377743040942704, + "flos": 19717580016000.0, + "grad_norm": 2.1901960514743934, + "language_loss": 0.81847239, + "learning_rate": 4.0410705827159886e-08, + "loss": 0.84097755, + "num_input_tokens_seen": 168551895, + "step": 7799, + "time_per_iteration": 3.6599738597869873 + }, + { + "auxiliary_loss_clip": 0.01176934, + "auxiliary_loss_mlp": 0.01024675, + "balance_loss_clip": 1.04057777, + "balance_loss_mlp": 1.01747513, + "epoch": 0.9378945469849095, + "flos": 15267530010240.0, + "grad_norm": 2.437993749468358, + "language_loss": 0.71011752, + "learning_rate": 4.0255055365472356e-08, + "loss": 0.73213357, + "num_input_tokens_seen": 168569990, + "step": 7800, + "time_per_iteration": 2.622145652770996 + }, + { + "auxiliary_loss_clip": 0.01206048, + "auxiliary_loss_mlp": 0.01024645, + "balance_loss_clip": 1.03548884, + "balance_loss_mlp": 1.01726604, + "epoch": 0.9380147898755486, + "flos": 20591394174720.0, + "grad_norm": 2.0494616224162705, + "language_loss": 0.74655032, + "learning_rate": 4.009970219553471e-08, + "loss": 0.76885724, + "num_input_tokens_seen": 168586940, + "step": 7801, + "time_per_iteration": 2.7894363403320312 + }, + { + "auxiliary_loss_clip": 0.01181326, + "auxiliary_loss_mlp": 0.01023637, + "balance_loss_clip": 1.0466814, + "balance_loss_mlp": 1.01630592, + "epoch": 0.9381350327661877, + "flos": 26281116316800.0, + "grad_norm": 5.691337520229002, + "language_loss": 0.7663275, + "learning_rate": 3.99446463409141e-08, + "loss": 0.78837711, + "num_input_tokens_seen": 168604795, + "step": 7802, + "time_per_iteration": 2.6798899173736572 + }, + { + "auxiliary_loss_clip": 0.01182338, + "auxiliary_loss_mlp": 0.01026568, + "balance_loss_clip": 1.0459857, + "balance_loss_mlp": 1.01904631, + "epoch": 0.9382552756568268, + "flos": 23586344225280.0, + "grad_norm": 2.0979927557256204, + "language_loss": 0.6888299, + "learning_rate": 3.978988782513215e-08, + "loss": 0.71091902, + "num_input_tokens_seen": 168622290, + "step": 7803, + "time_per_iteration": 3.5201451778411865 + }, + { + "auxiliary_loss_clip": 0.01180348, + "auxiliary_loss_mlp": 0.01026866, + "balance_loss_clip": 1.04643679, + "balance_loss_mlp": 1.01937938, + "epoch": 0.9383755185474659, + "flos": 28438809275520.0, + "grad_norm": 1.7636392681828357, + "language_loss": 0.764525, + "learning_rate": 3.963542667166586e-08, + "loss": 0.78659713, + "num_input_tokens_seen": 168642395, + "step": 7804, + "time_per_iteration": 2.6585140228271484 + }, + { + "auxiliary_loss_clip": 0.01196066, + "auxiliary_loss_mlp": 0.01025315, + "balance_loss_clip": 1.04602981, + "balance_loss_mlp": 1.01795423, + "epoch": 0.938495761438105, + "flos": 20449583280000.0, + "grad_norm": 1.7714183347062267, + "language_loss": 0.68489754, + "learning_rate": 3.9481262903946486e-08, + "loss": 0.70711136, + "num_input_tokens_seen": 168661840, + "step": 7805, + "time_per_iteration": 3.6184566020965576 + }, + { + "auxiliary_loss_clip": 0.01113689, + "auxiliary_loss_mlp": 0.01001401, + "balance_loss_clip": 1.00992572, + "balance_loss_mlp": 1.00022638, + "epoch": 0.938616004328744, + "flos": 69302711658240.0, + "grad_norm": 0.7904284562757873, + "language_loss": 0.5453226, + "learning_rate": 3.932739654536066e-08, + "loss": 0.56647348, + "num_input_tokens_seen": 168724540, + "step": 7806, + "time_per_iteration": 3.2290704250335693 + }, + { + "auxiliary_loss_clip": 0.01177407, + "auxiliary_loss_mlp": 0.0102307, + "balance_loss_clip": 1.04759598, + "balance_loss_mlp": 1.0158875, + "epoch": 0.9387362472193832, + "flos": 18911636605440.0, + "grad_norm": 2.0545586211299502, + "language_loss": 0.74202752, + "learning_rate": 3.917382761925014e-08, + "loss": 0.7640323, + "num_input_tokens_seen": 168740375, + "step": 7807, + "time_per_iteration": 2.5922107696533203 + }, + { + "auxiliary_loss_clip": 0.01170905, + "auxiliary_loss_mlp": 0.01027301, + "balance_loss_clip": 1.04475653, + "balance_loss_mlp": 1.0201695, + "epoch": 0.9388564901100223, + "flos": 26501967089280.0, + "grad_norm": 1.7114516354311031, + "language_loss": 0.79089034, + "learning_rate": 3.9020556148910754e-08, + "loss": 0.81287241, + "num_input_tokens_seen": 168759730, + "step": 7808, + "time_per_iteration": 2.6428685188293457 + }, + { + "auxiliary_loss_clip": 0.01088247, + "auxiliary_loss_mlp": 0.00999903, + "balance_loss_clip": 1.00944996, + "balance_loss_mlp": 0.99888366, + "epoch": 0.9389767330006613, + "flos": 58941083157120.0, + "grad_norm": 0.7092155859165561, + "language_loss": 0.5668239, + "learning_rate": 3.8867582157593895e-08, + "loss": 0.58770543, + "num_input_tokens_seen": 168813935, + "step": 7809, + "time_per_iteration": 3.927100896835327 + }, + { + "auxiliary_loss_clip": 0.011729, + "auxiliary_loss_mlp": 0.01025057, + "balance_loss_clip": 1.04597282, + "balance_loss_mlp": 1.01814866, + "epoch": 0.9390969758913005, + "flos": 31102554994560.0, + "grad_norm": 1.5388339679948333, + "language_loss": 0.76694846, + "learning_rate": 3.871490566850544e-08, + "loss": 0.78892803, + "num_input_tokens_seen": 168838145, + "step": 7810, + "time_per_iteration": 2.8618102073669434 + }, + { + "auxiliary_loss_clip": 0.01178107, + "auxiliary_loss_mlp": 0.01027232, + "balance_loss_clip": 1.04228079, + "balance_loss_mlp": 1.01979351, + "epoch": 0.9392172187819395, + "flos": 22419391173120.0, + "grad_norm": 1.6444413036042622, + "language_loss": 0.70545238, + "learning_rate": 3.856252670480642e-08, + "loss": 0.7275058, + "num_input_tokens_seen": 168856805, + "step": 7811, + "time_per_iteration": 2.7023887634277344 + }, + { + "auxiliary_loss_clip": 0.01178717, + "auxiliary_loss_mlp": 0.01023002, + "balance_loss_clip": 1.04083848, + "balance_loss_mlp": 1.01599264, + "epoch": 0.9393374616725786, + "flos": 19719483436800.0, + "grad_norm": 3.3513171231453245, + "language_loss": 0.81113058, + "learning_rate": 3.841044528961279e-08, + "loss": 0.83314776, + "num_input_tokens_seen": 168874600, + "step": 7812, + "time_per_iteration": 2.7007110118865967 + }, + { + "auxiliary_loss_clip": 0.01167727, + "auxiliary_loss_mlp": 0.01021764, + "balance_loss_clip": 1.0458945, + "balance_loss_mlp": 1.01477814, + "epoch": 0.9394577045632178, + "flos": 24170215800960.0, + "grad_norm": 2.123398251929975, + "language_loss": 0.79252863, + "learning_rate": 3.825866144599477e-08, + "loss": 0.81442356, + "num_input_tokens_seen": 168893655, + "step": 7813, + "time_per_iteration": 2.586637020111084 + }, + { + "auxiliary_loss_clip": 0.01182702, + "auxiliary_loss_mlp": 0.01026349, + "balance_loss_clip": 1.04230475, + "balance_loss_mlp": 1.01907122, + "epoch": 0.9395779474538568, + "flos": 19023929498880.0, + "grad_norm": 2.051664438257648, + "language_loss": 0.75353181, + "learning_rate": 3.8107175196978145e-08, + "loss": 0.77562231, + "num_input_tokens_seen": 168909960, + "step": 7814, + "time_per_iteration": 2.740838050842285 + }, + { + "auxiliary_loss_clip": 0.01187947, + "auxiliary_loss_mlp": 0.01026159, + "balance_loss_clip": 1.0422051, + "balance_loss_mlp": 1.0193851, + "epoch": 0.9396981903444959, + "flos": 14319129260160.0, + "grad_norm": 1.8668336036048623, + "language_loss": 0.76996946, + "learning_rate": 3.7955986565542996e-08, + "loss": 0.7921105, + "num_input_tokens_seen": 168928040, + "step": 7815, + "time_per_iteration": 2.6475954055786133 + }, + { + "auxiliary_loss_clip": 0.01192265, + "auxiliary_loss_mlp": 0.01027597, + "balance_loss_clip": 1.04211617, + "balance_loss_mlp": 1.02044773, + "epoch": 0.9398184332351349, + "flos": 34787564202240.0, + "grad_norm": 1.8307270343626816, + "language_loss": 0.6843313, + "learning_rate": 3.780509557462497e-08, + "loss": 0.70652986, + "num_input_tokens_seen": 168948240, + "step": 7816, + "time_per_iteration": 2.78251314163208 + }, + { + "auxiliary_loss_clip": 0.01181675, + "auxiliary_loss_mlp": 0.0102607, + "balance_loss_clip": 1.04276645, + "balance_loss_mlp": 1.01804399, + "epoch": 0.9399386761257741, + "flos": 25372253462400.0, + "grad_norm": 1.6258889618667836, + "language_loss": 0.75371504, + "learning_rate": 3.765450224711375e-08, + "loss": 0.77579248, + "num_input_tokens_seen": 168968745, + "step": 7817, + "time_per_iteration": 2.6710259914398193 + }, + { + "auxiliary_loss_clip": 0.01180624, + "auxiliary_loss_mlp": 0.01021911, + "balance_loss_clip": 1.04472673, + "balance_loss_mlp": 1.01473498, + "epoch": 0.9400589190164131, + "flos": 27304965584640.0, + "grad_norm": 1.7217894066860928, + "language_loss": 0.79893959, + "learning_rate": 3.750420660585396e-08, + "loss": 0.82096493, + "num_input_tokens_seen": 168990685, + "step": 7818, + "time_per_iteration": 2.7300424575805664 + }, + { + "auxiliary_loss_clip": 0.01164873, + "auxiliary_loss_mlp": 0.01033228, + "balance_loss_clip": 1.04648769, + "balance_loss_mlp": 1.0263133, + "epoch": 0.9401791619070522, + "flos": 23399859790080.0, + "grad_norm": 1.625375827164733, + "language_loss": 0.79798365, + "learning_rate": 3.735420867364603e-08, + "loss": 0.81996465, + "num_input_tokens_seen": 169011665, + "step": 7819, + "time_per_iteration": 2.5684077739715576 + }, + { + "auxiliary_loss_clip": 0.01203468, + "auxiliary_loss_mlp": 0.01023712, + "balance_loss_clip": 1.03785086, + "balance_loss_mlp": 1.01697636, + "epoch": 0.9402994047976914, + "flos": 35881403120640.0, + "grad_norm": 1.7764425604093932, + "language_loss": 0.6174016, + "learning_rate": 3.7204508473244186e-08, + "loss": 0.63967335, + "num_input_tokens_seen": 169035290, + "step": 7820, + "time_per_iteration": 2.9204766750335693 + }, + { + "auxiliary_loss_clip": 0.01215281, + "auxiliary_loss_mlp": 0.01023156, + "balance_loss_clip": 1.03918219, + "balance_loss_mlp": 1.01598525, + "epoch": 0.9404196476883304, + "flos": 22236821320320.0, + "grad_norm": 1.6568454446855143, + "language_loss": 0.69213611, + "learning_rate": 3.7055106027357395e-08, + "loss": 0.71452051, + "num_input_tokens_seen": 169055155, + "step": 7821, + "time_per_iteration": 2.7979581356048584 + }, + { + "auxiliary_loss_clip": 0.01173715, + "auxiliary_loss_mlp": 0.01024745, + "balance_loss_clip": 1.04622102, + "balance_loss_mlp": 1.0177716, + "epoch": 0.9405398905789695, + "flos": 18915802583040.0, + "grad_norm": 2.232736883358314, + "language_loss": 0.7172603, + "learning_rate": 3.690600135865063e-08, + "loss": 0.73924494, + "num_input_tokens_seen": 169072080, + "step": 7822, + "time_per_iteration": 2.575455904006958 + }, + { + "auxiliary_loss_clip": 0.01110427, + "auxiliary_loss_mlp": 0.01002264, + "balance_loss_clip": 1.00894213, + "balance_loss_mlp": 1.0011971, + "epoch": 0.9406601334696086, + "flos": 70274130048000.0, + "grad_norm": 0.7907832711778058, + "language_loss": 0.58110058, + "learning_rate": 3.675719448974246e-08, + "loss": 0.60222745, + "num_input_tokens_seen": 169137170, + "step": 7823, + "time_per_iteration": 3.384734869003296 + }, + { + "auxiliary_loss_clip": 0.01197934, + "auxiliary_loss_mlp": 0.01378475, + "balance_loss_clip": 1.04191232, + "balance_loss_mlp": 1.00012755, + "epoch": 0.9407803763602477, + "flos": 22165071903360.0, + "grad_norm": 2.1840287963531226, + "language_loss": 0.60308206, + "learning_rate": 3.6608685443207054e-08, + "loss": 0.62884611, + "num_input_tokens_seen": 169156320, + "step": 7824, + "time_per_iteration": 2.760082721710205 + }, + { + "auxiliary_loss_clip": 0.01195305, + "auxiliary_loss_mlp": 0.01022035, + "balance_loss_clip": 1.04359865, + "balance_loss_mlp": 1.01522231, + "epoch": 0.9409006192508867, + "flos": 18879496911360.0, + "grad_norm": 2.085572424439765, + "language_loss": 0.66619414, + "learning_rate": 3.646047424157306e-08, + "loss": 0.68836749, + "num_input_tokens_seen": 169173295, + "step": 7825, + "time_per_iteration": 3.5783863067626953 + }, + { + "auxiliary_loss_clip": 0.0118257, + "auxiliary_loss_mlp": 0.0102842, + "balance_loss_clip": 1.04427862, + "balance_loss_mlp": 1.0208981, + "epoch": 0.9410208621415259, + "flos": 23368258800000.0, + "grad_norm": 2.2494285690890248, + "language_loss": 0.6810748, + "learning_rate": 3.631256090732382e-08, + "loss": 0.70318472, + "num_input_tokens_seen": 169193755, + "step": 7826, + "time_per_iteration": 2.6524181365966797 + }, + { + "auxiliary_loss_clip": 0.01198226, + "auxiliary_loss_mlp": 0.01024156, + "balance_loss_clip": 1.04717553, + "balance_loss_mlp": 1.0175513, + "epoch": 0.941141105032165, + "flos": 22742227635840.0, + "grad_norm": 1.899591809508871, + "language_loss": 0.826841, + "learning_rate": 3.6164945462897833e-08, + "loss": 0.84906483, + "num_input_tokens_seen": 169213045, + "step": 7827, + "time_per_iteration": 2.7027251720428467 + }, + { + "auxiliary_loss_clip": 0.01174329, + "auxiliary_loss_mlp": 0.01378397, + "balance_loss_clip": 1.04620075, + "balance_loss_mlp": 1.0000875, + "epoch": 0.941261347922804, + "flos": 20704908130560.0, + "grad_norm": 1.8637482957061327, + "language_loss": 0.75701606, + "learning_rate": 3.6017627930687856e-08, + "loss": 0.7825433, + "num_input_tokens_seen": 169232870, + "step": 7828, + "time_per_iteration": 2.6765573024749756 + }, + { + "auxiliary_loss_clip": 0.01194063, + "auxiliary_loss_mlp": 0.0102722, + "balance_loss_clip": 1.03714216, + "balance_loss_mlp": 1.01985931, + "epoch": 0.9413815908134432, + "flos": 19421998997760.0, + "grad_norm": 2.408657775289213, + "language_loss": 0.77218276, + "learning_rate": 3.587060833304267e-08, + "loss": 0.79439557, + "num_input_tokens_seen": 169251060, + "step": 7829, + "time_per_iteration": 3.5643179416656494 + }, + { + "auxiliary_loss_clip": 0.01180032, + "auxiliary_loss_mlp": 0.01025254, + "balance_loss_clip": 1.04734838, + "balance_loss_mlp": 1.01800585, + "epoch": 0.9415018337040822, + "flos": 17493452853120.0, + "grad_norm": 2.1554236553295767, + "language_loss": 0.6386627, + "learning_rate": 3.5723886692264225e-08, + "loss": 0.66071558, + "num_input_tokens_seen": 169268600, + "step": 7830, + "time_per_iteration": 2.6051759719848633 + }, + { + "auxiliary_loss_clip": 0.01181288, + "auxiliary_loss_mlp": 0.01029471, + "balance_loss_clip": 1.04253006, + "balance_loss_mlp": 1.02246785, + "epoch": 0.9416220765947213, + "flos": 31831613343360.0, + "grad_norm": 2.062597526122502, + "language_loss": 0.6173178, + "learning_rate": 3.557746303061071e-08, + "loss": 0.6394254, + "num_input_tokens_seen": 169290355, + "step": 7831, + "time_per_iteration": 3.6576995849609375 + }, + { + "auxiliary_loss_clip": 0.01182888, + "auxiliary_loss_mlp": 0.01028658, + "balance_loss_clip": 1.04445243, + "balance_loss_mlp": 1.02112985, + "epoch": 0.9417423194853605, + "flos": 23511973115520.0, + "grad_norm": 1.584166291348187, + "language_loss": 0.72505748, + "learning_rate": 3.543133737029391e-08, + "loss": 0.74717295, + "num_input_tokens_seen": 169310865, + "step": 7832, + "time_per_iteration": 2.6457176208496094 + }, + { + "auxiliary_loss_clip": 0.0117992, + "auxiliary_loss_mlp": 0.0102425, + "balance_loss_clip": 1.04647017, + "balance_loss_mlp": 1.01664472, + "epoch": 0.9418625623759995, + "flos": 23915106432000.0, + "grad_norm": 2.049672237123612, + "language_loss": 0.68713939, + "learning_rate": 3.5285509733481214e-08, + "loss": 0.70918113, + "num_input_tokens_seen": 169330590, + "step": 7833, + "time_per_iteration": 2.6344072818756104 + }, + { + "auxiliary_loss_clip": 0.01170012, + "auxiliary_loss_mlp": 0.01025862, + "balance_loss_clip": 1.04386652, + "balance_loss_mlp": 1.01860857, + "epoch": 0.9419828052666386, + "flos": 18076965292800.0, + "grad_norm": 2.1050798985615864, + "language_loss": 0.76527947, + "learning_rate": 3.513998014229469e-08, + "loss": 0.78723824, + "num_input_tokens_seen": 169349540, + "step": 7834, + "time_per_iteration": 2.6360387802124023 + }, + { + "auxiliary_loss_clip": 0.0118914, + "auxiliary_loss_mlp": 0.01022379, + "balance_loss_clip": 1.04779005, + "balance_loss_mlp": 1.01578712, + "epoch": 0.9421030481572777, + "flos": 17712328377600.0, + "grad_norm": 2.600681365026926, + "language_loss": 0.86399543, + "learning_rate": 3.499474861881069e-08, + "loss": 0.88611066, + "num_input_tokens_seen": 169366765, + "step": 7835, + "time_per_iteration": 3.4368443489074707 + }, + { + "auxiliary_loss_clip": 0.01211337, + "auxiliary_loss_mlp": 0.01023029, + "balance_loss_clip": 1.04018974, + "balance_loss_mlp": 1.01618898, + "epoch": 0.9422232910479168, + "flos": 20194114775040.0, + "grad_norm": 2.1170523455524934, + "language_loss": 0.68431163, + "learning_rate": 3.4849815185061136e-08, + "loss": 0.70665532, + "num_input_tokens_seen": 169386655, + "step": 7836, + "time_per_iteration": 2.703338384628296 + }, + { + "auxiliary_loss_clip": 0.01171717, + "auxiliary_loss_mlp": 0.01024029, + "balance_loss_clip": 1.0425247, + "balance_loss_mlp": 1.01679897, + "epoch": 0.9423435339385559, + "flos": 18442571875200.0, + "grad_norm": 1.7919121266566316, + "language_loss": 0.76089287, + "learning_rate": 3.470517986303223e-08, + "loss": 0.78285027, + "num_input_tokens_seen": 169405640, + "step": 7837, + "time_per_iteration": 2.5996344089508057 + }, + { + "auxiliary_loss_clip": 0.01190529, + "auxiliary_loss_mlp": 0.01028807, + "balance_loss_clip": 1.04504275, + "balance_loss_mlp": 1.02145803, + "epoch": 0.942463776829195, + "flos": 20080636732800.0, + "grad_norm": 1.9274375889246675, + "language_loss": 0.79443312, + "learning_rate": 3.4560842674664856e-08, + "loss": 0.81662649, + "num_input_tokens_seen": 169424155, + "step": 7838, + "time_per_iteration": 2.7252748012542725 + }, + { + "auxiliary_loss_clip": 0.01174355, + "auxiliary_loss_mlp": 0.01022099, + "balance_loss_clip": 1.04219115, + "balance_loss_mlp": 1.0147202, + "epoch": 0.9425840197198341, + "flos": 22636255536000.0, + "grad_norm": 2.0674101951669157, + "language_loss": 0.75394732, + "learning_rate": 3.441680364185506e-08, + "loss": 0.77591181, + "num_input_tokens_seen": 169444025, + "step": 7839, + "time_per_iteration": 2.6444005966186523 + }, + { + "auxiliary_loss_clip": 0.01192033, + "auxiliary_loss_mlp": 0.01031542, + "balance_loss_clip": 1.04902458, + "balance_loss_mlp": 1.02423453, + "epoch": 0.9427042626104731, + "flos": 19937892084480.0, + "grad_norm": 7.7115292592870555, + "language_loss": 0.74634546, + "learning_rate": 3.427306278645314e-08, + "loss": 0.76858121, + "num_input_tokens_seen": 169462480, + "step": 7840, + "time_per_iteration": 2.6884560585021973 + }, + { + "auxiliary_loss_clip": 0.01204864, + "auxiliary_loss_mlp": 0.01020681, + "balance_loss_clip": 1.04247141, + "balance_loss_mlp": 1.0137279, + "epoch": 0.9428245055011123, + "flos": 22856998567680.0, + "grad_norm": 1.681363416738494, + "language_loss": 0.72984767, + "learning_rate": 3.4129620130264767e-08, + "loss": 0.75210309, + "num_input_tokens_seen": 169480840, + "step": 7841, + "time_per_iteration": 2.7377214431762695 + }, + { + "auxiliary_loss_clip": 0.01187392, + "auxiliary_loss_mlp": 0.01378468, + "balance_loss_clip": 1.04507816, + "balance_loss_mlp": 1.00015807, + "epoch": 0.9429447483917514, + "flos": 20951757371520.0, + "grad_norm": 2.609455664937956, + "language_loss": 0.77832711, + "learning_rate": 3.398647569505009e-08, + "loss": 0.80398571, + "num_input_tokens_seen": 169498265, + "step": 7842, + "time_per_iteration": 2.6449289321899414 + }, + { + "auxiliary_loss_clip": 0.01199174, + "auxiliary_loss_mlp": 0.01025608, + "balance_loss_clip": 1.0437367, + "balance_loss_mlp": 1.01747763, + "epoch": 0.9430649912823904, + "flos": 18843658116480.0, + "grad_norm": 3.194429178400648, + "language_loss": 0.75072122, + "learning_rate": 3.384362950252373e-08, + "loss": 0.77296913, + "num_input_tokens_seen": 169515235, + "step": 7843, + "time_per_iteration": 2.6873764991760254 + }, + { + "auxiliary_loss_clip": 0.01181603, + "auxiliary_loss_mlp": 0.01029022, + "balance_loss_clip": 1.04307854, + "balance_loss_mlp": 1.02185142, + "epoch": 0.9431852341730296, + "flos": 32556038837760.0, + "grad_norm": 1.9234303450588353, + "language_loss": 0.56915742, + "learning_rate": 3.3701081574355473e-08, + "loss": 0.59126365, + "num_input_tokens_seen": 169537195, + "step": 7844, + "time_per_iteration": 2.7262182235717773 + }, + { + "auxiliary_loss_clip": 0.01087861, + "auxiliary_loss_mlp": 0.01001695, + "balance_loss_clip": 1.00918627, + "balance_loss_mlp": 1.00068188, + "epoch": 0.9433054770636686, + "flos": 66904490252160.0, + "grad_norm": 0.6918854757375554, + "language_loss": 0.51663202, + "learning_rate": 3.3558831932169796e-08, + "loss": 0.53752762, + "num_input_tokens_seen": 169605865, + "step": 7845, + "time_per_iteration": 3.317070245742798 + }, + { + "auxiliary_loss_clip": 0.01172722, + "auxiliary_loss_mlp": 0.01026668, + "balance_loss_clip": 1.04428077, + "balance_loss_mlp": 1.01956558, + "epoch": 0.9434257199543077, + "flos": 26140346916480.0, + "grad_norm": 2.016743898144662, + "language_loss": 0.88472176, + "learning_rate": 3.341688059754588e-08, + "loss": 0.90671569, + "num_input_tokens_seen": 169621520, + "step": 7846, + "time_per_iteration": 2.7120139598846436 + }, + { + "auxiliary_loss_clip": 0.01200364, + "auxiliary_loss_mlp": 0.01378457, + "balance_loss_clip": 1.04263401, + "balance_loss_mlp": 1.00009, + "epoch": 0.9435459628449467, + "flos": 25003486483200.0, + "grad_norm": 1.9045799879940122, + "language_loss": 0.77777058, + "learning_rate": 3.327522759201762e-08, + "loss": 0.80355877, + "num_input_tokens_seen": 169641390, + "step": 7847, + "time_per_iteration": 2.733912467956543 + }, + { + "auxiliary_loss_clip": 0.01190902, + "auxiliary_loss_mlp": 0.01025701, + "balance_loss_clip": 1.04304397, + "balance_loss_mlp": 1.01830661, + "epoch": 0.9436662057355859, + "flos": 22163240309760.0, + "grad_norm": 2.2648486286530134, + "language_loss": 0.67093372, + "learning_rate": 3.313387293707359e-08, + "loss": 0.69309974, + "num_input_tokens_seen": 169660095, + "step": 7848, + "time_per_iteration": 2.7453079223632812 + }, + { + "auxiliary_loss_clip": 0.01189087, + "auxiliary_loss_mlp": 0.01029085, + "balance_loss_clip": 1.04281545, + "balance_loss_mlp": 1.02124095, + "epoch": 0.943786448626225, + "flos": 20118522602880.0, + "grad_norm": 4.20035957004185, + "language_loss": 0.68138301, + "learning_rate": 3.29928166541571e-08, + "loss": 0.70356476, + "num_input_tokens_seen": 169679050, + "step": 7849, + "time_per_iteration": 2.7097692489624023 + }, + { + "auxiliary_loss_clip": 0.01173936, + "auxiliary_loss_mlp": 0.0101993, + "balance_loss_clip": 1.04238999, + "balance_loss_mlp": 1.012658, + "epoch": 0.943906691516864, + "flos": 22090808534400.0, + "grad_norm": 1.9564543141961521, + "language_loss": 0.80505884, + "learning_rate": 3.2852058764666346e-08, + "loss": 0.82699752, + "num_input_tokens_seen": 169698150, + "step": 7850, + "time_per_iteration": 2.684955358505249 + }, + { + "auxiliary_loss_clip": 0.01185526, + "auxiliary_loss_mlp": 0.01020507, + "balance_loss_clip": 1.04649329, + "balance_loss_mlp": 1.01374769, + "epoch": 0.9440269344075032, + "flos": 35298501212160.0, + "grad_norm": 1.8320232736862583, + "language_loss": 0.68660843, + "learning_rate": 3.2711599289954264e-08, + "loss": 0.70866877, + "num_input_tokens_seen": 169722185, + "step": 7851, + "time_per_iteration": 3.732515573501587 + }, + { + "auxiliary_loss_clip": 0.0121058, + "auxiliary_loss_mlp": 0.01029904, + "balance_loss_clip": 1.03969419, + "balance_loss_mlp": 1.02288294, + "epoch": 0.9441471772981422, + "flos": 19238136255360.0, + "grad_norm": 1.7246786487280472, + "language_loss": 0.77860868, + "learning_rate": 3.257143825132847e-08, + "loss": 0.80101353, + "num_input_tokens_seen": 169740355, + "step": 7852, + "time_per_iteration": 2.7688329219818115 + }, + { + "auxiliary_loss_clip": 0.01185498, + "auxiliary_loss_mlp": 0.01018985, + "balance_loss_clip": 1.04443145, + "balance_loss_mlp": 1.01233315, + "epoch": 0.9442674201887813, + "flos": 25739799379200.0, + "grad_norm": 1.6968222784022937, + "language_loss": 0.75964224, + "learning_rate": 3.243157567005106e-08, + "loss": 0.78168714, + "num_input_tokens_seen": 169758535, + "step": 7853, + "time_per_iteration": 2.7105188369750977 + }, + { + "auxiliary_loss_clip": 0.01173232, + "auxiliary_loss_mlp": 0.01022588, + "balance_loss_clip": 1.05123138, + "balance_loss_mlp": 1.01558423, + "epoch": 0.9443876630794205, + "flos": 15523321737600.0, + "grad_norm": 1.8787869255206688, + "language_loss": 0.63920772, + "learning_rate": 3.2292011567339296e-08, + "loss": 0.66116595, + "num_input_tokens_seen": 169776340, + "step": 7854, + "time_per_iteration": 2.574603319168091 + }, + { + "auxiliary_loss_clip": 0.01175972, + "auxiliary_loss_mlp": 0.01378188, + "balance_loss_clip": 1.04493082, + "balance_loss_mlp": 1.0001173, + "epoch": 0.9445079059700595, + "flos": 13400821128960.0, + "grad_norm": 2.203625224036494, + "language_loss": 0.56087351, + "learning_rate": 3.21527459643649e-08, + "loss": 0.58641511, + "num_input_tokens_seen": 169793225, + "step": 7855, + "time_per_iteration": 3.5303163528442383 + }, + { + "auxiliary_loss_clip": 0.01179395, + "auxiliary_loss_mlp": 0.01022772, + "balance_loss_clip": 1.04646301, + "balance_loss_mlp": 1.01576579, + "epoch": 0.9446281488606986, + "flos": 23659242877440.0, + "grad_norm": 2.11493387427436, + "language_loss": 0.73962677, + "learning_rate": 3.2013778882254536e-08, + "loss": 0.76164842, + "num_input_tokens_seen": 169812020, + "step": 7856, + "time_per_iteration": 2.682004451751709 + }, + { + "auxiliary_loss_clip": 0.01165917, + "auxiliary_loss_mlp": 0.01023328, + "balance_loss_clip": 1.04488325, + "balance_loss_mlp": 1.01648188, + "epoch": 0.9447483917513377, + "flos": 25557337267200.0, + "grad_norm": 2.007791268915012, + "language_loss": 0.75991535, + "learning_rate": 3.1875110342088676e-08, + "loss": 0.78180778, + "num_input_tokens_seen": 169833470, + "step": 7857, + "time_per_iteration": 3.5733439922332764 + }, + { + "auxiliary_loss_clip": 0.01180429, + "auxiliary_loss_mlp": 0.01026248, + "balance_loss_clip": 1.04546928, + "balance_loss_mlp": 1.01901186, + "epoch": 0.9448686346419768, + "flos": 24535463247360.0, + "grad_norm": 1.6159140780170484, + "language_loss": 0.65696543, + "learning_rate": 3.1736740364904035e-08, + "loss": 0.67903221, + "num_input_tokens_seen": 169854000, + "step": 7858, + "time_per_iteration": 2.6782848834991455 + }, + { + "auxiliary_loss_clip": 0.01199364, + "auxiliary_loss_mlp": 0.01378391, + "balance_loss_clip": 1.04218674, + "balance_loss_mlp": 1.00023067, + "epoch": 0.9449888775326158, + "flos": 14721256995840.0, + "grad_norm": 2.082099669914836, + "language_loss": 0.77133119, + "learning_rate": 3.159866897169094e-08, + "loss": 0.79710877, + "num_input_tokens_seen": 169872200, + "step": 7859, + "time_per_iteration": 2.670823335647583 + }, + { + "auxiliary_loss_clip": 0.01202557, + "auxiliary_loss_mlp": 0.01026268, + "balance_loss_clip": 1.04583597, + "balance_loss_mlp": 1.01893973, + "epoch": 0.945109120423255, + "flos": 15447873219840.0, + "grad_norm": 1.8647553362869813, + "language_loss": 0.75516498, + "learning_rate": 3.146089618339487e-08, + "loss": 0.77745318, + "num_input_tokens_seen": 169889055, + "step": 7860, + "time_per_iteration": 2.652198314666748 + }, + { + "auxiliary_loss_clip": 0.01195447, + "auxiliary_loss_mlp": 0.0102699, + "balance_loss_clip": 1.04446352, + "balance_loss_mlp": 1.01943779, + "epoch": 0.9452293633138941, + "flos": 25448097029760.0, + "grad_norm": 2.4478238905518657, + "language_loss": 0.68445909, + "learning_rate": 3.132342202091554e-08, + "loss": 0.70668352, + "num_input_tokens_seen": 169909280, + "step": 7861, + "time_per_iteration": 3.513662815093994 + }, + { + "auxiliary_loss_clip": 0.0116968, + "auxiliary_loss_mlp": 0.0102565, + "balance_loss_clip": 1.04782736, + "balance_loss_mlp": 1.01826477, + "epoch": 0.9453496062045331, + "flos": 21215342350080.0, + "grad_norm": 2.791225658425068, + "language_loss": 0.68457675, + "learning_rate": 3.1186246505107595e-08, + "loss": 0.70653003, + "num_input_tokens_seen": 169928420, + "step": 7862, + "time_per_iteration": 2.4502196311950684 + }, + { + "auxiliary_loss_clip": 0.01176427, + "auxiliary_loss_mlp": 0.01021001, + "balance_loss_clip": 1.04825497, + "balance_loss_mlp": 1.01405668, + "epoch": 0.9454698490951723, + "flos": 20010898477440.0, + "grad_norm": 2.0786747383191257, + "language_loss": 0.83689326, + "learning_rate": 3.104936965678084e-08, + "loss": 0.85886753, + "num_input_tokens_seen": 169946750, + "step": 7863, + "time_per_iteration": 2.469641923904419 + }, + { + "auxiliary_loss_clip": 0.01175797, + "auxiliary_loss_mlp": 0.010208, + "balance_loss_clip": 1.04420936, + "balance_loss_mlp": 1.01361775, + "epoch": 0.9455900919858113, + "flos": 21069652786560.0, + "grad_norm": 2.478955569744802, + "language_loss": 0.81909823, + "learning_rate": 3.091279149669956e-08, + "loss": 0.84106421, + "num_input_tokens_seen": 169965540, + "step": 7864, + "time_per_iteration": 2.481698751449585 + }, + { + "auxiliary_loss_clip": 0.01175353, + "auxiliary_loss_mlp": 0.01378468, + "balance_loss_clip": 1.04491258, + "balance_loss_mlp": 1.0000875, + "epoch": 0.9457103348764504, + "flos": 20740854666240.0, + "grad_norm": 2.153779905686746, + "language_loss": 0.73615932, + "learning_rate": 3.0776512045581624e-08, + "loss": 0.76169753, + "num_input_tokens_seen": 169984330, + "step": 7865, + "time_per_iteration": 2.584444999694824 + }, + { + "auxiliary_loss_clip": 0.01179105, + "auxiliary_loss_mlp": 0.01026993, + "balance_loss_clip": 1.04435384, + "balance_loss_mlp": 1.01923263, + "epoch": 0.9458305777670896, + "flos": 21428363957760.0, + "grad_norm": 2.6849565580589636, + "language_loss": 0.7800855, + "learning_rate": 3.0640531324101384e-08, + "loss": 0.80214643, + "num_input_tokens_seen": 170002095, + "step": 7866, + "time_per_iteration": 2.5936598777770996 + }, + { + "auxiliary_loss_clip": 0.0118089, + "auxiliary_loss_mlp": 0.01024458, + "balance_loss_clip": 1.0505929, + "balance_loss_mlp": 1.01620841, + "epoch": 0.9459508206577286, + "flos": 20011185786240.0, + "grad_norm": 1.7888126485084987, + "language_loss": 0.76057845, + "learning_rate": 3.0504849352886554e-08, + "loss": 0.78263193, + "num_input_tokens_seen": 170020240, + "step": 7867, + "time_per_iteration": 2.5696942806243896 + }, + { + "auxiliary_loss_clip": 0.01177819, + "auxiliary_loss_mlp": 0.01030121, + "balance_loss_clip": 1.04866719, + "balance_loss_mlp": 1.02196681, + "epoch": 0.9460710635483677, + "flos": 12166428291840.0, + "grad_norm": 2.212393083821583, + "language_loss": 0.71501076, + "learning_rate": 3.036946615252023e-08, + "loss": 0.73709017, + "num_input_tokens_seen": 170035770, + "step": 7868, + "time_per_iteration": 2.5790722370147705 + }, + { + "auxiliary_loss_clip": 0.01191938, + "auxiliary_loss_mlp": 0.01026536, + "balance_loss_clip": 1.0456512, + "balance_loss_mlp": 1.01944923, + "epoch": 0.9461913064390068, + "flos": 34276196229120.0, + "grad_norm": 2.4437905344169124, + "language_loss": 0.67261708, + "learning_rate": 3.0234381743539984e-08, + "loss": 0.69480181, + "num_input_tokens_seen": 170053385, + "step": 7869, + "time_per_iteration": 2.721407890319824 + }, + { + "auxiliary_loss_clip": 0.01187692, + "auxiliary_loss_mlp": 0.01026478, + "balance_loss_clip": 1.04313743, + "balance_loss_mlp": 1.01931345, + "epoch": 0.9463115493296459, + "flos": 19463763536640.0, + "grad_norm": 2.059736902585293, + "language_loss": 0.79822773, + "learning_rate": 3.0099596146437863e-08, + "loss": 0.82036948, + "num_input_tokens_seen": 170070490, + "step": 7870, + "time_per_iteration": 2.5962307453155518 + }, + { + "auxiliary_loss_clip": 0.01059567, + "auxiliary_loss_mlp": 0.01003174, + "balance_loss_clip": 1.00867462, + "balance_loss_mlp": 1.00214291, + "epoch": 0.946431792220285, + "flos": 70570824387840.0, + "grad_norm": 0.7786979246873001, + "language_loss": 0.60035902, + "learning_rate": 2.996510938166086e-08, + "loss": 0.6209864, + "num_input_tokens_seen": 170133465, + "step": 7871, + "time_per_iteration": 3.189035654067993 + }, + { + "auxiliary_loss_clip": 0.0117562, + "auxiliary_loss_mlp": 0.01021257, + "balance_loss_clip": 1.04660273, + "balance_loss_mlp": 1.01423824, + "epoch": 0.9465520351109241, + "flos": 18947906363520.0, + "grad_norm": 2.1543819149633094, + "language_loss": 0.7395305, + "learning_rate": 2.983092146960997e-08, + "loss": 0.76149929, + "num_input_tokens_seen": 170150810, + "step": 7872, + "time_per_iteration": 2.51602840423584 + }, + { + "auxiliary_loss_clip": 0.01186145, + "auxiliary_loss_mlp": 0.01026308, + "balance_loss_clip": 1.04380548, + "balance_loss_mlp": 1.01916134, + "epoch": 0.9466722780015632, + "flos": 19135647774720.0, + "grad_norm": 3.6314311088258835, + "language_loss": 0.8018465, + "learning_rate": 2.9697032430642256e-08, + "loss": 0.82397103, + "num_input_tokens_seen": 170169025, + "step": 7873, + "time_per_iteration": 2.565406084060669 + }, + { + "auxiliary_loss_clip": 0.01163745, + "auxiliary_loss_mlp": 0.01021819, + "balance_loss_clip": 1.04630232, + "balance_loss_mlp": 1.01538777, + "epoch": 0.9467925208922022, + "flos": 17237912520960.0, + "grad_norm": 2.2827524552140637, + "language_loss": 0.73529255, + "learning_rate": 2.9563442285067906e-08, + "loss": 0.75714815, + "num_input_tokens_seen": 170186070, + "step": 7874, + "time_per_iteration": 2.5927228927612305 + }, + { + "auxiliary_loss_clip": 0.01181731, + "auxiliary_loss_mlp": 0.01025002, + "balance_loss_clip": 1.04818559, + "balance_loss_mlp": 1.01788831, + "epoch": 0.9469127637828414, + "flos": 29169016859520.0, + "grad_norm": 2.0451869660456503, + "language_loss": 0.79805917, + "learning_rate": 2.943015105315294e-08, + "loss": 0.82012647, + "num_input_tokens_seen": 170206265, + "step": 7875, + "time_per_iteration": 2.669466257095337 + }, + { + "auxiliary_loss_clip": 0.01202099, + "auxiliary_loss_mlp": 0.01023339, + "balance_loss_clip": 1.04060698, + "balance_loss_mlp": 1.01524472, + "epoch": 0.9470330066734804, + "flos": 26030460234240.0, + "grad_norm": 2.5977755443795036, + "language_loss": 0.66655302, + "learning_rate": 2.929715875511718e-08, + "loss": 0.68880737, + "num_input_tokens_seen": 170225300, + "step": 7876, + "time_per_iteration": 2.785858392715454 + }, + { + "auxiliary_loss_clip": 0.0117668, + "auxiliary_loss_mlp": 0.01026255, + "balance_loss_clip": 1.04362369, + "balance_loss_mlp": 1.01929283, + "epoch": 0.9471532495641195, + "flos": 23440906056960.0, + "grad_norm": 1.9396745473771768, + "language_loss": 0.70077795, + "learning_rate": 2.9164465411135375e-08, + "loss": 0.72280723, + "num_input_tokens_seen": 170245070, + "step": 7877, + "time_per_iteration": 3.5203018188476562 + }, + { + "auxiliary_loss_clip": 0.01178196, + "auxiliary_loss_mlp": 0.01024202, + "balance_loss_clip": 1.04790461, + "balance_loss_mlp": 1.01703715, + "epoch": 0.9472734924547586, + "flos": 15815850099840.0, + "grad_norm": 1.9801077944440315, + "language_loss": 0.80912173, + "learning_rate": 2.9032071041337426e-08, + "loss": 0.83114564, + "num_input_tokens_seen": 170263305, + "step": 7878, + "time_per_iteration": 2.632666826248169 + }, + { + "auxiliary_loss_clip": 0.01177251, + "auxiliary_loss_mlp": 0.01025595, + "balance_loss_clip": 1.04396558, + "balance_loss_mlp": 1.01809072, + "epoch": 0.9473937353453977, + "flos": 11181793697280.0, + "grad_norm": 1.577815836907772, + "language_loss": 0.72548151, + "learning_rate": 2.889997566580704e-08, + "loss": 0.74750996, + "num_input_tokens_seen": 170281460, + "step": 7879, + "time_per_iteration": 2.592383623123169 + }, + { + "auxiliary_loss_clip": 0.01168305, + "auxiliary_loss_mlp": 0.01020787, + "balance_loss_clip": 1.04655814, + "balance_loss_mlp": 1.01270437, + "epoch": 0.9475139782360368, + "flos": 25775530433280.0, + "grad_norm": 1.6819917535810027, + "language_loss": 0.70269889, + "learning_rate": 2.8768179304583086e-08, + "loss": 0.72458982, + "num_input_tokens_seen": 170303515, + "step": 7880, + "time_per_iteration": 2.606621026992798 + }, + { + "auxiliary_loss_clip": 0.01194139, + "auxiliary_loss_mlp": 0.01026466, + "balance_loss_clip": 1.04699898, + "balance_loss_mlp": 1.0196054, + "epoch": 0.9476342211266758, + "flos": 22820046451200.0, + "grad_norm": 1.631215065000974, + "language_loss": 0.73702115, + "learning_rate": 2.8636681977659117e-08, + "loss": 0.75922722, + "num_input_tokens_seen": 170323165, + "step": 7881, + "time_per_iteration": 3.6160757541656494 + }, + { + "auxiliary_loss_clip": 0.01200618, + "auxiliary_loss_mlp": 0.01026389, + "balance_loss_clip": 1.04465365, + "balance_loss_mlp": 1.01927495, + "epoch": 0.947754464017315, + "flos": 20193611984640.0, + "grad_norm": 2.059812608659302, + "language_loss": 0.78130019, + "learning_rate": 2.850548370498318e-08, + "loss": 0.80357033, + "num_input_tokens_seen": 170341005, + "step": 7882, + "time_per_iteration": 2.6671254634857178 + }, + { + "auxiliary_loss_clip": 0.01173979, + "auxiliary_loss_mlp": 0.01023325, + "balance_loss_clip": 1.04412901, + "balance_loss_mlp": 1.01646733, + "epoch": 0.9478747069079541, + "flos": 24717925359360.0, + "grad_norm": 2.5995918655878314, + "language_loss": 0.71153688, + "learning_rate": 2.8374584506457798e-08, + "loss": 0.7335099, + "num_input_tokens_seen": 170362280, + "step": 7883, + "time_per_iteration": 3.5553853511810303 + }, + { + "auxiliary_loss_clip": 0.01182763, + "auxiliary_loss_mlp": 0.01021315, + "balance_loss_clip": 1.0459739, + "balance_loss_mlp": 1.0144453, + "epoch": 0.9479949497985931, + "flos": 21361355136000.0, + "grad_norm": 2.8739276222250303, + "language_loss": 0.67566115, + "learning_rate": 2.824398440193998e-08, + "loss": 0.69770193, + "num_input_tokens_seen": 170381080, + "step": 7884, + "time_per_iteration": 2.616255283355713 + }, + { + "auxiliary_loss_clip": 0.01191711, + "auxiliary_loss_mlp": 0.01030446, + "balance_loss_clip": 1.0393703, + "balance_loss_mlp": 1.02275705, + "epoch": 0.9481151926892323, + "flos": 18148606968960.0, + "grad_norm": 1.9987690906191513, + "language_loss": 0.71392095, + "learning_rate": 2.811368341124232e-08, + "loss": 0.73614252, + "num_input_tokens_seen": 170400150, + "step": 7885, + "time_per_iteration": 2.74189829826355 + }, + { + "auxiliary_loss_clip": 0.0118315, + "auxiliary_loss_mlp": 0.01027808, + "balance_loss_clip": 1.0436641, + "balance_loss_mlp": 1.02119493, + "epoch": 0.9482354355798713, + "flos": 22128012046080.0, + "grad_norm": 2.176987519881095, + "language_loss": 0.68031299, + "learning_rate": 2.7983681554131222e-08, + "loss": 0.70242262, + "num_input_tokens_seen": 170420410, + "step": 7886, + "time_per_iteration": 2.642841339111328 + }, + { + "auxiliary_loss_clip": 0.01180348, + "auxiliary_loss_mlp": 0.01025742, + "balance_loss_clip": 1.04291177, + "balance_loss_mlp": 1.01823139, + "epoch": 0.9483556784705104, + "flos": 19063072344960.0, + "grad_norm": 2.808359659323168, + "language_loss": 0.70681137, + "learning_rate": 2.7853978850327365e-08, + "loss": 0.72887224, + "num_input_tokens_seen": 170439580, + "step": 7887, + "time_per_iteration": 3.424467086791992 + }, + { + "auxiliary_loss_clip": 0.01195433, + "auxiliary_loss_mlp": 0.01025892, + "balance_loss_clip": 1.04656005, + "balance_loss_mlp": 1.01906157, + "epoch": 0.9484759213611496, + "flos": 25777110631680.0, + "grad_norm": 1.8246759011907532, + "language_loss": 0.87322366, + "learning_rate": 2.7724575319507225e-08, + "loss": 0.89543688, + "num_input_tokens_seen": 170459290, + "step": 7888, + "time_per_iteration": 2.683687686920166 + }, + { + "auxiliary_loss_clip": 0.01172009, + "auxiliary_loss_mlp": 0.01024807, + "balance_loss_clip": 1.04229784, + "balance_loss_mlp": 1.0177592, + "epoch": 0.9485961642517886, + "flos": 20667740532480.0, + "grad_norm": 1.9808486023840748, + "language_loss": 0.77073216, + "learning_rate": 2.759547098130044e-08, + "loss": 0.79270035, + "num_input_tokens_seen": 170478020, + "step": 7889, + "time_per_iteration": 2.6390879154205322 + }, + { + "auxiliary_loss_clip": 0.01165104, + "auxiliary_loss_mlp": 0.01023461, + "balance_loss_clip": 1.04677939, + "balance_loss_mlp": 1.01660609, + "epoch": 0.9487164071424277, + "flos": 22674069578880.0, + "grad_norm": 1.9698479503757487, + "language_loss": 0.76449919, + "learning_rate": 2.746666585529267e-08, + "loss": 0.78638482, + "num_input_tokens_seen": 170498295, + "step": 7890, + "time_per_iteration": 2.596662998199463 + }, + { + "auxiliary_loss_clip": 0.01166283, + "auxiliary_loss_mlp": 0.01026473, + "balance_loss_clip": 1.04463983, + "balance_loss_mlp": 1.01912379, + "epoch": 0.9488366500330668, + "flos": 38726461716480.0, + "grad_norm": 2.038830734461327, + "language_loss": 0.74142158, + "learning_rate": 2.73381599610234e-08, + "loss": 0.76334918, + "num_input_tokens_seen": 170518695, + "step": 7891, + "time_per_iteration": 2.7323691844940186 + }, + { + "auxiliary_loss_clip": 0.0117029, + "auxiliary_loss_mlp": 0.01024404, + "balance_loss_clip": 1.04279971, + "balance_loss_mlp": 1.01717424, + "epoch": 0.9489568929237059, + "flos": 27890920149120.0, + "grad_norm": 1.7401188324821253, + "language_loss": 0.71394187, + "learning_rate": 2.7209953317987033e-08, + "loss": 0.73588884, + "num_input_tokens_seen": 170539735, + "step": 7892, + "time_per_iteration": 2.6636712551116943 + }, + { + "auxiliary_loss_clip": 0.01177645, + "auxiliary_loss_mlp": 0.01028352, + "balance_loss_clip": 1.04634738, + "balance_loss_mlp": 1.02031755, + "epoch": 0.9490771358143449, + "flos": 33580642291200.0, + "grad_norm": 1.7934737492269301, + "language_loss": 0.78269863, + "learning_rate": 2.7082045945631793e-08, + "loss": 0.80475855, + "num_input_tokens_seen": 170561950, + "step": 7893, + "time_per_iteration": 2.68916916847229 + }, + { + "auxiliary_loss_clip": 0.0118248, + "auxiliary_loss_mlp": 0.01021752, + "balance_loss_clip": 1.04037404, + "balance_loss_mlp": 1.0147754, + "epoch": 0.9491973787049841, + "flos": 14793796512000.0, + "grad_norm": 2.001928266055011, + "language_loss": 0.69230354, + "learning_rate": 2.6954437863361712e-08, + "loss": 0.71434581, + "num_input_tokens_seen": 170579865, + "step": 7894, + "time_per_iteration": 2.7016501426696777 + }, + { + "auxiliary_loss_clip": 0.01215088, + "auxiliary_loss_mlp": 0.01024617, + "balance_loss_clip": 1.04112577, + "balance_loss_mlp": 1.01785731, + "epoch": 0.9493176215956232, + "flos": 25332535998720.0, + "grad_norm": 2.388876417407643, + "language_loss": 0.71302581, + "learning_rate": 2.6827129090534862e-08, + "loss": 0.73542285, + "num_input_tokens_seen": 170600165, + "step": 7895, + "time_per_iteration": 2.8630664348602295 + }, + { + "auxiliary_loss_clip": 0.01183115, + "auxiliary_loss_mlp": 0.01026576, + "balance_loss_clip": 1.04540491, + "balance_loss_mlp": 1.01870847, + "epoch": 0.9494378644862622, + "flos": 21029971236480.0, + "grad_norm": 1.8772368419634569, + "language_loss": 0.77726364, + "learning_rate": 2.670011964646335e-08, + "loss": 0.79936051, + "num_input_tokens_seen": 170618845, + "step": 7896, + "time_per_iteration": 2.644531011581421 + }, + { + "auxiliary_loss_clip": 0.01219214, + "auxiliary_loss_mlp": 0.01024588, + "balance_loss_clip": 1.03350842, + "balance_loss_mlp": 1.01704478, + "epoch": 0.9495581073769014, + "flos": 15195134148480.0, + "grad_norm": 2.097317671410859, + "language_loss": 0.68497032, + "learning_rate": 2.657340955041487e-08, + "loss": 0.70740837, + "num_input_tokens_seen": 170637620, + "step": 7897, + "time_per_iteration": 2.7765042781829834 + }, + { + "auxiliary_loss_clip": 0.01183258, + "auxiliary_loss_mlp": 0.01029145, + "balance_loss_clip": 1.04534948, + "balance_loss_mlp": 1.02175438, + "epoch": 0.9496783502675404, + "flos": 28616566705920.0, + "grad_norm": 1.799005028467176, + "language_loss": 0.71459329, + "learning_rate": 2.6446998821611167e-08, + "loss": 0.73671734, + "num_input_tokens_seen": 170657815, + "step": 7898, + "time_per_iteration": 2.7279322147369385 + }, + { + "auxiliary_loss_clip": 0.01203688, + "auxiliary_loss_mlp": 0.01022923, + "balance_loss_clip": 1.04187322, + "balance_loss_mlp": 1.01589584, + "epoch": 0.9497985931581795, + "flos": 14866874732160.0, + "grad_norm": 2.98495279240996, + "language_loss": 0.71830803, + "learning_rate": 2.6320887479228228e-08, + "loss": 0.74057412, + "num_input_tokens_seen": 170674415, + "step": 7899, + "time_per_iteration": 2.6663155555725098 + }, + { + "auxiliary_loss_clip": 0.01187196, + "auxiliary_loss_mlp": 0.01023645, + "balance_loss_clip": 1.04405928, + "balance_loss_mlp": 1.01684666, + "epoch": 0.9499188360488187, + "flos": 27193319136000.0, + "grad_norm": 2.2736280506311126, + "language_loss": 0.72728354, + "learning_rate": 2.619507554239786e-08, + "loss": 0.74939191, + "num_input_tokens_seen": 170692975, + "step": 7900, + "time_per_iteration": 2.651700496673584 + }, + { + "auxiliary_loss_clip": 0.01181394, + "auxiliary_loss_mlp": 0.01023316, + "balance_loss_clip": 1.04294968, + "balance_loss_mlp": 1.01614547, + "epoch": 0.9500390789394577, + "flos": 24316479982080.0, + "grad_norm": 1.7595273329278152, + "language_loss": 0.69832337, + "learning_rate": 2.606956303020502e-08, + "loss": 0.72037053, + "num_input_tokens_seen": 170713780, + "step": 7901, + "time_per_iteration": 2.755941152572632 + }, + { + "auxiliary_loss_clip": 0.01177728, + "auxiliary_loss_mlp": 0.01027407, + "balance_loss_clip": 1.04769039, + "balance_loss_mlp": 1.02032626, + "epoch": 0.9501593218300968, + "flos": 14354752573440.0, + "grad_norm": 1.7612391112750518, + "language_loss": 0.84225214, + "learning_rate": 2.5944349961690036e-08, + "loss": 0.86430353, + "num_input_tokens_seen": 170730800, + "step": 7902, + "time_per_iteration": 2.5821144580841064 + }, + { + "auxiliary_loss_clip": 0.01193984, + "auxiliary_loss_mlp": 0.01027148, + "balance_loss_clip": 1.04319668, + "balance_loss_mlp": 1.01930439, + "epoch": 0.9502795647207359, + "flos": 38728113742080.0, + "grad_norm": 1.5840721472548458, + "language_loss": 0.73285609, + "learning_rate": 2.581943635584749e-08, + "loss": 0.75506747, + "num_input_tokens_seen": 170753630, + "step": 7903, + "time_per_iteration": 3.7187469005584717 + }, + { + "auxiliary_loss_clip": 0.01173996, + "auxiliary_loss_mlp": 0.01022695, + "balance_loss_clip": 1.04417419, + "balance_loss_mlp": 1.01519656, + "epoch": 0.950399807611375, + "flos": 40808023799040.0, + "grad_norm": 1.9965681977191005, + "language_loss": 0.65230489, + "learning_rate": 2.569482223162689e-08, + "loss": 0.67427182, + "num_input_tokens_seen": 170777605, + "step": 7904, + "time_per_iteration": 2.870558738708496 + }, + { + "auxiliary_loss_clip": 0.01176012, + "auxiliary_loss_mlp": 0.01025975, + "balance_loss_clip": 1.04395401, + "balance_loss_mlp": 1.01826859, + "epoch": 0.950520050502014, + "flos": 23440403266560.0, + "grad_norm": 1.885934573762322, + "language_loss": 0.72535586, + "learning_rate": 2.5570507607932e-08, + "loss": 0.74737567, + "num_input_tokens_seen": 170797520, + "step": 7905, + "time_per_iteration": 2.6401379108428955 + }, + { + "auxiliary_loss_clip": 0.01180038, + "auxiliary_loss_mlp": 0.01024636, + "balance_loss_clip": 1.04681659, + "balance_loss_mlp": 1.01722765, + "epoch": 0.9506402933926532, + "flos": 17783718658560.0, + "grad_norm": 3.2743642650206586, + "language_loss": 0.63832748, + "learning_rate": 2.54464925036213e-08, + "loss": 0.66037428, + "num_input_tokens_seen": 170814810, + "step": 7906, + "time_per_iteration": 3.480067491531372 + }, + { + "auxiliary_loss_clip": 0.01174246, + "auxiliary_loss_mlp": 0.01023475, + "balance_loss_clip": 1.04522133, + "balance_loss_mlp": 1.01653719, + "epoch": 0.9507605362832923, + "flos": 32561928668160.0, + "grad_norm": 1.8796007738090905, + "language_loss": 0.60589862, + "learning_rate": 2.532277693750773e-08, + "loss": 0.6278758, + "num_input_tokens_seen": 170835735, + "step": 7907, + "time_per_iteration": 2.6632297039031982 + }, + { + "auxiliary_loss_clip": 0.01200478, + "auxiliary_loss_mlp": 0.01029699, + "balance_loss_clip": 1.04568052, + "balance_loss_mlp": 1.02281511, + "epoch": 0.9508807791739313, + "flos": 19602054898560.0, + "grad_norm": 1.922821152143249, + "language_loss": 0.75810635, + "learning_rate": 2.5199360928358948e-08, + "loss": 0.78040814, + "num_input_tokens_seen": 170852970, + "step": 7908, + "time_per_iteration": 3.637299060821533 + }, + { + "auxiliary_loss_clip": 0.011645, + "auxiliary_loss_mlp": 0.01377934, + "balance_loss_clip": 1.04275405, + "balance_loss_mlp": 1.00009704, + "epoch": 0.9510010220645704, + "flos": 21471852349440.0, + "grad_norm": 1.8055852098037843, + "language_loss": 0.87208921, + "learning_rate": 2.507624449489665e-08, + "loss": 0.89751351, + "num_input_tokens_seen": 170871600, + "step": 7909, + "time_per_iteration": 2.658816337585449 + }, + { + "auxiliary_loss_clip": 0.01186118, + "auxiliary_loss_mlp": 0.01027647, + "balance_loss_clip": 1.04514623, + "balance_loss_mlp": 1.02075934, + "epoch": 0.9511212649552095, + "flos": 18879999701760.0, + "grad_norm": 2.205816272803594, + "language_loss": 0.64726734, + "learning_rate": 2.495342765579811e-08, + "loss": 0.66940498, + "num_input_tokens_seen": 170890260, + "step": 7910, + "time_per_iteration": 2.614304304122925 + }, + { + "auxiliary_loss_clip": 0.01200845, + "auxiliary_loss_mlp": 0.01027007, + "balance_loss_clip": 1.04309118, + "balance_loss_mlp": 1.01974773, + "epoch": 0.9512415078458486, + "flos": 20810521094400.0, + "grad_norm": 1.72373043566932, + "language_loss": 0.71557456, + "learning_rate": 2.4830910429693984e-08, + "loss": 0.73785311, + "num_input_tokens_seen": 170910220, + "step": 7911, + "time_per_iteration": 2.841581106185913 + }, + { + "auxiliary_loss_clip": 0.01168414, + "auxiliary_loss_mlp": 0.01025837, + "balance_loss_clip": 1.04783416, + "balance_loss_mlp": 1.01854157, + "epoch": 0.9513617507364877, + "flos": 18369565482240.0, + "grad_norm": 1.9902368841392881, + "language_loss": 0.79289848, + "learning_rate": 2.470869283517052e-08, + "loss": 0.81484103, + "num_input_tokens_seen": 170928255, + "step": 7912, + "time_per_iteration": 3.3664004802703857 + }, + { + "auxiliary_loss_clip": 0.01167598, + "auxiliary_loss_mlp": 0.01029827, + "balance_loss_clip": 1.04359794, + "balance_loss_mlp": 1.02277601, + "epoch": 0.9514819936271268, + "flos": 25010166412800.0, + "grad_norm": 1.5472760374844128, + "language_loss": 0.76995289, + "learning_rate": 2.458677489076777e-08, + "loss": 0.7919271, + "num_input_tokens_seen": 170949265, + "step": 7913, + "time_per_iteration": 2.6369616985321045 + }, + { + "auxiliary_loss_clip": 0.01163258, + "auxiliary_loss_mlp": 0.01019479, + "balance_loss_clip": 1.04199946, + "balance_loss_mlp": 1.01240671, + "epoch": 0.9516022365177659, + "flos": 18662129758080.0, + "grad_norm": 2.029996048511996, + "language_loss": 0.83122241, + "learning_rate": 2.446515661498072e-08, + "loss": 0.85304976, + "num_input_tokens_seen": 170968595, + "step": 7914, + "time_per_iteration": 2.544128656387329 + }, + { + "auxiliary_loss_clip": 0.01210553, + "auxiliary_loss_mlp": 0.01026925, + "balance_loss_clip": 1.04070222, + "balance_loss_mlp": 1.01980793, + "epoch": 0.9517224794084049, + "flos": 25372109808000.0, + "grad_norm": 2.289593619335178, + "language_loss": 0.75162971, + "learning_rate": 2.434383802625861e-08, + "loss": 0.77400446, + "num_input_tokens_seen": 170987550, + "step": 7915, + "time_per_iteration": 2.76873517036438 + }, + { + "auxiliary_loss_clip": 0.01191993, + "auxiliary_loss_mlp": 0.0102685, + "balance_loss_clip": 1.04119837, + "balance_loss_mlp": 1.01956058, + "epoch": 0.9518427222990441, + "flos": 21470918595840.0, + "grad_norm": 1.8279842668347677, + "language_loss": 0.73981094, + "learning_rate": 2.4222819143005168e-08, + "loss": 0.76199937, + "num_input_tokens_seen": 171007145, + "step": 7916, + "time_per_iteration": 2.6582484245300293 + }, + { + "auxiliary_loss_clip": 0.01166155, + "auxiliary_loss_mlp": 0.01022561, + "balance_loss_clip": 1.04783726, + "balance_loss_mlp": 1.01556039, + "epoch": 0.9519629651896832, + "flos": 21033634423680.0, + "grad_norm": 1.812082125761576, + "language_loss": 0.80984312, + "learning_rate": 2.4102099983579706e-08, + "loss": 0.83173031, + "num_input_tokens_seen": 171026295, + "step": 7917, + "time_per_iteration": 2.5719892978668213 + }, + { + "auxiliary_loss_clip": 0.01173379, + "auxiliary_loss_mlp": 0.01032621, + "balance_loss_clip": 1.04418397, + "balance_loss_mlp": 1.02441955, + "epoch": 0.9520832080803222, + "flos": 21689219502720.0, + "grad_norm": 1.697466838169946, + "language_loss": 0.77415586, + "learning_rate": 2.3981680566294236e-08, + "loss": 0.79621577, + "num_input_tokens_seen": 171045895, + "step": 7918, + "time_per_iteration": 2.64448881149292 + }, + { + "auxiliary_loss_clip": 0.01166932, + "auxiliary_loss_mlp": 0.01024756, + "balance_loss_clip": 1.04839838, + "balance_loss_mlp": 1.01809812, + "epoch": 0.9522034509709614, + "flos": 23145289125120.0, + "grad_norm": 2.083737078821117, + "language_loss": 0.73498094, + "learning_rate": 2.3861560909416822e-08, + "loss": 0.75689781, + "num_input_tokens_seen": 171065445, + "step": 7919, + "time_per_iteration": 2.5723423957824707 + }, + { + "auxiliary_loss_clip": 0.01201851, + "auxiliary_loss_mlp": 0.01021628, + "balance_loss_clip": 1.042781, + "balance_loss_mlp": 1.01480913, + "epoch": 0.9523236938616004, + "flos": 24679428958080.0, + "grad_norm": 1.6382096433838724, + "language_loss": 0.82709247, + "learning_rate": 2.3741741031169325e-08, + "loss": 0.84932733, + "num_input_tokens_seen": 171085015, + "step": 7920, + "time_per_iteration": 2.7010409832000732 + }, + { + "auxiliary_loss_clip": 0.01195315, + "auxiliary_loss_mlp": 0.010269, + "balance_loss_clip": 1.03907585, + "balance_loss_mlp": 1.01982188, + "epoch": 0.9524439367522395, + "flos": 22672309812480.0, + "grad_norm": 1.8737949966149243, + "language_loss": 0.71677619, + "learning_rate": 2.3622220949728544e-08, + "loss": 0.73899829, + "num_input_tokens_seen": 171103900, + "step": 7921, + "time_per_iteration": 2.703613042831421 + }, + { + "auxiliary_loss_clip": 0.01165693, + "auxiliary_loss_mlp": 0.0102779, + "balance_loss_clip": 1.04325855, + "balance_loss_mlp": 1.02019691, + "epoch": 0.9525641796428787, + "flos": 34055525024640.0, + "grad_norm": 2.747911001025932, + "language_loss": 0.60920703, + "learning_rate": 2.3503000683225526e-08, + "loss": 0.63114184, + "num_input_tokens_seen": 171121615, + "step": 7922, + "time_per_iteration": 2.7420654296875 + }, + { + "auxiliary_loss_clip": 0.01166267, + "auxiliary_loss_mlp": 0.010234, + "balance_loss_clip": 1.04578447, + "balance_loss_mlp": 1.01560032, + "epoch": 0.9526844225335177, + "flos": 16727083251840.0, + "grad_norm": 2.112624304987332, + "language_loss": 0.84461355, + "learning_rate": 2.3384080249745585e-08, + "loss": 0.86651021, + "num_input_tokens_seen": 171139505, + "step": 7923, + "time_per_iteration": 2.5985491275787354 + }, + { + "auxiliary_loss_clip": 0.01206146, + "auxiliary_loss_mlp": 0.01022682, + "balance_loss_clip": 1.04188633, + "balance_loss_mlp": 1.01552081, + "epoch": 0.9528046654241568, + "flos": 36939367330560.0, + "grad_norm": 2.124288211496096, + "language_loss": 0.83124322, + "learning_rate": 2.3265459667329178e-08, + "loss": 0.85353148, + "num_input_tokens_seen": 171158995, + "step": 7924, + "time_per_iteration": 2.8664088249206543 + }, + { + "auxiliary_loss_clip": 0.01187157, + "auxiliary_loss_mlp": 0.0102111, + "balance_loss_clip": 1.04687905, + "balance_loss_mlp": 1.0139451, + "epoch": 0.9529249083147959, + "flos": 18255010032000.0, + "grad_norm": 2.2897573403769496, + "language_loss": 0.86652762, + "learning_rate": 2.31471389539708e-08, + "loss": 0.88861018, + "num_input_tokens_seen": 171176120, + "step": 7925, + "time_per_iteration": 2.598381996154785 + }, + { + "auxiliary_loss_clip": 0.01178266, + "auxiliary_loss_mlp": 0.01378338, + "balance_loss_clip": 1.04722333, + "balance_loss_mlp": 1.00008631, + "epoch": 0.953045151205435, + "flos": 28658438985600.0, + "grad_norm": 1.805648634160502, + "language_loss": 0.72871447, + "learning_rate": 2.3029118127619872e-08, + "loss": 0.75428045, + "num_input_tokens_seen": 171195835, + "step": 7926, + "time_per_iteration": 2.704482078552246 + }, + { + "auxiliary_loss_clip": 0.01174494, + "auxiliary_loss_mlp": 0.01021902, + "balance_loss_clip": 1.04315889, + "balance_loss_mlp": 1.01465404, + "epoch": 0.953165394096074, + "flos": 21835232288640.0, + "grad_norm": 2.1204245391254526, + "language_loss": 0.87010956, + "learning_rate": 2.2911397206179628e-08, + "loss": 0.89207351, + "num_input_tokens_seen": 171212585, + "step": 7927, + "time_per_iteration": 2.6242263317108154 + }, + { + "auxiliary_loss_clip": 0.01166774, + "auxiliary_loss_mlp": 0.01023, + "balance_loss_clip": 1.04768527, + "balance_loss_mlp": 1.01635456, + "epoch": 0.9532856369867132, + "flos": 19975059682560.0, + "grad_norm": 1.865264207609896, + "language_loss": 0.63080728, + "learning_rate": 2.279397620750845e-08, + "loss": 0.65270507, + "num_input_tokens_seen": 171231630, + "step": 7928, + "time_per_iteration": 2.6109635829925537 + }, + { + "auxiliary_loss_clip": 0.01177246, + "auxiliary_loss_mlp": 0.01026754, + "balance_loss_clip": 1.04212976, + "balance_loss_mlp": 1.01973569, + "epoch": 0.9534058798773523, + "flos": 15049588239360.0, + "grad_norm": 2.3206545000053347, + "language_loss": 0.78829408, + "learning_rate": 2.2676855149419195e-08, + "loss": 0.81033409, + "num_input_tokens_seen": 171248800, + "step": 7929, + "time_per_iteration": 3.5679969787597656 + }, + { + "auxiliary_loss_clip": 0.01181445, + "auxiliary_loss_mlp": 0.01026074, + "balance_loss_clip": 1.04653251, + "balance_loss_mlp": 1.01943755, + "epoch": 0.9535261227679913, + "flos": 17602800831360.0, + "grad_norm": 2.24137995434547, + "language_loss": 0.75228763, + "learning_rate": 2.2560034049678988e-08, + "loss": 0.77436286, + "num_input_tokens_seen": 171263150, + "step": 7930, + "time_per_iteration": 2.6088168621063232 + }, + { + "auxiliary_loss_clip": 0.01169582, + "auxiliary_loss_mlp": 0.01018684, + "balance_loss_clip": 1.04799151, + "balance_loss_mlp": 1.0116446, + "epoch": 0.9536463656586305, + "flos": 23142954741120.0, + "grad_norm": 2.684946224409654, + "language_loss": 0.75282669, + "learning_rate": 2.2443512926008988e-08, + "loss": 0.77470934, + "num_input_tokens_seen": 171282480, + "step": 7931, + "time_per_iteration": 2.6887295246124268 + }, + { + "auxiliary_loss_clip": 0.01194222, + "auxiliary_loss_mlp": 0.01026597, + "balance_loss_clip": 1.04170465, + "balance_loss_mlp": 1.01988232, + "epoch": 0.9537666085492695, + "flos": 18625033987200.0, + "grad_norm": 2.357433373916841, + "language_loss": 0.70016837, + "learning_rate": 2.2327291796085946e-08, + "loss": 0.72237659, + "num_input_tokens_seen": 171300840, + "step": 7932, + "time_per_iteration": 3.6372971534729004 + }, + { + "auxiliary_loss_clip": 0.0116797, + "auxiliary_loss_mlp": 0.01025314, + "balance_loss_clip": 1.04649627, + "balance_loss_mlp": 1.01738071, + "epoch": 0.9538868514399086, + "flos": 18989347680000.0, + "grad_norm": 2.7067873248488414, + "language_loss": 0.77292651, + "learning_rate": 2.2211370677540197e-08, + "loss": 0.79485941, + "num_input_tokens_seen": 171317365, + "step": 7933, + "time_per_iteration": 2.5712270736694336 + }, + { + "auxiliary_loss_clip": 0.01168674, + "auxiliary_loss_mlp": 0.01023498, + "balance_loss_clip": 1.04704535, + "balance_loss_mlp": 1.01671839, + "epoch": 0.9540070943305478, + "flos": 16800556521600.0, + "grad_norm": 2.739432802084937, + "language_loss": 0.78393495, + "learning_rate": 2.2095749587957012e-08, + "loss": 0.8058567, + "num_input_tokens_seen": 171335270, + "step": 7934, + "time_per_iteration": 3.4707016944885254 + }, + { + "auxiliary_loss_clip": 0.01179893, + "auxiliary_loss_mlp": 0.01021813, + "balance_loss_clip": 1.04302108, + "balance_loss_mlp": 1.01393032, + "epoch": 0.9541273372211868, + "flos": 20156911263360.0, + "grad_norm": 10.869194331429066, + "language_loss": 0.69731098, + "learning_rate": 2.1980428544876138e-08, + "loss": 0.71932799, + "num_input_tokens_seen": 171353910, + "step": 7935, + "time_per_iteration": 2.615004301071167 + }, + { + "auxiliary_loss_clip": 0.01189164, + "auxiliary_loss_mlp": 0.0102524, + "balance_loss_clip": 1.03612399, + "balance_loss_mlp": 1.01780105, + "epoch": 0.9542475801118259, + "flos": 26725511381760.0, + "grad_norm": 1.534552934669465, + "language_loss": 0.74314547, + "learning_rate": 2.1865407565791584e-08, + "loss": 0.7652896, + "num_input_tokens_seen": 171375480, + "step": 7936, + "time_per_iteration": 2.74215030670166 + }, + { + "auxiliary_loss_clip": 0.01183211, + "auxiliary_loss_mlp": 0.01020457, + "balance_loss_clip": 1.0412457, + "balance_loss_mlp": 1.01297939, + "epoch": 0.954367823002465, + "flos": 23330911633920.0, + "grad_norm": 2.133683555307259, + "language_loss": 0.7707094, + "learning_rate": 2.175068666815183e-08, + "loss": 0.79274607, + "num_input_tokens_seen": 171396320, + "step": 7937, + "time_per_iteration": 2.6172852516174316 + }, + { + "auxiliary_loss_clip": 0.01192412, + "auxiliary_loss_mlp": 0.01027366, + "balance_loss_clip": 1.0429287, + "balance_loss_mlp": 1.02021933, + "epoch": 0.9544880658931041, + "flos": 14902713527040.0, + "grad_norm": 2.800638775609954, + "language_loss": 0.7881943, + "learning_rate": 2.163626586935985e-08, + "loss": 0.81039202, + "num_input_tokens_seen": 171412860, + "step": 7938, + "time_per_iteration": 3.500763177871704 + }, + { + "auxiliary_loss_clip": 0.01171925, + "auxiliary_loss_mlp": 0.01024477, + "balance_loss_clip": 1.04366612, + "balance_loss_mlp": 1.01683021, + "epoch": 0.9546083087837431, + "flos": 29095902725760.0, + "grad_norm": 2.8015335580199276, + "language_loss": 0.63247216, + "learning_rate": 2.1522145186773755e-08, + "loss": 0.65443623, + "num_input_tokens_seen": 171431780, + "step": 7939, + "time_per_iteration": 2.639981269836426 + }, + { + "auxiliary_loss_clip": 0.01186265, + "auxiliary_loss_mlp": 0.01024282, + "balance_loss_clip": 1.04555511, + "balance_loss_mlp": 1.01754093, + "epoch": 0.9547285516743822, + "flos": 21142335957120.0, + "grad_norm": 1.7231165338694647, + "language_loss": 0.85329401, + "learning_rate": 2.140832463770481e-08, + "loss": 0.87539947, + "num_input_tokens_seen": 171450975, + "step": 7940, + "time_per_iteration": 2.702608346939087 + }, + { + "auxiliary_loss_clip": 0.01189223, + "auxiliary_loss_mlp": 0.01027182, + "balance_loss_clip": 1.04424977, + "balance_loss_mlp": 1.01989269, + "epoch": 0.9548487945650214, + "flos": 27490157130240.0, + "grad_norm": 1.9094283624308843, + "language_loss": 0.75804007, + "learning_rate": 2.129480423941987e-08, + "loss": 0.78020406, + "num_input_tokens_seen": 171467645, + "step": 7941, + "time_per_iteration": 2.699990749359131 + }, + { + "auxiliary_loss_clip": 0.01188136, + "auxiliary_loss_mlp": 0.0102277, + "balance_loss_clip": 1.04518092, + "balance_loss_mlp": 1.01590335, + "epoch": 0.9549690374556604, + "flos": 22273198819200.0, + "grad_norm": 1.6064018638406603, + "language_loss": 0.801404, + "learning_rate": 2.1181584009140052e-08, + "loss": 0.82351309, + "num_input_tokens_seen": 171487185, + "step": 7942, + "time_per_iteration": 2.627187967300415 + }, + { + "auxiliary_loss_clip": 0.01198988, + "auxiliary_loss_mlp": 0.01024917, + "balance_loss_clip": 1.04154634, + "balance_loss_mlp": 1.01829195, + "epoch": 0.9550892803462995, + "flos": 17595294888960.0, + "grad_norm": 2.121505274361009, + "language_loss": 0.8393482, + "learning_rate": 2.10686639640405e-08, + "loss": 0.86158723, + "num_input_tokens_seen": 171501275, + "step": 7943, + "time_per_iteration": 2.6289148330688477 + }, + { + "auxiliary_loss_clip": 0.01179075, + "auxiliary_loss_mlp": 0.01024606, + "balance_loss_clip": 1.04525578, + "balance_loss_mlp": 1.01801109, + "epoch": 0.9552095232369386, + "flos": 24353144789760.0, + "grad_norm": 1.6386625722232246, + "language_loss": 0.81392586, + "learning_rate": 2.0956044121251294e-08, + "loss": 0.83596265, + "num_input_tokens_seen": 171520060, + "step": 7944, + "time_per_iteration": 2.6942026615142822 + }, + { + "auxiliary_loss_clip": 0.01191439, + "auxiliary_loss_mlp": 0.01025228, + "balance_loss_clip": 1.04336953, + "balance_loss_mlp": 1.01799774, + "epoch": 0.9553297661275777, + "flos": 22746860490240.0, + "grad_norm": 1.9401178358957543, + "language_loss": 0.80956399, + "learning_rate": 2.084372449785654e-08, + "loss": 0.83173066, + "num_input_tokens_seen": 171539895, + "step": 7945, + "time_per_iteration": 2.792404890060425 + }, + { + "auxiliary_loss_clip": 0.01177548, + "auxiliary_loss_mlp": 0.01027114, + "balance_loss_clip": 1.04171133, + "balance_loss_mlp": 1.02005124, + "epoch": 0.9554500090182168, + "flos": 15413866018560.0, + "grad_norm": 1.9768384059277941, + "language_loss": 0.68612003, + "learning_rate": 2.0731705110895282e-08, + "loss": 0.70816666, + "num_input_tokens_seen": 171557385, + "step": 7946, + "time_per_iteration": 2.5832293033599854 + }, + { + "auxiliary_loss_clip": 0.01179295, + "auxiliary_loss_mlp": 0.01028755, + "balance_loss_clip": 1.04811502, + "balance_loss_mlp": 1.02098823, + "epoch": 0.9555702519088559, + "flos": 23513517400320.0, + "grad_norm": 3.6036723018910424, + "language_loss": 0.86693692, + "learning_rate": 2.0619985977360587e-08, + "loss": 0.88901746, + "num_input_tokens_seen": 171575705, + "step": 7947, + "time_per_iteration": 2.592747211456299 + }, + { + "auxiliary_loss_clip": 0.01199763, + "auxiliary_loss_mlp": 0.01021835, + "balance_loss_clip": 1.03816128, + "balance_loss_mlp": 1.01492095, + "epoch": 0.955690494799495, + "flos": 22962072827520.0, + "grad_norm": 1.920917230183842, + "language_loss": 0.76796669, + "learning_rate": 2.0508567114200237e-08, + "loss": 0.79018271, + "num_input_tokens_seen": 171595620, + "step": 7948, + "time_per_iteration": 2.6971254348754883 + }, + { + "auxiliary_loss_clip": 0.01188381, + "auxiliary_loss_mlp": 0.01020941, + "balance_loss_clip": 1.04407263, + "balance_loss_mlp": 1.01457822, + "epoch": 0.955810737690134, + "flos": 26031250333440.0, + "grad_norm": 32.504262147788864, + "language_loss": 0.78580374, + "learning_rate": 2.0397448538316485e-08, + "loss": 0.80789697, + "num_input_tokens_seen": 171616660, + "step": 7949, + "time_per_iteration": 2.690518617630005 + }, + { + "auxiliary_loss_clip": 0.01187083, + "auxiliary_loss_mlp": 0.01024395, + "balance_loss_clip": 1.04219532, + "balance_loss_mlp": 1.01731133, + "epoch": 0.9559309805807732, + "flos": 20849951249280.0, + "grad_norm": 2.4460259473624597, + "language_loss": 0.67184907, + "learning_rate": 2.028663026656563e-08, + "loss": 0.69396389, + "num_input_tokens_seen": 171635515, + "step": 7950, + "time_per_iteration": 2.641770362854004 + }, + { + "auxiliary_loss_clip": 0.01164759, + "auxiliary_loss_mlp": 0.01378451, + "balance_loss_clip": 1.04717565, + "balance_loss_mlp": 1.00014019, + "epoch": 0.9560512234714122, + "flos": 21578219498880.0, + "grad_norm": 2.6812069022517266, + "language_loss": 0.71852762, + "learning_rate": 2.0176112315758885e-08, + "loss": 0.74395972, + "num_input_tokens_seen": 171653305, + "step": 7951, + "time_per_iteration": 2.598801851272583 + }, + { + "auxiliary_loss_clip": 0.01208017, + "auxiliary_loss_mlp": 0.01024832, + "balance_loss_clip": 1.04124236, + "balance_loss_mlp": 1.01767349, + "epoch": 0.9561714663620513, + "flos": 17450144029440.0, + "grad_norm": 2.236896025172401, + "language_loss": 0.68924284, + "learning_rate": 2.0065894702661957e-08, + "loss": 0.71157134, + "num_input_tokens_seen": 171669980, + "step": 7952, + "time_per_iteration": 2.6499736309051514 + }, + { + "auxiliary_loss_clip": 0.01184023, + "auxiliary_loss_mlp": 0.01378627, + "balance_loss_clip": 1.04159486, + "balance_loss_mlp": 1.00011384, + "epoch": 0.9562917092526905, + "flos": 26098510550400.0, + "grad_norm": 1.635288953180144, + "language_loss": 0.78194118, + "learning_rate": 1.9955977443994577e-08, + "loss": 0.80756772, + "num_input_tokens_seen": 171689970, + "step": 7953, + "time_per_iteration": 2.7349345684051514 + }, + { + "auxiliary_loss_clip": 0.01181629, + "auxiliary_loss_mlp": 0.01030317, + "balance_loss_clip": 1.04342484, + "balance_loss_mlp": 1.02156091, + "epoch": 0.9564119521433295, + "flos": 24096742531200.0, + "grad_norm": 2.069260241105606, + "language_loss": 0.6241405, + "learning_rate": 1.9846360556430965e-08, + "loss": 0.64626002, + "num_input_tokens_seen": 171708270, + "step": 7954, + "time_per_iteration": 2.6562232971191406 + }, + { + "auxiliary_loss_clip": 0.01164665, + "auxiliary_loss_mlp": 0.01018506, + "balance_loss_clip": 1.04578698, + "balance_loss_mlp": 1.01177406, + "epoch": 0.9565321950339686, + "flos": 32008903896960.0, + "grad_norm": 2.3081112001327475, + "language_loss": 0.61576933, + "learning_rate": 1.973704405660004e-08, + "loss": 0.63760102, + "num_input_tokens_seen": 171729385, + "step": 7955, + "time_per_iteration": 3.5475876331329346 + }, + { + "auxiliary_loss_clip": 0.0120823, + "auxiliary_loss_mlp": 0.01025255, + "balance_loss_clip": 1.03919005, + "balance_loss_mlp": 1.01826072, + "epoch": 0.9566524379246077, + "flos": 23588642695680.0, + "grad_norm": 2.3269000185168616, + "language_loss": 0.78171295, + "learning_rate": 1.9628027961085203e-08, + "loss": 0.80404776, + "num_input_tokens_seen": 171752615, + "step": 7956, + "time_per_iteration": 2.796542167663574 + }, + { + "auxiliary_loss_clip": 0.01177633, + "auxiliary_loss_mlp": 0.01028791, + "balance_loss_clip": 1.03831935, + "balance_loss_mlp": 1.02182055, + "epoch": 0.9567726808152468, + "flos": 38067716240640.0, + "grad_norm": 1.7088599344754167, + "language_loss": 0.84021121, + "learning_rate": 1.9519312286423894e-08, + "loss": 0.86227548, + "num_input_tokens_seen": 171775810, + "step": 7957, + "time_per_iteration": 2.8080971240997314 + }, + { + "auxiliary_loss_clip": 0.0117437, + "auxiliary_loss_mlp": 0.01024979, + "balance_loss_clip": 1.04609823, + "balance_loss_mlp": 1.01737654, + "epoch": 0.9568929237058859, + "flos": 22744059229440.0, + "grad_norm": 1.7291238247297622, + "language_loss": 0.77744555, + "learning_rate": 1.9410897049108255e-08, + "loss": 0.79943907, + "num_input_tokens_seen": 171795090, + "step": 7958, + "time_per_iteration": 3.518179416656494 + }, + { + "auxiliary_loss_clip": 0.01172652, + "auxiliary_loss_mlp": 0.01029752, + "balance_loss_clip": 1.04956818, + "balance_loss_mlp": 1.02164245, + "epoch": 0.957013166596525, + "flos": 23841633162240.0, + "grad_norm": 2.735553167501594, + "language_loss": 0.91459513, + "learning_rate": 1.9302782265584905e-08, + "loss": 0.93661916, + "num_input_tokens_seen": 171815755, + "step": 7959, + "time_per_iteration": 2.5864739418029785 + }, + { + "auxiliary_loss_clip": 0.01187782, + "auxiliary_loss_mlp": 0.01024535, + "balance_loss_clip": 1.03961682, + "balance_loss_mlp": 1.01748633, + "epoch": 0.9571334094871641, + "flos": 17639286071040.0, + "grad_norm": 2.1353422920186906, + "language_loss": 0.86840093, + "learning_rate": 1.9194967952254282e-08, + "loss": 0.89052403, + "num_input_tokens_seen": 171834330, + "step": 7960, + "time_per_iteration": 3.6016504764556885 + }, + { + "auxiliary_loss_clip": 0.01176693, + "auxiliary_loss_mlp": 0.01026586, + "balance_loss_clip": 1.04736209, + "balance_loss_mlp": 1.01875961, + "epoch": 0.9572536523778031, + "flos": 15369623441280.0, + "grad_norm": 2.3679924030860664, + "language_loss": 0.8088035, + "learning_rate": 1.9087454125472635e-08, + "loss": 0.8308363, + "num_input_tokens_seen": 171848805, + "step": 7961, + "time_per_iteration": 2.596151113510132 + }, + { + "auxiliary_loss_clip": 0.01166427, + "auxiliary_loss_mlp": 0.01029128, + "balance_loss_clip": 1.04644156, + "balance_loss_mlp": 1.02202618, + "epoch": 0.9573738952684423, + "flos": 24969838417920.0, + "grad_norm": 2.7842869159659953, + "language_loss": 0.78562105, + "learning_rate": 1.8980240801548696e-08, + "loss": 0.8075766, + "num_input_tokens_seen": 171867995, + "step": 7962, + "time_per_iteration": 2.5639843940734863 + }, + { + "auxiliary_loss_clip": 0.01181695, + "auxiliary_loss_mlp": 0.01024451, + "balance_loss_clip": 1.04508519, + "balance_loss_mlp": 1.01772463, + "epoch": 0.9574941381590814, + "flos": 25769461034880.0, + "grad_norm": 1.9109906993139985, + "language_loss": 0.74088091, + "learning_rate": 1.8873327996747458e-08, + "loss": 0.76294237, + "num_input_tokens_seen": 171886495, + "step": 7963, + "time_per_iteration": 2.741687059402466 + }, + { + "auxiliary_loss_clip": 0.01176213, + "auxiliary_loss_mlp": 0.01023445, + "balance_loss_clip": 1.04414535, + "balance_loss_mlp": 1.01629257, + "epoch": 0.9576143810497204, + "flos": 32307178435200.0, + "grad_norm": 1.9979819184025092, + "language_loss": 0.66068935, + "learning_rate": 1.8766715727287053e-08, + "loss": 0.68268597, + "num_input_tokens_seen": 171908200, + "step": 7964, + "time_per_iteration": 2.6596953868865967 + }, + { + "auxiliary_loss_clip": 0.01179767, + "auxiliary_loss_mlp": 0.01378601, + "balance_loss_clip": 1.04463542, + "balance_loss_mlp": 1.00022161, + "epoch": 0.9577346239403596, + "flos": 27745733376000.0, + "grad_norm": 1.789019689959651, + "language_loss": 0.79288733, + "learning_rate": 1.8660404009340546e-08, + "loss": 0.81847101, + "num_input_tokens_seen": 171928650, + "step": 7965, + "time_per_iteration": 3.554100751876831 + }, + { + "auxiliary_loss_clip": 0.01073182, + "auxiliary_loss_mlp": 0.01000466, + "balance_loss_clip": 1.00832283, + "balance_loss_mlp": 0.99944633, + "epoch": 0.9578548668309986, + "flos": 57468313710720.0, + "grad_norm": 0.8703765387373714, + "language_loss": 0.59523529, + "learning_rate": 1.8554392859035485e-08, + "loss": 0.61597174, + "num_input_tokens_seen": 171986400, + "step": 7966, + "time_per_iteration": 3.2047946453094482 + }, + { + "auxiliary_loss_clip": 0.01222616, + "auxiliary_loss_mlp": 0.01022011, + "balance_loss_clip": 1.03833699, + "balance_loss_mlp": 1.01491535, + "epoch": 0.9579751097216377, + "flos": 19756040503680.0, + "grad_norm": 9.764687970915283, + "language_loss": 0.7900005, + "learning_rate": 1.8448682292453444e-08, + "loss": 0.81244677, + "num_input_tokens_seen": 172005475, + "step": 7967, + "time_per_iteration": 2.819157600402832 + }, + { + "auxiliary_loss_clip": 0.01167166, + "auxiliary_loss_mlp": 0.01026816, + "balance_loss_clip": 1.047001, + "balance_loss_mlp": 1.01938033, + "epoch": 0.9580953526122769, + "flos": 18041270152320.0, + "grad_norm": 1.7555567727162855, + "language_loss": 0.66609293, + "learning_rate": 1.8343272325631154e-08, + "loss": 0.68803275, + "num_input_tokens_seen": 172024420, + "step": 7968, + "time_per_iteration": 2.5797340869903564 + }, + { + "auxiliary_loss_clip": 0.01223108, + "auxiliary_loss_mlp": 0.01379018, + "balance_loss_clip": 1.03883469, + "balance_loss_mlp": 1.00021374, + "epoch": 0.9582155955029159, + "flos": 24270154416000.0, + "grad_norm": 2.416901734359382, + "language_loss": 0.77931416, + "learning_rate": 1.8238162974558492e-08, + "loss": 0.8053354, + "num_input_tokens_seen": 172038350, + "step": 7969, + "time_per_iteration": 2.757838726043701 + }, + { + "auxiliary_loss_clip": 0.01185145, + "auxiliary_loss_mlp": 0.01030705, + "balance_loss_clip": 1.04726672, + "balance_loss_mlp": 1.02374005, + "epoch": 0.958335838393555, + "flos": 22783309816320.0, + "grad_norm": 1.905719358962909, + "language_loss": 0.74962914, + "learning_rate": 1.8133354255181144e-08, + "loss": 0.77178764, + "num_input_tokens_seen": 172058665, + "step": 7970, + "time_per_iteration": 2.7022156715393066 + }, + { + "auxiliary_loss_clip": 0.01168205, + "auxiliary_loss_mlp": 0.01030433, + "balance_loss_clip": 1.04256868, + "balance_loss_mlp": 1.02322972, + "epoch": 0.958456081284194, + "flos": 16911484698240.0, + "grad_norm": 2.161242907746701, + "language_loss": 0.74370235, + "learning_rate": 1.802884618339795e-08, + "loss": 0.76568878, + "num_input_tokens_seen": 172077470, + "step": 7971, + "time_per_iteration": 2.579522132873535 + }, + { + "auxiliary_loss_clip": 0.01179549, + "auxiliary_loss_mlp": 0.01026502, + "balance_loss_clip": 1.04869771, + "balance_loss_mlp": 1.0189327, + "epoch": 0.9585763241748332, + "flos": 19974951941760.0, + "grad_norm": 1.9721173011927364, + "language_loss": 0.8083123, + "learning_rate": 1.7924638775062894e-08, + "loss": 0.83037281, + "num_input_tokens_seen": 172096590, + "step": 7972, + "time_per_iteration": 2.6081061363220215 + }, + { + "auxiliary_loss_clip": 0.01189058, + "auxiliary_loss_mlp": 0.01020774, + "balance_loss_clip": 1.045645, + "balance_loss_mlp": 1.01417542, + "epoch": 0.9586965670654722, + "flos": 21395649646080.0, + "grad_norm": 1.847834612150888, + "language_loss": 0.8152473, + "learning_rate": 1.7820732045984444e-08, + "loss": 0.8373456, + "num_input_tokens_seen": 172116735, + "step": 7973, + "time_per_iteration": 2.6868276596069336 + }, + { + "auxiliary_loss_clip": 0.01170116, + "auxiliary_loss_mlp": 0.01026038, + "balance_loss_clip": 1.04430807, + "balance_loss_mlp": 1.0184269, + "epoch": 0.9588168099561113, + "flos": 21435115714560.0, + "grad_norm": 1.7345031842177931, + "language_loss": 0.73813641, + "learning_rate": 1.7717126011924655e-08, + "loss": 0.76009798, + "num_input_tokens_seen": 172138320, + "step": 7974, + "time_per_iteration": 2.644552230834961 + }, + { + "auxiliary_loss_clip": 0.01192557, + "auxiliary_loss_mlp": 0.01031651, + "balance_loss_clip": 1.03762901, + "balance_loss_mlp": 1.02418804, + "epoch": 0.9589370528467505, + "flos": 11763761852160.0, + "grad_norm": 3.1447908641959232, + "language_loss": 0.76434922, + "learning_rate": 1.7613820688600957e-08, + "loss": 0.78659129, + "num_input_tokens_seen": 172154225, + "step": 7975, + "time_per_iteration": 2.7012951374053955 + }, + { + "auxiliary_loss_clip": 0.01190951, + "auxiliary_loss_mlp": 0.01028145, + "balance_loss_clip": 1.04446816, + "balance_loss_mlp": 1.02075982, + "epoch": 0.9590572957373895, + "flos": 23441516588160.0, + "grad_norm": 1.7445014109067596, + "language_loss": 0.78420413, + "learning_rate": 1.7510816091684588e-08, + "loss": 0.80639511, + "num_input_tokens_seen": 172174150, + "step": 7976, + "time_per_iteration": 2.70542573928833 + }, + { + "auxiliary_loss_clip": 0.01185673, + "auxiliary_loss_mlp": 0.01024303, + "balance_loss_clip": 1.04630876, + "balance_loss_mlp": 1.01639378, + "epoch": 0.9591775386280286, + "flos": 22528272274560.0, + "grad_norm": 2.21027226672016, + "language_loss": 0.7859515, + "learning_rate": 1.740811223680083e-08, + "loss": 0.80805123, + "num_input_tokens_seen": 172191005, + "step": 7977, + "time_per_iteration": 2.631293773651123 + }, + { + "auxiliary_loss_clip": 0.01167764, + "auxiliary_loss_mlp": 0.0102372, + "balance_loss_clip": 1.04874182, + "balance_loss_mlp": 1.01669288, + "epoch": 0.9592977815186677, + "flos": 18186959715840.0, + "grad_norm": 2.3541318140493894, + "language_loss": 0.74042708, + "learning_rate": 1.7305709139530334e-08, + "loss": 0.76234186, + "num_input_tokens_seen": 172209785, + "step": 7978, + "time_per_iteration": 2.6371102333068848 + }, + { + "auxiliary_loss_clip": 0.01166885, + "auxiliary_loss_mlp": 0.0102435, + "balance_loss_clip": 1.04319441, + "balance_loss_mlp": 1.01647651, + "epoch": 0.9594180244093068, + "flos": 16537797555840.0, + "grad_norm": 2.980997690106469, + "language_loss": 0.74961096, + "learning_rate": 1.7203606815407334e-08, + "loss": 0.77152336, + "num_input_tokens_seen": 172224380, + "step": 7979, + "time_per_iteration": 2.745439052581787 + }, + { + "auxiliary_loss_clip": 0.01189506, + "auxiliary_loss_mlp": 0.01027447, + "balance_loss_clip": 1.04611218, + "balance_loss_mlp": 1.02011299, + "epoch": 0.9595382672999458, + "flos": 20554334317440.0, + "grad_norm": 1.8681823246310985, + "language_loss": 0.79447663, + "learning_rate": 1.7101805279920557e-08, + "loss": 0.8166461, + "num_input_tokens_seen": 172242540, + "step": 7980, + "time_per_iteration": 2.6220850944519043 + }, + { + "auxiliary_loss_clip": 0.01168712, + "auxiliary_loss_mlp": 0.01024901, + "balance_loss_clip": 1.04850888, + "balance_loss_mlp": 1.01745677, + "epoch": 0.959658510190585, + "flos": 22638266697600.0, + "grad_norm": 1.9216840332540366, + "language_loss": 0.80853933, + "learning_rate": 1.7000304548513643e-08, + "loss": 0.83047545, + "num_input_tokens_seen": 172262645, + "step": 7981, + "time_per_iteration": 3.5196900367736816 + }, + { + "auxiliary_loss_clip": 0.01191521, + "auxiliary_loss_mlp": 0.01026723, + "balance_loss_clip": 1.04233193, + "balance_loss_mlp": 1.01857543, + "epoch": 0.9597787530812241, + "flos": 19135252725120.0, + "grad_norm": 1.9850649253149772, + "language_loss": 0.82707655, + "learning_rate": 1.6899104636583394e-08, + "loss": 0.84925902, + "num_input_tokens_seen": 172280695, + "step": 7982, + "time_per_iteration": 2.6496903896331787 + }, + { + "auxiliary_loss_clip": 0.01073714, + "auxiliary_loss_mlp": 0.0100158, + "balance_loss_clip": 1.00862551, + "balance_loss_mlp": 1.00050092, + "epoch": 0.9598989959718631, + "flos": 60098124055680.0, + "grad_norm": 0.7309499438775974, + "language_loss": 0.62106109, + "learning_rate": 1.6798205559482638e-08, + "loss": 0.64181399, + "num_input_tokens_seen": 172343075, + "step": 7983, + "time_per_iteration": 3.326014757156372 + }, + { + "auxiliary_loss_clip": 0.01196658, + "auxiliary_loss_mlp": 0.01024084, + "balance_loss_clip": 1.04428363, + "balance_loss_mlp": 1.01673782, + "epoch": 0.9600192388625023, + "flos": 20886795624960.0, + "grad_norm": 1.9571757532130716, + "language_loss": 0.76519746, + "learning_rate": 1.669760733251713e-08, + "loss": 0.78740489, + "num_input_tokens_seen": 172361950, + "step": 7984, + "time_per_iteration": 3.6186366081237793 + }, + { + "auxiliary_loss_clip": 0.01217097, + "auxiliary_loss_mlp": 0.01022361, + "balance_loss_clip": 1.04065275, + "balance_loss_mlp": 1.01566434, + "epoch": 0.9601394817531413, + "flos": 20445740524800.0, + "grad_norm": 1.758561837107692, + "language_loss": 0.82366168, + "learning_rate": 1.659730997094755e-08, + "loss": 0.84605628, + "num_input_tokens_seen": 172380440, + "step": 7985, + "time_per_iteration": 2.7133355140686035 + }, + { + "auxiliary_loss_clip": 0.01166418, + "auxiliary_loss_mlp": 0.01023172, + "balance_loss_clip": 1.0435077, + "balance_loss_mlp": 1.01601362, + "epoch": 0.9602597246437804, + "flos": 21507152440320.0, + "grad_norm": 1.8604406877256832, + "language_loss": 0.61960578, + "learning_rate": 1.6497313489989283e-08, + "loss": 0.64150167, + "num_input_tokens_seen": 172400265, + "step": 7986, + "time_per_iteration": 3.5773725509643555 + }, + { + "auxiliary_loss_clip": 0.01194703, + "auxiliary_loss_mlp": 0.01025498, + "balance_loss_clip": 1.03507185, + "balance_loss_mlp": 1.01803803, + "epoch": 0.9603799675344196, + "flos": 29935099152000.0, + "grad_norm": 2.1279821366739577, + "language_loss": 0.69570136, + "learning_rate": 1.639761790481131e-08, + "loss": 0.71790338, + "num_input_tokens_seen": 172421145, + "step": 7987, + "time_per_iteration": 2.7609593868255615 + }, + { + "auxiliary_loss_clip": 0.01180099, + "auxiliary_loss_mlp": 0.01030747, + "balance_loss_clip": 1.047791, + "balance_loss_mlp": 1.02359486, + "epoch": 0.9605002104250586, + "flos": 28001525103360.0, + "grad_norm": 1.9534078228340435, + "language_loss": 0.79189855, + "learning_rate": 1.6298223230537754e-08, + "loss": 0.81400704, + "num_input_tokens_seen": 172438945, + "step": 7988, + "time_per_iteration": 2.6909539699554443 + }, + { + "auxiliary_loss_clip": 0.01185582, + "auxiliary_loss_mlp": 0.01378995, + "balance_loss_clip": 1.0460453, + "balance_loss_mlp": 1.00018764, + "epoch": 0.9606204533156977, + "flos": 35590490870400.0, + "grad_norm": 2.1363359350126485, + "language_loss": 0.69747186, + "learning_rate": 1.619912948224611e-08, + "loss": 0.72311759, + "num_input_tokens_seen": 172460150, + "step": 7989, + "time_per_iteration": 2.8254292011260986 + }, + { + "auxiliary_loss_clip": 0.01186037, + "auxiliary_loss_mlp": 0.01030406, + "balance_loss_clip": 1.04261637, + "balance_loss_mlp": 1.02223992, + "epoch": 0.9607406962063368, + "flos": 26574614346240.0, + "grad_norm": 2.442258279845257, + "language_loss": 0.60934281, + "learning_rate": 1.6100336674969682e-08, + "loss": 0.63150728, + "num_input_tokens_seen": 172478990, + "step": 7990, + "time_per_iteration": 3.630671977996826 + }, + { + "auxiliary_loss_clip": 0.01204412, + "auxiliary_loss_mlp": 0.01027862, + "balance_loss_clip": 1.04039407, + "balance_loss_mlp": 1.02002716, + "epoch": 0.9608609390969759, + "flos": 25331781813120.0, + "grad_norm": 1.7455468638683846, + "language_loss": 0.76369154, + "learning_rate": 1.600184482369449e-08, + "loss": 0.78601432, + "num_input_tokens_seen": 172498905, + "step": 7991, + "time_per_iteration": 2.7898263931274414 + }, + { + "auxiliary_loss_clip": 0.01196819, + "auxiliary_loss_mlp": 0.01024731, + "balance_loss_clip": 1.04301429, + "balance_loss_mlp": 1.01621914, + "epoch": 0.960981181987615, + "flos": 21069114082560.0, + "grad_norm": 2.1027472696656204, + "language_loss": 0.89194286, + "learning_rate": 1.5903653943362126e-08, + "loss": 0.91415834, + "num_input_tokens_seen": 172517900, + "step": 7992, + "time_per_iteration": 2.6638236045837402 + }, + { + "auxiliary_loss_clip": 0.01187518, + "auxiliary_loss_mlp": 0.01024912, + "balance_loss_clip": 1.04586697, + "balance_loss_mlp": 1.01778018, + "epoch": 0.9611014248782541, + "flos": 17823256554240.0, + "grad_norm": 1.9473473366138725, + "language_loss": 0.76794779, + "learning_rate": 1.580576404886802e-08, + "loss": 0.79007208, + "num_input_tokens_seen": 172536430, + "step": 7993, + "time_per_iteration": 2.6770193576812744 + }, + { + "auxiliary_loss_clip": 0.01175956, + "auxiliary_loss_mlp": 0.01024165, + "balance_loss_clip": 1.04597461, + "balance_loss_mlp": 1.01744163, + "epoch": 0.9612216677688932, + "flos": 19354631040000.0, + "grad_norm": 2.0981439497244345, + "language_loss": 0.79685903, + "learning_rate": 1.570817515506162e-08, + "loss": 0.81886023, + "num_input_tokens_seen": 172555120, + "step": 7994, + "time_per_iteration": 2.605721950531006 + }, + { + "auxiliary_loss_clip": 0.01167485, + "auxiliary_loss_mlp": 0.01023364, + "balance_loss_clip": 1.04881108, + "balance_loss_mlp": 1.01627755, + "epoch": 0.9613419106595322, + "flos": 15808739207040.0, + "grad_norm": 2.458804813565086, + "language_loss": 0.81742924, + "learning_rate": 1.561088727674753e-08, + "loss": 0.83933777, + "num_input_tokens_seen": 172569330, + "step": 7995, + "time_per_iteration": 2.5296075344085693 + }, + { + "auxiliary_loss_clip": 0.01214628, + "auxiliary_loss_mlp": 0.01022378, + "balance_loss_clip": 1.04196239, + "balance_loss_mlp": 1.01483202, + "epoch": 0.9614621535501714, + "flos": 25702488126720.0, + "grad_norm": 2.0541077135349863, + "language_loss": 0.71090031, + "learning_rate": 1.551390042868417e-08, + "loss": 0.73327041, + "num_input_tokens_seen": 172591100, + "step": 7996, + "time_per_iteration": 2.751762866973877 + }, + { + "auxiliary_loss_clip": 0.01176931, + "auxiliary_loss_mlp": 0.01021151, + "balance_loss_clip": 1.04663599, + "balance_loss_mlp": 1.01408792, + "epoch": 0.9615823964408104, + "flos": 17819054663040.0, + "grad_norm": 1.913891030687383, + "language_loss": 0.70822585, + "learning_rate": 1.5417214625584207e-08, + "loss": 0.73020667, + "num_input_tokens_seen": 172608755, + "step": 7997, + "time_per_iteration": 2.585134744644165 + }, + { + "auxiliary_loss_clip": 0.01169506, + "auxiliary_loss_mlp": 0.01031315, + "balance_loss_clip": 1.04422951, + "balance_loss_mlp": 1.02395368, + "epoch": 0.9617026393314495, + "flos": 20190020624640.0, + "grad_norm": 1.6299594516574176, + "language_loss": 0.85243738, + "learning_rate": 1.5320829882114806e-08, + "loss": 0.87444556, + "num_input_tokens_seen": 172626830, + "step": 7998, + "time_per_iteration": 2.635711669921875 + }, + { + "auxiliary_loss_clip": 0.01165431, + "auxiliary_loss_mlp": 0.0102357, + "balance_loss_clip": 1.04481208, + "balance_loss_mlp": 1.01635814, + "epoch": 0.9618228822220887, + "flos": 20267013427200.0, + "grad_norm": 1.920543044086676, + "language_loss": 0.79101807, + "learning_rate": 1.5224746212897378e-08, + "loss": 0.81290805, + "num_input_tokens_seen": 172646125, + "step": 7999, + "time_per_iteration": 2.5438179969787598 + }, + { + "auxiliary_loss_clip": 0.01162676, + "auxiliary_loss_mlp": 0.01031561, + "balance_loss_clip": 1.04549384, + "balance_loss_mlp": 1.02411604, + "epoch": 0.9619431251127277, + "flos": 21031300039680.0, + "grad_norm": 1.6564562558411031, + "language_loss": 0.77238065, + "learning_rate": 1.512896363250804e-08, + "loss": 0.79432303, + "num_input_tokens_seen": 172666235, + "step": 8000, + "time_per_iteration": 2.5871317386627197 + }, + { + "auxiliary_loss_clip": 0.01178752, + "auxiliary_loss_mlp": 0.01024957, + "balance_loss_clip": 1.04556286, + "balance_loss_mlp": 1.01771188, + "epoch": 0.9620633680033668, + "flos": 22382654538240.0, + "grad_norm": 3.77317939122289, + "language_loss": 0.75552726, + "learning_rate": 1.503348215547673e-08, + "loss": 0.77756441, + "num_input_tokens_seen": 172687325, + "step": 8001, + "time_per_iteration": 2.6221468448638916 + }, + { + "auxiliary_loss_clip": 0.01180424, + "auxiliary_loss_mlp": 0.01025368, + "balance_loss_clip": 1.04165912, + "balance_loss_mlp": 1.01789308, + "epoch": 0.962183610894006, + "flos": 18471730740480.0, + "grad_norm": 1.8056863308768285, + "language_loss": 0.80950326, + "learning_rate": 1.4938301796288078e-08, + "loss": 0.83156115, + "num_input_tokens_seen": 172703895, + "step": 8002, + "time_per_iteration": 2.6453654766082764 + }, + { + "auxiliary_loss_clip": 0.01166799, + "auxiliary_loss_mlp": 0.01026196, + "balance_loss_clip": 1.04715991, + "balance_loss_mlp": 1.0190022, + "epoch": 0.962303853784645, + "flos": 18435245500800.0, + "grad_norm": 2.214978168215993, + "language_loss": 0.81837046, + "learning_rate": 1.4843422569380537e-08, + "loss": 0.84030038, + "num_input_tokens_seen": 172720650, + "step": 8003, + "time_per_iteration": 2.5336596965789795 + }, + { + "auxiliary_loss_clip": 0.01200155, + "auxiliary_loss_mlp": 0.01024365, + "balance_loss_clip": 1.03997278, + "balance_loss_mlp": 1.0170157, + "epoch": 0.9624240966752841, + "flos": 26391074826240.0, + "grad_norm": 1.6996596059039746, + "language_loss": 0.8291676, + "learning_rate": 1.4748844489147483e-08, + "loss": 0.85141277, + "num_input_tokens_seen": 172737640, + "step": 8004, + "time_per_iteration": 2.775167465209961 + }, + { + "auxiliary_loss_clip": 0.01181058, + "auxiliary_loss_mlp": 0.01024641, + "balance_loss_clip": 1.04067516, + "balance_loss_mlp": 1.01795673, + "epoch": 0.9625443395659231, + "flos": 14647675985280.0, + "grad_norm": 1.9282012940308844, + "language_loss": 0.70749909, + "learning_rate": 1.4654567569936326e-08, + "loss": 0.72955608, + "num_input_tokens_seen": 172755215, + "step": 8005, + "time_per_iteration": 2.598421096801758 + }, + { + "auxiliary_loss_clip": 0.01195412, + "auxiliary_loss_mlp": 0.0102364, + "balance_loss_clip": 1.04182172, + "balance_loss_mlp": 1.01620722, + "epoch": 0.9626645824565623, + "flos": 18367626147840.0, + "grad_norm": 1.8264450334911357, + "language_loss": 0.83073944, + "learning_rate": 1.456059182604874e-08, + "loss": 0.85293001, + "num_input_tokens_seen": 172774020, + "step": 8006, + "time_per_iteration": 2.646111249923706 + }, + { + "auxiliary_loss_clip": 0.01168792, + "auxiliary_loss_mlp": 0.01023065, + "balance_loss_clip": 1.04964924, + "balance_loss_mlp": 1.01507783, + "epoch": 0.9627848253472013, + "flos": 16580424021120.0, + "grad_norm": 1.9337125889337756, + "language_loss": 0.76542515, + "learning_rate": 1.4466917271740653e-08, + "loss": 0.78734374, + "num_input_tokens_seen": 172792220, + "step": 8007, + "time_per_iteration": 3.465574264526367 + }, + { + "auxiliary_loss_clip": 0.0117944, + "auxiliary_loss_mlp": 0.01024693, + "balance_loss_clip": 1.04328012, + "balance_loss_mlp": 1.01714659, + "epoch": 0.9629050682378404, + "flos": 20886867452160.0, + "grad_norm": 2.0789798772783836, + "language_loss": 0.6763072, + "learning_rate": 1.4373543921222697e-08, + "loss": 0.69834852, + "num_input_tokens_seen": 172811805, + "step": 8008, + "time_per_iteration": 2.661207437515259 + }, + { + "auxiliary_loss_clip": 0.0118021, + "auxiliary_loss_mlp": 0.0101993, + "balance_loss_clip": 1.04289019, + "balance_loss_mlp": 1.01255345, + "epoch": 0.9630253111284796, + "flos": 17019252478080.0, + "grad_norm": 1.7148169814555836, + "language_loss": 0.77901483, + "learning_rate": 1.428047178865932e-08, + "loss": 0.80101621, + "num_input_tokens_seen": 172828595, + "step": 8009, + "time_per_iteration": 2.6031832695007324 + }, + { + "auxiliary_loss_clip": 0.01181455, + "auxiliary_loss_mlp": 0.01032024, + "balance_loss_clip": 1.04071641, + "balance_loss_mlp": 1.02446055, + "epoch": 0.9631455540191186, + "flos": 20338942412160.0, + "grad_norm": 1.951110881688728, + "language_loss": 0.7461859, + "learning_rate": 1.4187700888169451e-08, + "loss": 0.76832074, + "num_input_tokens_seen": 172847770, + "step": 8010, + "time_per_iteration": 3.4985480308532715 + }, + { + "auxiliary_loss_clip": 0.01072944, + "auxiliary_loss_mlp": 0.01001218, + "balance_loss_clip": 1.00929499, + "balance_loss_mlp": 1.00017452, + "epoch": 0.9632657969097577, + "flos": 65956700033280.0, + "grad_norm": 0.7527100844620331, + "language_loss": 0.57009864, + "learning_rate": 1.40952312338265e-08, + "loss": 0.59084022, + "num_input_tokens_seen": 172912415, + "step": 8011, + "time_per_iteration": 3.286627769470215 + }, + { + "auxiliary_loss_clip": 0.01196058, + "auxiliary_loss_mlp": 0.01020454, + "balance_loss_clip": 1.04241562, + "balance_loss_mlp": 1.0134747, + "epoch": 0.9633860398003968, + "flos": 44419523823360.0, + "grad_norm": 1.6571526899702138, + "language_loss": 0.68308425, + "learning_rate": 1.4003062839657909e-08, + "loss": 0.70524943, + "num_input_tokens_seen": 172934895, + "step": 8012, + "time_per_iteration": 3.7917745113372803 + }, + { + "auxiliary_loss_clip": 0.01194919, + "auxiliary_loss_mlp": 0.01027137, + "balance_loss_clip": 1.04162836, + "balance_loss_mlp": 1.01963258, + "epoch": 0.9635062826910359, + "flos": 24827704300800.0, + "grad_norm": 1.6123178516992949, + "language_loss": 0.80135375, + "learning_rate": 1.391119571964583e-08, + "loss": 0.8235743, + "num_input_tokens_seen": 172955835, + "step": 8013, + "time_per_iteration": 2.694241762161255 + }, + { + "auxiliary_loss_clip": 0.01175693, + "auxiliary_loss_mlp": 0.01023625, + "balance_loss_clip": 1.04636049, + "balance_loss_mlp": 1.0167737, + "epoch": 0.9636265255816749, + "flos": 15961360095360.0, + "grad_norm": 1.7976897258228965, + "language_loss": 0.72940141, + "learning_rate": 1.3819629887726225e-08, + "loss": 0.75139463, + "num_input_tokens_seen": 172973925, + "step": 8014, + "time_per_iteration": 2.635942220687866 + }, + { + "auxiliary_loss_clip": 0.01188838, + "auxiliary_loss_mlp": 0.01023913, + "balance_loss_clip": 1.04521406, + "balance_loss_mlp": 1.01672447, + "epoch": 0.9637467684723141, + "flos": 22601781457920.0, + "grad_norm": 1.8857262901306486, + "language_loss": 0.76461107, + "learning_rate": 1.3728365357789317e-08, + "loss": 0.78673857, + "num_input_tokens_seen": 172993290, + "step": 8015, + "time_per_iteration": 2.615537405014038 + }, + { + "auxiliary_loss_clip": 0.01197376, + "auxiliary_loss_mlp": 0.01024648, + "balance_loss_clip": 1.03864002, + "balance_loss_mlp": 1.01803184, + "epoch": 0.9638670113629532, + "flos": 17565812801280.0, + "grad_norm": 2.564475130855009, + "language_loss": 0.76476657, + "learning_rate": 1.3637402143680254e-08, + "loss": 0.78698683, + "num_input_tokens_seen": 173008190, + "step": 8016, + "time_per_iteration": 2.7497217655181885 + }, + { + "auxiliary_loss_clip": 0.01095985, + "auxiliary_loss_mlp": 0.01003032, + "balance_loss_clip": 1.01051712, + "balance_loss_mlp": 1.0019238, + "epoch": 0.9639872542535922, + "flos": 55072139379840.0, + "grad_norm": 1.4433997859966206, + "language_loss": 0.55092883, + "learning_rate": 1.3546740259197998e-08, + "loss": 0.57191896, + "num_input_tokens_seen": 173061000, + "step": 8017, + "time_per_iteration": 4.0337748527526855 + }, + { + "auxiliary_loss_clip": 0.01184316, + "auxiliary_loss_mlp": 0.010225, + "balance_loss_clip": 1.04477489, + "balance_loss_mlp": 1.0149895, + "epoch": 0.9641074971442314, + "flos": 24134484746880.0, + "grad_norm": 2.087857512213157, + "language_loss": 0.70600772, + "learning_rate": 1.3456379718095989e-08, + "loss": 0.72807586, + "num_input_tokens_seen": 173081415, + "step": 8018, + "time_per_iteration": 2.66992449760437 + }, + { + "auxiliary_loss_clip": 0.01081654, + "auxiliary_loss_mlp": 0.00998945, + "balance_loss_clip": 1.00891197, + "balance_loss_mlp": 0.99787849, + "epoch": 0.9642277400348704, + "flos": 66747416077440.0, + "grad_norm": 0.8402974378243918, + "language_loss": 0.62003905, + "learning_rate": 1.3366320534081487e-08, + "loss": 0.640845, + "num_input_tokens_seen": 173144095, + "step": 8019, + "time_per_iteration": 3.2435712814331055 + }, + { + "auxiliary_loss_clip": 0.01174247, + "auxiliary_loss_mlp": 0.01022782, + "balance_loss_clip": 1.04544413, + "balance_loss_mlp": 1.01522458, + "epoch": 0.9643479829255095, + "flos": 30920272450560.0, + "grad_norm": 2.528944435427056, + "language_loss": 0.75668514, + "learning_rate": 1.3276562720816675e-08, + "loss": 0.77865547, + "num_input_tokens_seen": 173165605, + "step": 8020, + "time_per_iteration": 2.6778950691223145 + }, + { + "auxiliary_loss_clip": 0.01167229, + "auxiliary_loss_mlp": 0.01025136, + "balance_loss_clip": 1.04615927, + "balance_loss_mlp": 1.01745248, + "epoch": 0.9644682258161487, + "flos": 20048245643520.0, + "grad_norm": 2.310564609503307, + "language_loss": 0.825445, + "learning_rate": 1.3187106291917549e-08, + "loss": 0.8473686, + "num_input_tokens_seen": 173182595, + "step": 8021, + "time_per_iteration": 2.581456184387207 + }, + { + "auxiliary_loss_clip": 0.01173255, + "auxiliary_loss_mlp": 0.01026472, + "balance_loss_clip": 1.0469718, + "balance_loss_mlp": 1.01908064, + "epoch": 0.9645884687067877, + "flos": 21178713456000.0, + "grad_norm": 1.8618863565910584, + "language_loss": 0.70307082, + "learning_rate": 1.309795126095503e-08, + "loss": 0.72506809, + "num_input_tokens_seen": 173200895, + "step": 8022, + "time_per_iteration": 2.582960844039917 + }, + { + "auxiliary_loss_clip": 0.01213261, + "auxiliary_loss_mlp": 0.01020678, + "balance_loss_clip": 1.03552461, + "balance_loss_mlp": 1.01331711, + "epoch": 0.9647087115974268, + "flos": 18945967029120.0, + "grad_norm": 2.5193643385498286, + "language_loss": 0.80814976, + "learning_rate": 1.3009097641453192e-08, + "loss": 0.8304891, + "num_input_tokens_seen": 173218745, + "step": 8023, + "time_per_iteration": 2.852233648300171 + }, + { + "auxiliary_loss_clip": 0.01185475, + "auxiliary_loss_mlp": 0.01025255, + "balance_loss_clip": 1.04565752, + "balance_loss_mlp": 1.018031, + "epoch": 0.9648289544880659, + "flos": 16545088016640.0, + "grad_norm": 1.6625975227578502, + "language_loss": 0.75987142, + "learning_rate": 1.2920545446891474e-08, + "loss": 0.78197873, + "num_input_tokens_seen": 173235465, + "step": 8024, + "time_per_iteration": 2.706174373626709 + }, + { + "auxiliary_loss_clip": 0.01188201, + "auxiliary_loss_mlp": 0.01023825, + "balance_loss_clip": 1.04651237, + "balance_loss_mlp": 1.01671386, + "epoch": 0.964949197378705, + "flos": 24057527857920.0, + "grad_norm": 1.6715382253739768, + "language_loss": 0.70945603, + "learning_rate": 1.2832294690703127e-08, + "loss": 0.73157632, + "num_input_tokens_seen": 173254440, + "step": 8025, + "time_per_iteration": 2.7194104194641113 + }, + { + "auxiliary_loss_clip": 0.01177548, + "auxiliary_loss_mlp": 0.01023449, + "balance_loss_clip": 1.04856217, + "balance_loss_mlp": 1.01626635, + "epoch": 0.965069440269344, + "flos": 23365565280000.0, + "grad_norm": 3.0141205050788784, + "language_loss": 0.77478778, + "learning_rate": 1.2744345386275668e-08, + "loss": 0.79679775, + "num_input_tokens_seen": 173273980, + "step": 8026, + "time_per_iteration": 2.6510698795318604 + }, + { + "auxiliary_loss_clip": 0.01193265, + "auxiliary_loss_mlp": 0.01023868, + "balance_loss_clip": 1.04789686, + "balance_loss_mlp": 1.01657522, + "epoch": 0.9651896831599832, + "flos": 25374875155200.0, + "grad_norm": 2.307891447971173, + "language_loss": 0.78589183, + "learning_rate": 1.265669754695109e-08, + "loss": 0.80806315, + "num_input_tokens_seen": 173293550, + "step": 8027, + "time_per_iteration": 2.681586503982544 + }, + { + "auxiliary_loss_clip": 0.01210959, + "auxiliary_loss_mlp": 0.01029482, + "balance_loss_clip": 1.03795302, + "balance_loss_mlp": 1.0218879, + "epoch": 0.9653099260506223, + "flos": 22272875596800.0, + "grad_norm": 2.356414889688796, + "language_loss": 0.82101202, + "learning_rate": 1.2569351186025201e-08, + "loss": 0.84341639, + "num_input_tokens_seen": 173312005, + "step": 8028, + "time_per_iteration": 2.7647197246551514 + }, + { + "auxiliary_loss_clip": 0.01182221, + "auxiliary_loss_mlp": 0.01026182, + "balance_loss_clip": 1.0409447, + "balance_loss_mlp": 1.01910365, + "epoch": 0.9654301689412613, + "flos": 26760847386240.0, + "grad_norm": 1.4466527497048303, + "language_loss": 0.75764549, + "learning_rate": 1.2482306316748737e-08, + "loss": 0.77972955, + "num_input_tokens_seen": 173332450, + "step": 8029, + "time_per_iteration": 2.724609375 + }, + { + "auxiliary_loss_clip": 0.01179691, + "auxiliary_loss_mlp": 0.01024151, + "balance_loss_clip": 1.04423964, + "balance_loss_mlp": 1.01693869, + "epoch": 0.9655504118319005, + "flos": 17412689122560.0, + "grad_norm": 2.2235240587135365, + "language_loss": 0.78377235, + "learning_rate": 1.2395562952326021e-08, + "loss": 0.80581081, + "num_input_tokens_seen": 173349610, + "step": 8030, + "time_per_iteration": 2.6535146236419678 + }, + { + "auxiliary_loss_clip": 0.01198316, + "auxiliary_loss_mlp": 0.01029691, + "balance_loss_clip": 1.04980516, + "balance_loss_mlp": 1.02196074, + "epoch": 0.9656706547225395, + "flos": 22126970551680.0, + "grad_norm": 2.382093952799146, + "language_loss": 0.81476033, + "learning_rate": 1.2309121105916309e-08, + "loss": 0.83704042, + "num_input_tokens_seen": 173367900, + "step": 8031, + "time_per_iteration": 2.6106791496276855 + }, + { + "auxiliary_loss_clip": 0.01180901, + "auxiliary_loss_mlp": 0.01025928, + "balance_loss_clip": 1.04622412, + "balance_loss_mlp": 1.018507, + "epoch": 0.9657908976131786, + "flos": 37049289926400.0, + "grad_norm": 2.0168093497728004, + "language_loss": 0.69380617, + "learning_rate": 1.222298079063222e-08, + "loss": 0.71587449, + "num_input_tokens_seen": 173389040, + "step": 8032, + "time_per_iteration": 2.761599540710449 + }, + { + "auxiliary_loss_clip": 0.01178082, + "auxiliary_loss_mlp": 0.01025146, + "balance_loss_clip": 1.04690552, + "balance_loss_mlp": 1.01826215, + "epoch": 0.9659111405038178, + "flos": 24389809597440.0, + "grad_norm": 2.025597949479992, + "language_loss": 0.72503471, + "learning_rate": 1.2137142019541524e-08, + "loss": 0.74706697, + "num_input_tokens_seen": 173407595, + "step": 8033, + "time_per_iteration": 3.5749852657318115 + }, + { + "auxiliary_loss_clip": 0.01190366, + "auxiliary_loss_mlp": 0.01026687, + "balance_loss_clip": 1.04427528, + "balance_loss_mlp": 1.01959753, + "epoch": 0.9660313833944568, + "flos": 25009412227200.0, + "grad_norm": 2.090551501176105, + "language_loss": 0.73241448, + "learning_rate": 1.2051604805666027e-08, + "loss": 0.75458503, + "num_input_tokens_seen": 173424720, + "step": 8034, + "time_per_iteration": 2.772613763809204 + }, + { + "auxiliary_loss_clip": 0.0116711, + "auxiliary_loss_mlp": 0.01378592, + "balance_loss_clip": 1.04777932, + "balance_loss_mlp": 1.00018787, + "epoch": 0.9661516262850959, + "flos": 11801575895040.0, + "grad_norm": 2.7816219110201144, + "language_loss": 0.78498179, + "learning_rate": 1.196636916198135e-08, + "loss": 0.81043887, + "num_input_tokens_seen": 173442260, + "step": 8035, + "time_per_iteration": 2.573953866958618 + }, + { + "auxiliary_loss_clip": 0.01169079, + "auxiliary_loss_mlp": 0.01029188, + "balance_loss_clip": 1.04705894, + "balance_loss_mlp": 1.02206492, + "epoch": 0.9662718691757349, + "flos": 20047778766720.0, + "grad_norm": 1.862646700387711, + "language_loss": 0.76879162, + "learning_rate": 1.1881435101418036e-08, + "loss": 0.79077429, + "num_input_tokens_seen": 173461675, + "step": 8036, + "time_per_iteration": 3.4812228679656982 + }, + { + "auxiliary_loss_clip": 0.0108555, + "auxiliary_loss_mlp": 0.01003308, + "balance_loss_clip": 1.00923395, + "balance_loss_mlp": 1.00228882, + "epoch": 0.9663921120663741, + "flos": 68027703517440.0, + "grad_norm": 0.7254898863470339, + "language_loss": 0.65501094, + "learning_rate": 1.1796802636860003e-08, + "loss": 0.67589951, + "num_input_tokens_seen": 173530205, + "step": 8037, + "time_per_iteration": 3.253828525543213 + }, + { + "auxiliary_loss_clip": 0.01170994, + "auxiliary_loss_mlp": 0.01027135, + "balance_loss_clip": 1.0501107, + "balance_loss_mlp": 1.01991343, + "epoch": 0.9665123549570132, + "flos": 26322916769280.0, + "grad_norm": 2.2218235944751137, + "language_loss": 0.73909235, + "learning_rate": 1.1712471781146316e-08, + "loss": 0.76107371, + "num_input_tokens_seen": 173549540, + "step": 8038, + "time_per_iteration": 2.610583782196045 + }, + { + "auxiliary_loss_clip": 0.01165443, + "auxiliary_loss_mlp": 0.01028176, + "balance_loss_clip": 1.04543829, + "balance_loss_mlp": 1.0208683, + "epoch": 0.9666325978476522, + "flos": 43941121557120.0, + "grad_norm": 1.940540392237262, + "language_loss": 0.67163473, + "learning_rate": 1.1628442547069628e-08, + "loss": 0.69357091, + "num_input_tokens_seen": 173571740, + "step": 8039, + "time_per_iteration": 3.6019468307495117 + }, + { + "auxiliary_loss_clip": 0.01177878, + "auxiliary_loss_mlp": 0.01378781, + "balance_loss_clip": 1.04515243, + "balance_loss_mlp": 1.0001688, + "epoch": 0.9667528407382914, + "flos": 21543422198400.0, + "grad_norm": 1.8958092128226542, + "language_loss": 0.77058333, + "learning_rate": 1.1544714947377521e-08, + "loss": 0.79614991, + "num_input_tokens_seen": 173589425, + "step": 8040, + "time_per_iteration": 2.6907835006713867 + }, + { + "auxiliary_loss_clip": 0.0117122, + "auxiliary_loss_mlp": 0.01027217, + "balance_loss_clip": 1.04981899, + "balance_loss_mlp": 1.01922393, + "epoch": 0.9668730836289304, + "flos": 23878585278720.0, + "grad_norm": 4.088801755598811, + "language_loss": 0.70292461, + "learning_rate": 1.1461288994770945e-08, + "loss": 0.72490895, + "num_input_tokens_seen": 173608500, + "step": 8041, + "time_per_iteration": 2.5634610652923584 + }, + { + "auxiliary_loss_clip": 0.01170161, + "auxiliary_loss_mlp": 0.01022536, + "balance_loss_clip": 1.04633856, + "balance_loss_mlp": 1.01507926, + "epoch": 0.9669933265195695, + "flos": 28293011971200.0, + "grad_norm": 1.6713588268714414, + "language_loss": 0.77242255, + "learning_rate": 1.1378164701906002e-08, + "loss": 0.79434943, + "num_input_tokens_seen": 173630265, + "step": 8042, + "time_per_iteration": 2.65283203125 + }, + { + "auxiliary_loss_clip": 0.01170684, + "auxiliary_loss_mlp": 0.01026078, + "balance_loss_clip": 1.04781425, + "balance_loss_mlp": 1.01848733, + "epoch": 0.9671135694102087, + "flos": 22454763091200.0, + "grad_norm": 1.8827914633431122, + "language_loss": 0.66519994, + "learning_rate": 1.1295342081392156e-08, + "loss": 0.68716758, + "num_input_tokens_seen": 173649625, + "step": 8043, + "time_per_iteration": 3.4858925342559814 + }, + { + "auxiliary_loss_clip": 0.0118715, + "auxiliary_loss_mlp": 0.01023389, + "balance_loss_clip": 1.04421723, + "balance_loss_mlp": 1.01620698, + "epoch": 0.9672338123008477, + "flos": 20155941596160.0, + "grad_norm": 1.6428717779790205, + "language_loss": 0.69830543, + "learning_rate": 1.1212821145793804e-08, + "loss": 0.72041082, + "num_input_tokens_seen": 173669240, + "step": 8044, + "time_per_iteration": 2.640302896499634 + }, + { + "auxiliary_loss_clip": 0.01181435, + "auxiliary_loss_mlp": 0.01027415, + "balance_loss_clip": 1.04164124, + "balance_loss_mlp": 1.02022958, + "epoch": 0.9673540551914868, + "flos": 16977487939200.0, + "grad_norm": 1.8064406858215012, + "language_loss": 0.79027599, + "learning_rate": 1.1130601907629156e-08, + "loss": 0.81236446, + "num_input_tokens_seen": 173686970, + "step": 8045, + "time_per_iteration": 2.7194783687591553 + }, + { + "auxiliary_loss_clip": 0.01072971, + "auxiliary_loss_mlp": 0.01002724, + "balance_loss_clip": 1.00868189, + "balance_loss_mlp": 1.00168645, + "epoch": 0.9674742980821259, + "flos": 61892903952000.0, + "grad_norm": 0.8139560486441844, + "language_loss": 0.64828217, + "learning_rate": 1.1048684379370899e-08, + "loss": 0.66903913, + "num_input_tokens_seen": 173747655, + "step": 8046, + "time_per_iteration": 3.1915602684020996 + }, + { + "auxiliary_loss_clip": 0.01170925, + "auxiliary_loss_mlp": 0.01021024, + "balance_loss_clip": 1.0420264, + "balance_loss_mlp": 1.01458049, + "epoch": 0.967594540972765, + "flos": 18697824898560.0, + "grad_norm": 1.909654270487694, + "language_loss": 0.74269271, + "learning_rate": 1.0967068573445759e-08, + "loss": 0.7646122, + "num_input_tokens_seen": 173765140, + "step": 8047, + "time_per_iteration": 2.621072769165039 + }, + { + "auxiliary_loss_clip": 0.01180307, + "auxiliary_loss_mlp": 0.01024465, + "balance_loss_clip": 1.04226851, + "balance_loss_mlp": 1.01730394, + "epoch": 0.967714783863404, + "flos": 20777411733120.0, + "grad_norm": 2.44694456817161, + "language_loss": 0.65869373, + "learning_rate": 1.0885754502234945e-08, + "loss": 0.68074143, + "num_input_tokens_seen": 173784800, + "step": 8048, + "time_per_iteration": 2.6420323848724365 + }, + { + "auxiliary_loss_clip": 0.01191273, + "auxiliary_loss_mlp": 0.01020527, + "balance_loss_clip": 1.04237974, + "balance_loss_mlp": 1.01373184, + "epoch": 0.9678350267540432, + "flos": 23185473465600.0, + "grad_norm": 2.294548447563964, + "language_loss": 0.77920389, + "learning_rate": 1.08047421780737e-08, + "loss": 0.80132186, + "num_input_tokens_seen": 173803990, + "step": 8049, + "time_per_iteration": 2.6512672901153564 + }, + { + "auxiliary_loss_clip": 0.01192916, + "auxiliary_loss_mlp": 0.01378519, + "balance_loss_clip": 1.04631102, + "balance_loss_mlp": 1.00016367, + "epoch": 0.9679552696446823, + "flos": 21726063878400.0, + "grad_norm": 2.0369810073414523, + "language_loss": 0.73730123, + "learning_rate": 1.0724031613251305e-08, + "loss": 0.76301557, + "num_input_tokens_seen": 173821890, + "step": 8050, + "time_per_iteration": 2.6463544368743896 + }, + { + "auxiliary_loss_clip": 0.01184793, + "auxiliary_loss_mlp": 0.01025721, + "balance_loss_clip": 1.04902458, + "balance_loss_mlp": 1.01831245, + "epoch": 0.9680755125353213, + "flos": 26869046129280.0, + "grad_norm": 2.3772070141229853, + "language_loss": 0.66956186, + "learning_rate": 1.0643622820011744e-08, + "loss": 0.69166702, + "num_input_tokens_seen": 173842945, + "step": 8051, + "time_per_iteration": 2.6376290321350098 + }, + { + "auxiliary_loss_clip": 0.01172571, + "auxiliary_loss_mlp": 0.0102489, + "balance_loss_clip": 1.04887307, + "balance_loss_mlp": 1.01720715, + "epoch": 0.9681957554259605, + "flos": 28325008010880.0, + "grad_norm": 2.230424446274845, + "language_loss": 0.68222195, + "learning_rate": 1.0563515810552814e-08, + "loss": 0.70419657, + "num_input_tokens_seen": 173859915, + "step": 8052, + "time_per_iteration": 2.5395851135253906 + }, + { + "auxiliary_loss_clip": 0.01168277, + "auxiliary_loss_mlp": 0.01029998, + "balance_loss_clip": 1.04830837, + "balance_loss_mlp": 1.0230484, + "epoch": 0.9683159983165995, + "flos": 20557674282240.0, + "grad_norm": 2.286542587779488, + "language_loss": 0.7361989, + "learning_rate": 1.0483710597026795e-08, + "loss": 0.75818169, + "num_input_tokens_seen": 173879775, + "step": 8053, + "time_per_iteration": 2.609600782394409 + }, + { + "auxiliary_loss_clip": 0.01195154, + "auxiliary_loss_mlp": 0.01023295, + "balance_loss_clip": 1.04261565, + "balance_loss_mlp": 1.01628864, + "epoch": 0.9684362412072386, + "flos": 24207958016640.0, + "grad_norm": 2.3696512799360057, + "language_loss": 0.73934531, + "learning_rate": 1.0404207191540227e-08, + "loss": 0.76152986, + "num_input_tokens_seen": 173900230, + "step": 8054, + "time_per_iteration": 2.688913106918335 + }, + { + "auxiliary_loss_clip": 0.0116773, + "auxiliary_loss_mlp": 0.01021818, + "balance_loss_clip": 1.04762721, + "balance_loss_mlp": 1.01483226, + "epoch": 0.9685564840978778, + "flos": 22346241125760.0, + "grad_norm": 1.9065306795608965, + "language_loss": 0.74727416, + "learning_rate": 1.0325005606153236e-08, + "loss": 0.76916969, + "num_input_tokens_seen": 173919690, + "step": 8055, + "time_per_iteration": 2.5932376384735107 + }, + { + "auxiliary_loss_clip": 0.01207175, + "auxiliary_loss_mlp": 0.01027252, + "balance_loss_clip": 1.04246879, + "balance_loss_mlp": 1.01945519, + "epoch": 0.9686767269885168, + "flos": 14386389477120.0, + "grad_norm": 2.522997131945559, + "language_loss": 0.79180908, + "learning_rate": 1.0246105852881104e-08, + "loss": 0.81415331, + "num_input_tokens_seen": 173934790, + "step": 8056, + "time_per_iteration": 2.6516788005828857 + }, + { + "auxiliary_loss_clip": 0.0116794, + "auxiliary_loss_mlp": 0.0102921, + "balance_loss_clip": 1.0460676, + "balance_loss_mlp": 1.0213896, + "epoch": 0.9687969698791559, + "flos": 21287630471040.0, + "grad_norm": 1.856664061357211, + "language_loss": 0.78477019, + "learning_rate": 1.0167507943692476e-08, + "loss": 0.80674165, + "num_input_tokens_seen": 173953875, + "step": 8057, + "time_per_iteration": 2.585435390472412 + }, + { + "auxiliary_loss_clip": 0.01174762, + "auxiliary_loss_mlp": 0.01027419, + "balance_loss_clip": 1.04784346, + "balance_loss_mlp": 1.02004278, + "epoch": 0.968917212769795, + "flos": 19828328624640.0, + "grad_norm": 2.256776495041318, + "language_loss": 0.71218109, + "learning_rate": 1.008921189051093e-08, + "loss": 0.73420286, + "num_input_tokens_seen": 173971220, + "step": 8058, + "time_per_iteration": 2.6012964248657227 + }, + { + "auxiliary_loss_clip": 0.01170256, + "auxiliary_loss_mlp": 0.01026741, + "balance_loss_clip": 1.04931211, + "balance_loss_mlp": 1.01845598, + "epoch": 0.9690374556604341, + "flos": 21681749473920.0, + "grad_norm": 4.145056309096784, + "language_loss": 0.77274513, + "learning_rate": 1.0011217705213848e-08, + "loss": 0.79471511, + "num_input_tokens_seen": 173989095, + "step": 8059, + "time_per_iteration": 3.476640224456787 + }, + { + "auxiliary_loss_clip": 0.01175617, + "auxiliary_loss_mlp": 0.01028386, + "balance_loss_clip": 1.04754436, + "balance_loss_mlp": 1.02138209, + "epoch": 0.9691576985510731, + "flos": 32635437851520.0, + "grad_norm": 1.8999062136973375, + "language_loss": 0.7446965, + "learning_rate": 9.933525399632658e-09, + "loss": 0.76673651, + "num_input_tokens_seen": 174007330, + "step": 8060, + "time_per_iteration": 2.691225051879883 + }, + { + "auxiliary_loss_clip": 0.01181461, + "auxiliary_loss_mlp": 0.01026243, + "balance_loss_clip": 1.04285192, + "balance_loss_mlp": 1.01826787, + "epoch": 0.9692779414417123, + "flos": 35663174040960.0, + "grad_norm": 1.7590242588326181, + "language_loss": 0.65521443, + "learning_rate": 9.856134985553488e-09, + "loss": 0.67729151, + "num_input_tokens_seen": 174027055, + "step": 8061, + "time_per_iteration": 2.6957755088806152 + }, + { + "auxiliary_loss_clip": 0.01165189, + "auxiliary_loss_mlp": 0.01023578, + "balance_loss_clip": 1.04594254, + "balance_loss_mlp": 1.01633251, + "epoch": 0.9693981843323514, + "flos": 28366952117760.0, + "grad_norm": 1.7404422947953855, + "language_loss": 0.73769844, + "learning_rate": 9.77904647471628e-09, + "loss": 0.7595861, + "num_input_tokens_seen": 174050235, + "step": 8062, + "time_per_iteration": 3.532994508743286 + }, + { + "auxiliary_loss_clip": 0.01191729, + "auxiliary_loss_mlp": 0.01025637, + "balance_loss_clip": 1.03838158, + "balance_loss_mlp": 1.01859176, + "epoch": 0.9695184272229904, + "flos": 23622865378560.0, + "grad_norm": 1.5667260731783592, + "language_loss": 0.74168605, + "learning_rate": 9.702259878815454e-09, + "loss": 0.76385975, + "num_input_tokens_seen": 174070560, + "step": 8063, + "time_per_iteration": 2.6842405796051025 + }, + { + "auxiliary_loss_clip": 0.01180998, + "auxiliary_loss_mlp": 0.010283, + "balance_loss_clip": 1.0484097, + "balance_loss_mlp": 1.02053905, + "epoch": 0.9696386701136296, + "flos": 23294677789440.0, + "grad_norm": 2.206103038586952, + "language_loss": 0.74448574, + "learning_rate": 9.625775209499254e-09, + "loss": 0.76657867, + "num_input_tokens_seen": 174090565, + "step": 8064, + "time_per_iteration": 2.627938985824585 + }, + { + "auxiliary_loss_clip": 0.01184392, + "auxiliary_loss_mlp": 0.01024957, + "balance_loss_clip": 1.0411334, + "balance_loss_mlp": 1.01780462, + "epoch": 0.9697589130042686, + "flos": 15121876360320.0, + "grad_norm": 2.180411185992323, + "language_loss": 0.74333823, + "learning_rate": 9.549592478370172e-09, + "loss": 0.76543176, + "num_input_tokens_seen": 174108745, + "step": 8065, + "time_per_iteration": 3.569671630859375 + }, + { + "auxiliary_loss_clip": 0.01178861, + "auxiliary_loss_mlp": 0.01022681, + "balance_loss_clip": 1.04568231, + "balance_loss_mlp": 1.01538563, + "epoch": 0.9698791558949077, + "flos": 18879532824960.0, + "grad_norm": 1.6979887902957793, + "language_loss": 0.79125631, + "learning_rate": 9.473711696985632e-09, + "loss": 0.8132717, + "num_input_tokens_seen": 174128075, + "step": 8066, + "time_per_iteration": 2.5596041679382324 + }, + { + "auxiliary_loss_clip": 0.0118355, + "auxiliary_loss_mlp": 0.01023298, + "balance_loss_clip": 1.04456079, + "balance_loss_mlp": 1.01587772, + "epoch": 0.9699993987855468, + "flos": 17931455297280.0, + "grad_norm": 2.2266225439953713, + "language_loss": 0.75988746, + "learning_rate": 9.398132876856201e-09, + "loss": 0.78195596, + "num_input_tokens_seen": 174147040, + "step": 8067, + "time_per_iteration": 2.6683573722839355 + }, + { + "auxiliary_loss_clip": 0.01110652, + "auxiliary_loss_mlp": 0.01000802, + "balance_loss_clip": 1.00893021, + "balance_loss_mlp": 0.99962789, + "epoch": 0.9701196416761859, + "flos": 67182186297600.0, + "grad_norm": 0.7763553724371846, + "language_loss": 0.60814625, + "learning_rate": 9.322856029447379e-09, + "loss": 0.62926084, + "num_input_tokens_seen": 174208225, + "step": 8068, + "time_per_iteration": 4.002663850784302 + }, + { + "auxiliary_loss_clip": 0.01165885, + "auxiliary_loss_mlp": 0.01017539, + "balance_loss_clip": 1.04804349, + "balance_loss_mlp": 1.00996017, + "epoch": 0.970239884566825, + "flos": 24277804012800.0, + "grad_norm": 2.4181133193192688, + "language_loss": 0.8018229, + "learning_rate": 9.247881166178695e-09, + "loss": 0.82365716, + "num_input_tokens_seen": 174226935, + "step": 8069, + "time_per_iteration": 2.6099495887756348 + }, + { + "auxiliary_loss_clip": 0.01201073, + "auxiliary_loss_mlp": 0.01025529, + "balance_loss_clip": 1.0435878, + "balance_loss_mlp": 1.01840043, + "epoch": 0.970360127457464, + "flos": 25301689194240.0, + "grad_norm": 2.361590245034372, + "language_loss": 0.76719999, + "learning_rate": 9.173208298423274e-09, + "loss": 0.78946602, + "num_input_tokens_seen": 174248140, + "step": 8070, + "time_per_iteration": 2.863656520843506 + }, + { + "auxiliary_loss_clip": 0.01199356, + "auxiliary_loss_mlp": 0.01378509, + "balance_loss_clip": 1.04044223, + "balance_loss_mlp": 1.00017142, + "epoch": 0.9704803703481032, + "flos": 29572473398400.0, + "grad_norm": 1.5230172124944248, + "language_loss": 0.76184762, + "learning_rate": 9.09883743750961e-09, + "loss": 0.78762627, + "num_input_tokens_seen": 174271030, + "step": 8071, + "time_per_iteration": 2.780150890350342 + }, + { + "auxiliary_loss_clip": 0.01181412, + "auxiliary_loss_mlp": 0.01022301, + "balance_loss_clip": 1.04348755, + "balance_loss_mlp": 1.01558924, + "epoch": 0.9706006132387422, + "flos": 17380046638080.0, + "grad_norm": 1.592930316199334, + "language_loss": 0.84254348, + "learning_rate": 9.024768594719124e-09, + "loss": 0.86458057, + "num_input_tokens_seen": 174289410, + "step": 8072, + "time_per_iteration": 2.7416577339172363 + }, + { + "auxiliary_loss_clip": 0.01193854, + "auxiliary_loss_mlp": 0.01021957, + "balance_loss_clip": 1.04338646, + "balance_loss_mlp": 1.01475072, + "epoch": 0.9707208561293813, + "flos": 18186421011840.0, + "grad_norm": 2.318621053676266, + "language_loss": 0.72601694, + "learning_rate": 8.95100178128816e-09, + "loss": 0.74817502, + "num_input_tokens_seen": 174308550, + "step": 8073, + "time_per_iteration": 2.6385881900787354 + }, + { + "auxiliary_loss_clip": 0.01186583, + "auxiliary_loss_mlp": 0.0102287, + "balance_loss_clip": 1.04618907, + "balance_loss_mlp": 1.01518095, + "epoch": 0.9708410990200205, + "flos": 31248388212480.0, + "grad_norm": 2.041218353436397, + "language_loss": 0.69936228, + "learning_rate": 8.877537008407321e-09, + "loss": 0.72145677, + "num_input_tokens_seen": 174328600, + "step": 8074, + "time_per_iteration": 2.7813045978546143 + }, + { + "auxiliary_loss_clip": 0.01188245, + "auxiliary_loss_mlp": 0.01023438, + "balance_loss_clip": 1.04415321, + "balance_loss_mlp": 1.01560628, + "epoch": 0.9709613419106595, + "flos": 30554450386560.0, + "grad_norm": 1.6944349806544645, + "language_loss": 0.68636388, + "learning_rate": 8.804374287221028e-09, + "loss": 0.70848072, + "num_input_tokens_seen": 174349835, + "step": 8075, + "time_per_iteration": 2.730463743209839 + }, + { + "auxiliary_loss_clip": 0.01183164, + "auxiliary_loss_mlp": 0.01027432, + "balance_loss_clip": 1.03877997, + "balance_loss_mlp": 1.02005327, + "epoch": 0.9710815848012986, + "flos": 23730166281600.0, + "grad_norm": 1.7513967022253343, + "language_loss": 0.84599191, + "learning_rate": 8.731513628827958e-09, + "loss": 0.8680979, + "num_input_tokens_seen": 174369200, + "step": 8076, + "time_per_iteration": 2.713466167449951 + }, + { + "auxiliary_loss_clip": 0.01178217, + "auxiliary_loss_mlp": 0.01021838, + "balance_loss_clip": 1.04678702, + "balance_loss_mlp": 1.01484597, + "epoch": 0.9712018276919377, + "flos": 23761875012480.0, + "grad_norm": 2.1039747604616843, + "language_loss": 0.82712305, + "learning_rate": 8.658955044280825e-09, + "loss": 0.84912366, + "num_input_tokens_seen": 174388125, + "step": 8077, + "time_per_iteration": 2.730912446975708 + }, + { + "auxiliary_loss_clip": 0.01174331, + "auxiliary_loss_mlp": 0.0102265, + "balance_loss_clip": 1.04630208, + "balance_loss_mlp": 1.01583767, + "epoch": 0.9713220705825768, + "flos": 23330983461120.0, + "grad_norm": 1.6712497779987634, + "language_loss": 0.77503091, + "learning_rate": 8.586698544587268e-09, + "loss": 0.79700071, + "num_input_tokens_seen": 174409735, + "step": 8078, + "time_per_iteration": 2.65451717376709 + }, + { + "auxiliary_loss_clip": 0.01173925, + "auxiliary_loss_mlp": 0.01025686, + "balance_loss_clip": 1.04214406, + "balance_loss_mlp": 1.01788998, + "epoch": 0.9714423134732159, + "flos": 22200946611840.0, + "grad_norm": 1.8222876688370344, + "language_loss": 0.73995793, + "learning_rate": 8.514744140707853e-09, + "loss": 0.76195395, + "num_input_tokens_seen": 174428875, + "step": 8079, + "time_per_iteration": 2.6984763145446777 + }, + { + "auxiliary_loss_clip": 0.01166526, + "auxiliary_loss_mlp": 0.01022231, + "balance_loss_clip": 1.0474062, + "balance_loss_mlp": 1.01543665, + "epoch": 0.971562556363855, + "flos": 20229917656320.0, + "grad_norm": 1.8108634838952364, + "language_loss": 0.76132566, + "learning_rate": 8.443091843558515e-09, + "loss": 0.78321326, + "num_input_tokens_seen": 174447960, + "step": 8080, + "time_per_iteration": 2.604602098464966 + }, + { + "auxiliary_loss_clip": 0.01178746, + "auxiliary_loss_mlp": 0.01022558, + "balance_loss_clip": 1.04374945, + "balance_loss_mlp": 1.01541483, + "epoch": 0.9716827992544941, + "flos": 24970197553920.0, + "grad_norm": 2.225976523048117, + "language_loss": 0.64466655, + "learning_rate": 8.37174166400878e-09, + "loss": 0.66667956, + "num_input_tokens_seen": 174463535, + "step": 8081, + "time_per_iteration": 2.6905605792999268 + }, + { + "auxiliary_loss_clip": 0.01167332, + "auxiliary_loss_mlp": 0.01028921, + "balance_loss_clip": 1.04865837, + "balance_loss_mlp": 1.02130961, + "epoch": 0.9718030421451331, + "flos": 24681476033280.0, + "grad_norm": 2.2439992588931807, + "language_loss": 0.84895796, + "learning_rate": 8.300693612881992e-09, + "loss": 0.87092054, + "num_input_tokens_seen": 174483600, + "step": 8082, + "time_per_iteration": 2.615447521209717 + }, + { + "auxiliary_loss_clip": 0.01174481, + "auxiliary_loss_mlp": 0.01378627, + "balance_loss_clip": 1.04574609, + "balance_loss_mlp": 1.00013041, + "epoch": 0.9719232850357723, + "flos": 22090700793600.0, + "grad_norm": 2.2446195634907036, + "language_loss": 0.81305492, + "learning_rate": 8.22994770095664e-09, + "loss": 0.83858597, + "num_input_tokens_seen": 174502175, + "step": 8083, + "time_per_iteration": 2.6529533863067627 + }, + { + "auxiliary_loss_clip": 0.01185764, + "auxiliary_loss_mlp": 0.01031801, + "balance_loss_clip": 1.04636157, + "balance_loss_mlp": 1.02452672, + "epoch": 0.9720435279264114, + "flos": 23656908493440.0, + "grad_norm": 2.8608103673089698, + "language_loss": 0.75382447, + "learning_rate": 8.159503938964585e-09, + "loss": 0.77600014, + "num_input_tokens_seen": 174519495, + "step": 8084, + "time_per_iteration": 2.6193418502807617 + }, + { + "auxiliary_loss_clip": 0.01186021, + "auxiliary_loss_mlp": 0.01030383, + "balance_loss_clip": 1.04142165, + "balance_loss_mlp": 1.02326608, + "epoch": 0.9721637708170504, + "flos": 28365910623360.0, + "grad_norm": 1.7922078997240822, + "language_loss": 0.70351422, + "learning_rate": 8.089362337592164e-09, + "loss": 0.72567821, + "num_input_tokens_seen": 174543120, + "step": 8085, + "time_per_iteration": 3.665126323699951 + }, + { + "auxiliary_loss_clip": 0.01182808, + "auxiliary_loss_mlp": 0.01022118, + "balance_loss_clip": 1.04637551, + "balance_loss_mlp": 1.01521266, + "epoch": 0.9722840137076896, + "flos": 29130807767040.0, + "grad_norm": 1.5996072595876905, + "language_loss": 0.72325838, + "learning_rate": 8.019522907479536e-09, + "loss": 0.74530768, + "num_input_tokens_seen": 174563480, + "step": 8086, + "time_per_iteration": 2.711486577987671 + }, + { + "auxiliary_loss_clip": 0.01180763, + "auxiliary_loss_mlp": 0.01025737, + "balance_loss_clip": 1.04847491, + "balance_loss_mlp": 1.01871622, + "epoch": 0.9724042565983286, + "flos": 19243954258560.0, + "grad_norm": 2.3992177981970033, + "language_loss": 0.77345061, + "learning_rate": 7.949985659221558e-09, + "loss": 0.7955156, + "num_input_tokens_seen": 174580745, + "step": 8087, + "time_per_iteration": 2.6796162128448486 + }, + { + "auxiliary_loss_clip": 0.01189041, + "auxiliary_loss_mlp": 0.01024847, + "balance_loss_clip": 1.04497957, + "balance_loss_mlp": 1.01779592, + "epoch": 0.9725244994889677, + "flos": 23039676161280.0, + "grad_norm": 2.0293115078747257, + "language_loss": 0.78760314, + "learning_rate": 7.880750603366904e-09, + "loss": 0.80974197, + "num_input_tokens_seen": 174599615, + "step": 8088, + "time_per_iteration": 3.5887598991394043 + }, + { + "auxiliary_loss_clip": 0.01201064, + "auxiliary_loss_mlp": 0.01026342, + "balance_loss_clip": 1.04340744, + "balance_loss_mlp": 1.01848006, + "epoch": 0.9726447423796069, + "flos": 23367468700800.0, + "grad_norm": 2.223253214627558, + "language_loss": 0.79652268, + "learning_rate": 7.811817750418282e-09, + "loss": 0.81879675, + "num_input_tokens_seen": 174618375, + "step": 8089, + "time_per_iteration": 2.675347328186035 + }, + { + "auxiliary_loss_clip": 0.01191406, + "auxiliary_loss_mlp": 0.01029577, + "balance_loss_clip": 1.04354763, + "balance_loss_mlp": 1.02184629, + "epoch": 0.9727649852702459, + "flos": 26541648639360.0, + "grad_norm": 1.67319592072457, + "language_loss": 0.80066973, + "learning_rate": 7.743187110833105e-09, + "loss": 0.82287961, + "num_input_tokens_seen": 174641135, + "step": 8090, + "time_per_iteration": 2.776275634765625 + }, + { + "auxiliary_loss_clip": 0.01186742, + "auxiliary_loss_mlp": 0.01021975, + "balance_loss_clip": 1.04292262, + "balance_loss_mlp": 1.01491499, + "epoch": 0.972885228160885, + "flos": 20522338277760.0, + "grad_norm": 1.4676817863306282, + "language_loss": 0.806656, + "learning_rate": 7.674858695022602e-09, + "loss": 0.82874316, + "num_input_tokens_seen": 174659490, + "step": 8091, + "time_per_iteration": 3.5480682849884033 + }, + { + "auxiliary_loss_clip": 0.01170077, + "auxiliary_loss_mlp": 0.01026414, + "balance_loss_clip": 1.04893208, + "balance_loss_mlp": 1.01881433, + "epoch": 0.9730054710515241, + "flos": 17566064196480.0, + "grad_norm": 6.0102743692865666, + "language_loss": 0.76155335, + "learning_rate": 7.606832513351591e-09, + "loss": 0.78351825, + "num_input_tokens_seen": 174677440, + "step": 8092, + "time_per_iteration": 2.579463243484497 + }, + { + "auxiliary_loss_clip": 0.01059655, + "auxiliary_loss_mlp": 0.01373654, + "balance_loss_clip": 1.00857997, + "balance_loss_mlp": 0.99979419, + "epoch": 0.9731257139421632, + "flos": 68972010117120.0, + "grad_norm": 0.8299982544655193, + "language_loss": 0.63938487, + "learning_rate": 7.539108576140264e-09, + "loss": 0.66371799, + "num_input_tokens_seen": 174741550, + "step": 8093, + "time_per_iteration": 3.2650418281555176 + }, + { + "auxiliary_loss_clip": 0.01201492, + "auxiliary_loss_mlp": 0.01025253, + "balance_loss_clip": 1.04308558, + "balance_loss_mlp": 1.01859522, + "epoch": 0.9732459568328022, + "flos": 18478841633280.0, + "grad_norm": 3.2401152369517323, + "language_loss": 0.70547378, + "learning_rate": 7.471686893661732e-09, + "loss": 0.72774124, + "num_input_tokens_seen": 174759845, + "step": 8094, + "time_per_iteration": 3.5785751342773438 + }, + { + "auxiliary_loss_clip": 0.01185768, + "auxiliary_loss_mlp": 0.01024565, + "balance_loss_clip": 1.04683375, + "balance_loss_mlp": 1.0175643, + "epoch": 0.9733661997234414, + "flos": 20883886623360.0, + "grad_norm": 1.7448971927758334, + "language_loss": 0.64233792, + "learning_rate": 7.4045674761442636e-09, + "loss": 0.66444123, + "num_input_tokens_seen": 174777175, + "step": 8095, + "time_per_iteration": 2.6914403438568115 + }, + { + "auxiliary_loss_clip": 0.01165708, + "auxiliary_loss_mlp": 0.01378752, + "balance_loss_clip": 1.04613042, + "balance_loss_mlp": 1.00013232, + "epoch": 0.9734864426140805, + "flos": 23766795175680.0, + "grad_norm": 5.443078660572433, + "language_loss": 0.7432791, + "learning_rate": 7.337750333769488e-09, + "loss": 0.76872367, + "num_input_tokens_seen": 174796980, + "step": 8096, + "time_per_iteration": 2.65468692779541 + }, + { + "auxiliary_loss_clip": 0.01187386, + "auxiliary_loss_mlp": 0.01030925, + "balance_loss_clip": 1.04122937, + "balance_loss_mlp": 1.02362359, + "epoch": 0.9736066855047195, + "flos": 35042422176000.0, + "grad_norm": 1.8844505194539891, + "language_loss": 0.72924715, + "learning_rate": 7.2712354766737425e-09, + "loss": 0.75143027, + "num_input_tokens_seen": 174817310, + "step": 8097, + "time_per_iteration": 2.7477498054504395 + }, + { + "auxiliary_loss_clip": 0.01185633, + "auxiliary_loss_mlp": 0.01027642, + "balance_loss_clip": 1.04429197, + "balance_loss_mlp": 1.02011955, + "epoch": 0.9737269283953586, + "flos": 20410620001920.0, + "grad_norm": 1.6382276631126553, + "language_loss": 0.80752015, + "learning_rate": 7.2050229149469565e-09, + "loss": 0.82965291, + "num_input_tokens_seen": 174837320, + "step": 8098, + "time_per_iteration": 2.757800817489624 + }, + { + "auxiliary_loss_clip": 0.0119472, + "auxiliary_loss_mlp": 0.01024382, + "balance_loss_clip": 1.04052365, + "balance_loss_mlp": 1.01759303, + "epoch": 0.9738471712859977, + "flos": 28911680847360.0, + "grad_norm": 1.8280588007797318, + "language_loss": 0.63780653, + "learning_rate": 7.139112658633984e-09, + "loss": 0.65999758, + "num_input_tokens_seen": 174857470, + "step": 8099, + "time_per_iteration": 2.74727463722229 + }, + { + "auxiliary_loss_clip": 0.01192439, + "auxiliary_loss_mlp": 0.01021971, + "balance_loss_clip": 1.04453278, + "balance_loss_mlp": 1.01520324, + "epoch": 0.9739674141766368, + "flos": 27782326356480.0, + "grad_norm": 2.176180288478959, + "language_loss": 0.70406854, + "learning_rate": 7.073504717733048e-09, + "loss": 0.72621262, + "num_input_tokens_seen": 174877035, + "step": 8100, + "time_per_iteration": 2.819019317626953 + }, + { + "auxiliary_loss_clip": 0.01110086, + "auxiliary_loss_mlp": 0.01002635, + "balance_loss_clip": 1.01055837, + "balance_loss_mlp": 1.001508, + "epoch": 0.9740876570672758, + "flos": 68863057188480.0, + "grad_norm": 0.7318880887675816, + "language_loss": 0.57215011, + "learning_rate": 7.008199102196855e-09, + "loss": 0.59327734, + "num_input_tokens_seen": 174938460, + "step": 8101, + "time_per_iteration": 3.251774311065674 + }, + { + "auxiliary_loss_clip": 0.01085475, + "auxiliary_loss_mlp": 0.01000718, + "balance_loss_clip": 1.01595247, + "balance_loss_mlp": 0.99966258, + "epoch": 0.974207899957915, + "flos": 58236622646400.0, + "grad_norm": 0.7980972852632402, + "language_loss": 0.58983892, + "learning_rate": 6.9431958219321464e-09, + "loss": 0.61070091, + "num_input_tokens_seen": 174994625, + "step": 8102, + "time_per_iteration": 3.210480213165283 + }, + { + "auxiliary_loss_clip": 0.01182441, + "auxiliary_loss_mlp": 0.01030465, + "balance_loss_clip": 1.04352045, + "balance_loss_mlp": 1.02299392, + "epoch": 0.9743281428485541, + "flos": 22600057605120.0, + "grad_norm": 1.6188067152321075, + "language_loss": 0.7783615, + "learning_rate": 6.878494886800146e-09, + "loss": 0.80049062, + "num_input_tokens_seen": 175015400, + "step": 8103, + "time_per_iteration": 2.6794118881225586 + }, + { + "auxiliary_loss_clip": 0.01186648, + "auxiliary_loss_mlp": 0.01019564, + "balance_loss_clip": 1.04588199, + "balance_loss_mlp": 1.01251316, + "epoch": 0.9744483857391931, + "flos": 20008815488640.0, + "grad_norm": 1.9369790722920148, + "language_loss": 0.76403511, + "learning_rate": 6.814096306615669e-09, + "loss": 0.78609723, + "num_input_tokens_seen": 175033540, + "step": 8104, + "time_per_iteration": 2.657581329345703 + }, + { + "auxiliary_loss_clip": 0.01188711, + "auxiliary_loss_mlp": 0.01026233, + "balance_loss_clip": 1.04224241, + "balance_loss_mlp": 1.01921511, + "epoch": 0.9745686286298323, + "flos": 17675268520320.0, + "grad_norm": 4.2445083695554064, + "language_loss": 0.65354824, + "learning_rate": 6.750000091148011e-09, + "loss": 0.67569768, + "num_input_tokens_seen": 175050835, + "step": 8105, + "time_per_iteration": 2.646092414855957 + }, + { + "auxiliary_loss_clip": 0.01170063, + "auxiliary_loss_mlp": 0.01028553, + "balance_loss_clip": 1.04941797, + "balance_loss_mlp": 1.02128482, + "epoch": 0.9746888715204713, + "flos": 29460252332160.0, + "grad_norm": 1.8587174609209332, + "language_loss": 0.72544128, + "learning_rate": 6.686206250120729e-09, + "loss": 0.74742746, + "num_input_tokens_seen": 175072330, + "step": 8106, + "time_per_iteration": 2.658350944519043 + }, + { + "auxiliary_loss_clip": 0.01196326, + "auxiliary_loss_mlp": 0.01022844, + "balance_loss_clip": 1.04097962, + "balance_loss_mlp": 1.01541162, + "epoch": 0.9748091144111104, + "flos": 18479308510080.0, + "grad_norm": 2.128221917947624, + "language_loss": 0.7457056, + "learning_rate": 6.622714793210749e-09, + "loss": 0.76789731, + "num_input_tokens_seen": 175091250, + "step": 8107, + "time_per_iteration": 2.662277936935425 + }, + { + "auxiliary_loss_clip": 0.01167502, + "auxiliary_loss_mlp": 0.01027145, + "balance_loss_clip": 1.04708755, + "balance_loss_mlp": 1.01992643, + "epoch": 0.9749293573017496, + "flos": 20665154753280.0, + "grad_norm": 1.6419769842435306, + "language_loss": 0.78479105, + "learning_rate": 6.559525730050364e-09, + "loss": 0.80673754, + "num_input_tokens_seen": 175111350, + "step": 8108, + "time_per_iteration": 2.575286626815796 + }, + { + "auxiliary_loss_clip": 0.01194232, + "auxiliary_loss_mlp": 0.01026824, + "balance_loss_clip": 1.04225612, + "balance_loss_mlp": 1.01937103, + "epoch": 0.9750496001923886, + "flos": 18478590238080.0, + "grad_norm": 1.953938351365, + "language_loss": 0.76199985, + "learning_rate": 6.496639070224574e-09, + "loss": 0.78421044, + "num_input_tokens_seen": 175129835, + "step": 8109, + "time_per_iteration": 2.6900932788848877 + }, + { + "auxiliary_loss_clip": 0.01180535, + "auxiliary_loss_mlp": 0.01024826, + "balance_loss_clip": 1.0483253, + "balance_loss_mlp": 1.01766109, + "epoch": 0.9751698430830277, + "flos": 19572967860480.0, + "grad_norm": 2.2258327687845085, + "language_loss": 0.83792055, + "learning_rate": 6.4340548232739714e-09, + "loss": 0.85997415, + "num_input_tokens_seen": 175146035, + "step": 8110, + "time_per_iteration": 2.5958287715911865 + }, + { + "auxiliary_loss_clip": 0.01197508, + "auxiliary_loss_mlp": 0.01022154, + "balance_loss_clip": 1.04209697, + "balance_loss_mlp": 1.01512957, + "epoch": 0.9752900859736668, + "flos": 23550325862400.0, + "grad_norm": 2.0753375721414615, + "language_loss": 0.79110169, + "learning_rate": 6.371772998692071e-09, + "loss": 0.81329829, + "num_input_tokens_seen": 175165290, + "step": 8111, + "time_per_iteration": 3.7181570529937744 + }, + { + "auxiliary_loss_clip": 0.01197034, + "auxiliary_loss_mlp": 0.01028694, + "balance_loss_clip": 1.04081881, + "balance_loss_mlp": 1.02106452, + "epoch": 0.9754103288643059, + "flos": 20303211358080.0, + "grad_norm": 3.6806571691238736, + "language_loss": 0.64943832, + "learning_rate": 6.309793605927094e-09, + "loss": 0.67169559, + "num_input_tokens_seen": 175183610, + "step": 8112, + "time_per_iteration": 2.79683256149292 + }, + { + "auxiliary_loss_clip": 0.01188934, + "auxiliary_loss_mlp": 0.01022599, + "balance_loss_clip": 1.04352748, + "balance_loss_mlp": 1.01627243, + "epoch": 0.975530571754945, + "flos": 19350680544000.0, + "grad_norm": 1.701191521709829, + "language_loss": 0.80079758, + "learning_rate": 6.248116654381297e-09, + "loss": 0.82291293, + "num_input_tokens_seen": 175202080, + "step": 8113, + "time_per_iteration": 2.8039212226867676 + }, + { + "auxiliary_loss_clip": 0.01187618, + "auxiliary_loss_mlp": 0.01022535, + "balance_loss_clip": 1.0430913, + "balance_loss_mlp": 1.01616883, + "epoch": 0.9756508146455841, + "flos": 23583399310080.0, + "grad_norm": 1.845626278966021, + "language_loss": 0.72502863, + "learning_rate": 6.186742153410751e-09, + "loss": 0.74713016, + "num_input_tokens_seen": 175221575, + "step": 8114, + "time_per_iteration": 3.689767837524414 + }, + { + "auxiliary_loss_clip": 0.01185294, + "auxiliary_loss_mlp": 0.01026314, + "balance_loss_clip": 1.04653811, + "balance_loss_mlp": 1.01842856, + "epoch": 0.9757710575362232, + "flos": 22966921163520.0, + "grad_norm": 1.8849583254752222, + "language_loss": 0.87534952, + "learning_rate": 6.125670112326453e-09, + "loss": 0.89746559, + "num_input_tokens_seen": 175240835, + "step": 8115, + "time_per_iteration": 2.6147403717041016 + }, + { + "auxiliary_loss_clip": 0.01174569, + "auxiliary_loss_mlp": 0.01023125, + "balance_loss_clip": 1.04289842, + "balance_loss_mlp": 1.01576376, + "epoch": 0.9758913004268622, + "flos": 27966009530880.0, + "grad_norm": 2.740625375858279, + "language_loss": 0.70521444, + "learning_rate": 6.064900540392548e-09, + "loss": 0.72719133, + "num_input_tokens_seen": 175262930, + "step": 8116, + "time_per_iteration": 2.7633488178253174 + }, + { + "auxiliary_loss_clip": 0.01178515, + "auxiliary_loss_mlp": 0.01023497, + "balance_loss_clip": 1.04439485, + "balance_loss_mlp": 1.01696777, + "epoch": 0.9760115433175014, + "flos": 22200156512640.0, + "grad_norm": 2.077355376669016, + "language_loss": 0.78765112, + "learning_rate": 6.0044334468278835e-09, + "loss": 0.80967128, + "num_input_tokens_seen": 175282275, + "step": 8117, + "time_per_iteration": 2.6413660049438477 + }, + { + "auxiliary_loss_clip": 0.01203965, + "auxiliary_loss_mlp": 0.0102732, + "balance_loss_clip": 1.04080629, + "balance_loss_mlp": 1.02031946, + "epoch": 0.9761317862081405, + "flos": 26250736389120.0, + "grad_norm": 1.8171811942136176, + "language_loss": 0.71674776, + "learning_rate": 5.944268840805345e-09, + "loss": 0.73906064, + "num_input_tokens_seen": 175303020, + "step": 8118, + "time_per_iteration": 3.690756320953369 + }, + { + "auxiliary_loss_clip": 0.01187875, + "auxiliary_loss_mlp": 0.0102283, + "balance_loss_clip": 1.04333127, + "balance_loss_mlp": 1.01577842, + "epoch": 0.9762520290987795, + "flos": 26575440359040.0, + "grad_norm": 2.230647851334847, + "language_loss": 0.64219326, + "learning_rate": 5.88440673145163e-09, + "loss": 0.66430032, + "num_input_tokens_seen": 175324070, + "step": 8119, + "time_per_iteration": 2.7646536827087402 + }, + { + "auxiliary_loss_clip": 0.01176887, + "auxiliary_loss_mlp": 0.01020736, + "balance_loss_clip": 1.04948378, + "balance_loss_mlp": 1.01400983, + "epoch": 0.9763722719894187, + "flos": 18005036307840.0, + "grad_norm": 2.4297356848167624, + "language_loss": 0.82626212, + "learning_rate": 5.824847127848142e-09, + "loss": 0.84823835, + "num_input_tokens_seen": 175342595, + "step": 8120, + "time_per_iteration": 3.501160144805908 + }, + { + "auxiliary_loss_clip": 0.01208285, + "auxiliary_loss_mlp": 0.01026045, + "balance_loss_clip": 1.0435586, + "balance_loss_mlp": 1.01873183, + "epoch": 0.9764925148800577, + "flos": 22455660931200.0, + "grad_norm": 1.7851631756628468, + "language_loss": 0.78384984, + "learning_rate": 5.765590039029433e-09, + "loss": 0.80619317, + "num_input_tokens_seen": 175361915, + "step": 8121, + "time_per_iteration": 2.7989747524261475 + }, + { + "auxiliary_loss_clip": 0.01167915, + "auxiliary_loss_mlp": 0.01027946, + "balance_loss_clip": 1.04982901, + "balance_loss_mlp": 1.02092493, + "epoch": 0.9766127577706968, + "flos": 36757084786560.0, + "grad_norm": 2.9313375203544316, + "language_loss": 0.71150565, + "learning_rate": 5.706635473985422e-09, + "loss": 0.7334643, + "num_input_tokens_seen": 175385785, + "step": 8122, + "time_per_iteration": 2.745201349258423 + }, + { + "auxiliary_loss_clip": 0.01174897, + "auxiliary_loss_mlp": 0.01020926, + "balance_loss_clip": 1.04616332, + "balance_loss_mlp": 1.01406884, + "epoch": 0.976733000661336, + "flos": 22309971367680.0, + "grad_norm": 1.8427857247595416, + "language_loss": 0.84708589, + "learning_rate": 5.6479834416591764e-09, + "loss": 0.86904413, + "num_input_tokens_seen": 175405145, + "step": 8123, + "time_per_iteration": 2.654662609100342 + }, + { + "auxiliary_loss_clip": 0.01175462, + "auxiliary_loss_mlp": 0.01378879, + "balance_loss_clip": 1.04697382, + "balance_loss_mlp": 1.00014305, + "epoch": 0.976853243551975, + "flos": 25810938264960.0, + "grad_norm": 1.8484181866885814, + "language_loss": 0.68583524, + "learning_rate": 5.589633950947803e-09, + "loss": 0.71137857, + "num_input_tokens_seen": 175422645, + "step": 8124, + "time_per_iteration": 2.656273126602173 + }, + { + "auxiliary_loss_clip": 0.01180202, + "auxiliary_loss_mlp": 0.0102607, + "balance_loss_clip": 1.04432094, + "balance_loss_mlp": 1.01842904, + "epoch": 0.9769734864426141, + "flos": 21397445326080.0, + "grad_norm": 2.1238313645895563, + "language_loss": 0.69771534, + "learning_rate": 5.5315870107035535e-09, + "loss": 0.71977806, + "num_input_tokens_seen": 175440695, + "step": 8125, + "time_per_iteration": 2.6375086307525635 + }, + { + "auxiliary_loss_clip": 0.01181143, + "auxiliary_loss_mlp": 0.01025146, + "balance_loss_clip": 1.04565048, + "balance_loss_mlp": 1.01811552, + "epoch": 0.9770937293332532, + "flos": 13990977584640.0, + "grad_norm": 1.7994126626332676, + "language_loss": 0.7889663, + "learning_rate": 5.473842629731607e-09, + "loss": 0.8110292, + "num_input_tokens_seen": 175459195, + "step": 8126, + "time_per_iteration": 2.664846420288086 + }, + { + "auxiliary_loss_clip": 0.01191394, + "auxiliary_loss_mlp": 0.01378977, + "balance_loss_clip": 1.04453981, + "balance_loss_mlp": 1.00016892, + "epoch": 0.9772139722238923, + "flos": 17931994001280.0, + "grad_norm": 1.8680678546124607, + "language_loss": 0.7802673, + "learning_rate": 5.416400816792066e-09, + "loss": 0.80597103, + "num_input_tokens_seen": 175476710, + "step": 8127, + "time_per_iteration": 2.658621072769165 + }, + { + "auxiliary_loss_clip": 0.01165693, + "auxiliary_loss_mlp": 0.01019148, + "balance_loss_clip": 1.04567122, + "balance_loss_mlp": 1.0120666, + "epoch": 0.9773342151145313, + "flos": 20446171488000.0, + "grad_norm": 2.3682100599360307, + "language_loss": 0.78522849, + "learning_rate": 5.359261580598407e-09, + "loss": 0.80707687, + "num_input_tokens_seen": 175492550, + "step": 8128, + "time_per_iteration": 2.7112009525299072 + }, + { + "auxiliary_loss_clip": 0.01178908, + "auxiliary_loss_mlp": 0.01023071, + "balance_loss_clip": 1.04806578, + "balance_loss_mlp": 1.01494098, + "epoch": 0.9774544580051704, + "flos": 11837306949120.0, + "grad_norm": 2.339560872442355, + "language_loss": 0.78164536, + "learning_rate": 5.302424929819027e-09, + "loss": 0.80366516, + "num_input_tokens_seen": 175506560, + "step": 8129, + "time_per_iteration": 2.5984718799591064 + }, + { + "auxiliary_loss_clip": 0.01176721, + "auxiliary_loss_mlp": 0.01023453, + "balance_loss_clip": 1.04253745, + "balance_loss_mlp": 1.01597238, + "epoch": 0.9775747008958096, + "flos": 13479932833920.0, + "grad_norm": 2.277504154590759, + "language_loss": 0.73384219, + "learning_rate": 5.24589087307592e-09, + "loss": 0.75584388, + "num_input_tokens_seen": 175524180, + "step": 8130, + "time_per_iteration": 2.672300100326538 + }, + { + "auxiliary_loss_clip": 0.01168331, + "auxiliary_loss_mlp": 0.01021099, + "balance_loss_clip": 1.04648221, + "balance_loss_mlp": 1.01398802, + "epoch": 0.9776949437864486, + "flos": 59532314042880.0, + "grad_norm": 1.6042262435036734, + "language_loss": 0.65067756, + "learning_rate": 5.189659418944891e-09, + "loss": 0.67257184, + "num_input_tokens_seen": 175554355, + "step": 8131, + "time_per_iteration": 2.9813320636749268 + }, + { + "auxiliary_loss_clip": 0.01167799, + "auxiliary_loss_mlp": 0.0102395, + "balance_loss_clip": 1.04842353, + "balance_loss_mlp": 1.01705337, + "epoch": 0.9778151866770877, + "flos": 21178605715200.0, + "grad_norm": 1.8268160648187066, + "language_loss": 0.78333724, + "learning_rate": 5.133730575956674e-09, + "loss": 0.80525476, + "num_input_tokens_seen": 175574025, + "step": 8132, + "time_per_iteration": 2.598093271255493 + }, + { + "auxiliary_loss_clip": 0.01185566, + "auxiliary_loss_mlp": 0.01019721, + "balance_loss_clip": 1.04273546, + "balance_loss_mlp": 1.01254416, + "epoch": 0.9779354295677268, + "flos": 20886795624960.0, + "grad_norm": 2.1742283011178745, + "language_loss": 0.72053444, + "learning_rate": 5.0781043525953696e-09, + "loss": 0.74258733, + "num_input_tokens_seen": 175592090, + "step": 8133, + "time_per_iteration": 2.6609926223754883 + }, + { + "auxiliary_loss_clip": 0.01181122, + "auxiliary_loss_mlp": 0.01026369, + "balance_loss_clip": 1.04667306, + "balance_loss_mlp": 1.01971138, + "epoch": 0.9780556724583659, + "flos": 23440618748160.0, + "grad_norm": 1.7908303026024428, + "language_loss": 0.73532665, + "learning_rate": 5.0227807572995605e-09, + "loss": 0.75740159, + "num_input_tokens_seen": 175614065, + "step": 8134, + "time_per_iteration": 2.7848503589630127 + }, + { + "auxiliary_loss_clip": 0.01184437, + "auxiliary_loss_mlp": 0.01022349, + "balance_loss_clip": 1.04287815, + "balance_loss_mlp": 1.01511931, + "epoch": 0.9781759153490049, + "flos": 20923244951040.0, + "grad_norm": 2.8490748201822456, + "language_loss": 0.67125571, + "learning_rate": 4.967759798461646e-09, + "loss": 0.69332361, + "num_input_tokens_seen": 175632410, + "step": 8135, + "time_per_iteration": 2.6469216346740723 + }, + { + "auxiliary_loss_clip": 0.01165461, + "auxiliary_loss_mlp": 0.01021949, + "balance_loss_clip": 1.04704261, + "balance_loss_mlp": 1.01504397, + "epoch": 0.9782961582396441, + "flos": 28293191539200.0, + "grad_norm": 1.9092881721385666, + "language_loss": 0.74428707, + "learning_rate": 4.913041484428282e-09, + "loss": 0.7661612, + "num_input_tokens_seen": 175652885, + "step": 8136, + "time_per_iteration": 2.6305549144744873 + }, + { + "auxiliary_loss_clip": 0.01180479, + "auxiliary_loss_mlp": 0.01019752, + "balance_loss_clip": 1.04758966, + "balance_loss_mlp": 1.01285541, + "epoch": 0.9784164011302832, + "flos": 25552955808000.0, + "grad_norm": 1.754940586354205, + "language_loss": 0.73989916, + "learning_rate": 4.858625823500384e-09, + "loss": 0.7619015, + "num_input_tokens_seen": 175670585, + "step": 8137, + "time_per_iteration": 3.5746476650238037 + }, + { + "auxiliary_loss_clip": 0.01181623, + "auxiliary_loss_mlp": 0.01028665, + "balance_loss_clip": 1.04656208, + "balance_loss_mlp": 1.02088058, + "epoch": 0.9785366440209222, + "flos": 29965945956480.0, + "grad_norm": 2.1968883009560587, + "language_loss": 0.7338593, + "learning_rate": 4.80451282393246e-09, + "loss": 0.75596225, + "num_input_tokens_seen": 175690570, + "step": 8138, + "time_per_iteration": 2.660250186920166 + }, + { + "auxiliary_loss_clip": 0.01185744, + "auxiliary_loss_mlp": 0.0102143, + "balance_loss_clip": 1.04449534, + "balance_loss_mlp": 1.0138607, + "epoch": 0.9786568869115614, + "flos": 32343591847680.0, + "grad_norm": 1.8330168161875455, + "language_loss": 0.67320037, + "learning_rate": 4.750702493933722e-09, + "loss": 0.69527209, + "num_input_tokens_seen": 175710455, + "step": 8139, + "time_per_iteration": 2.7706589698791504 + }, + { + "auxiliary_loss_clip": 0.0118416, + "auxiliary_loss_mlp": 0.01378375, + "balance_loss_clip": 1.04596126, + "balance_loss_mlp": 1.00008178, + "epoch": 0.9787771298022004, + "flos": 23331414424320.0, + "grad_norm": 3.0428496021986233, + "language_loss": 0.85180736, + "learning_rate": 4.697194841666974e-09, + "loss": 0.8774327, + "num_input_tokens_seen": 175729380, + "step": 8140, + "time_per_iteration": 3.505873918533325 + }, + { + "auxiliary_loss_clip": 0.01175935, + "auxiliary_loss_mlp": 0.01021747, + "balance_loss_clip": 1.04439175, + "balance_loss_mlp": 1.01402211, + "epoch": 0.9788973726928395, + "flos": 21468548298240.0, + "grad_norm": 2.7254500330045834, + "language_loss": 0.81861442, + "learning_rate": 4.6439898752492764e-09, + "loss": 0.84059125, + "num_input_tokens_seen": 175749520, + "step": 8141, + "time_per_iteration": 2.6631619930267334 + }, + { + "auxiliary_loss_clip": 0.01073961, + "auxiliary_loss_mlp": 0.01373785, + "balance_loss_clip": 1.00927329, + "balance_loss_mlp": 0.99982196, + "epoch": 0.9790176155834787, + "flos": 68897459439360.0, + "grad_norm": 0.7511298201729295, + "language_loss": 0.63682908, + "learning_rate": 4.591087602751731e-09, + "loss": 0.66130656, + "num_input_tokens_seen": 175811380, + "step": 8142, + "time_per_iteration": 3.300305128097534 + }, + { + "auxiliary_loss_clip": 0.0117473, + "auxiliary_loss_mlp": 0.01025369, + "balance_loss_clip": 1.0461427, + "balance_loss_mlp": 1.01854432, + "epoch": 0.9791378584741177, + "flos": 21430877909760.0, + "grad_norm": 1.5802615544772103, + "language_loss": 0.71902633, + "learning_rate": 4.538488032199916e-09, + "loss": 0.74102736, + "num_input_tokens_seen": 175829480, + "step": 8143, + "time_per_iteration": 2.6595356464385986 + }, + { + "auxiliary_loss_clip": 0.01179481, + "auxiliary_loss_mlp": 0.01022752, + "balance_loss_clip": 1.04410481, + "balance_loss_mlp": 1.01531386, + "epoch": 0.9792581013647568, + "flos": 20153032594560.0, + "grad_norm": 1.8928172234564316, + "language_loss": 0.69488561, + "learning_rate": 4.486191171572784e-09, + "loss": 0.71690792, + "num_input_tokens_seen": 175846750, + "step": 8144, + "time_per_iteration": 3.6152381896972656 + }, + { + "auxiliary_loss_clip": 0.01180756, + "auxiliary_loss_mlp": 0.01025115, + "balance_loss_clip": 1.04679728, + "balance_loss_mlp": 1.01781631, + "epoch": 0.9793783442553959, + "flos": 23728191033600.0, + "grad_norm": 1.5521210165626662, + "language_loss": 0.77495074, + "learning_rate": 4.434197028803766e-09, + "loss": 0.79700941, + "num_input_tokens_seen": 175865975, + "step": 8145, + "time_per_iteration": 2.711440086364746 + }, + { + "auxiliary_loss_clip": 0.01198641, + "auxiliary_loss_mlp": 0.01032571, + "balance_loss_clip": 1.04234147, + "balance_loss_mlp": 1.02494502, + "epoch": 0.979498587146035, + "flos": 23038742407680.0, + "grad_norm": 1.984915525532278, + "language_loss": 0.82163042, + "learning_rate": 4.3825056117805514e-09, + "loss": 0.84394252, + "num_input_tokens_seen": 175881860, + "step": 8146, + "time_per_iteration": 3.5443344116210938 + }, + { + "auxiliary_loss_clip": 0.01166416, + "auxiliary_loss_mlp": 0.01025773, + "balance_loss_clip": 1.04596865, + "balance_loss_mlp": 1.01806056, + "epoch": 0.979618830036674, + "flos": 14318841951360.0, + "grad_norm": 2.057221829727088, + "language_loss": 0.7949295, + "learning_rate": 4.331116928344425e-09, + "loss": 0.81685138, + "num_input_tokens_seen": 175898175, + "step": 8147, + "time_per_iteration": 2.574799060821533 + }, + { + "auxiliary_loss_clip": 0.01190824, + "auxiliary_loss_mlp": 0.01378878, + "balance_loss_clip": 1.04448748, + "balance_loss_mlp": 1.00013375, + "epoch": 0.9797390729273132, + "flos": 16727514215040.0, + "grad_norm": 1.8671031858168299, + "language_loss": 0.62661326, + "learning_rate": 4.28003098629115e-09, + "loss": 0.65231031, + "num_input_tokens_seen": 175914310, + "step": 8148, + "time_per_iteration": 2.627579689025879 + }, + { + "auxiliary_loss_clip": 0.01187526, + "auxiliary_loss_mlp": 0.01024255, + "balance_loss_clip": 1.03655636, + "balance_loss_mlp": 1.01700068, + "epoch": 0.9798593158179523, + "flos": 24532661986560.0, + "grad_norm": 1.8321311094022805, + "language_loss": 0.78558713, + "learning_rate": 4.229247793370305e-09, + "loss": 0.80770493, + "num_input_tokens_seen": 175933435, + "step": 8149, + "time_per_iteration": 2.7161800861358643 + }, + { + "auxiliary_loss_clip": 0.01170322, + "auxiliary_loss_mlp": 0.01024094, + "balance_loss_clip": 1.04928625, + "balance_loss_mlp": 1.01702452, + "epoch": 0.9799795587085913, + "flos": 27308808339840.0, + "grad_norm": 1.6649992096125426, + "language_loss": 0.70020872, + "learning_rate": 4.178767357285951e-09, + "loss": 0.72215289, + "num_input_tokens_seen": 175955065, + "step": 8150, + "time_per_iteration": 2.607344388961792 + }, + { + "auxiliary_loss_clip": 0.01178159, + "auxiliary_loss_mlp": 0.01378541, + "balance_loss_clip": 1.04669118, + "balance_loss_mlp": 1.00023174, + "epoch": 0.9800998015992305, + "flos": 26286575184000.0, + "grad_norm": 2.148142189426372, + "language_loss": 0.71717185, + "learning_rate": 4.128589685695516e-09, + "loss": 0.74273884, + "num_input_tokens_seen": 175975490, + "step": 8151, + "time_per_iteration": 2.685570001602173 + }, + { + "auxiliary_loss_clip": 0.01167896, + "auxiliary_loss_mlp": 0.01022303, + "balance_loss_clip": 1.04717112, + "balance_loss_mlp": 1.01551366, + "epoch": 0.9802200444898695, + "flos": 16723635546240.0, + "grad_norm": 1.817021794060209, + "language_loss": 0.8410688, + "learning_rate": 4.078714786211135e-09, + "loss": 0.86297083, + "num_input_tokens_seen": 175991340, + "step": 8152, + "time_per_iteration": 2.5147666931152344 + }, + { + "auxiliary_loss_clip": 0.01175408, + "auxiliary_loss_mlp": 0.01022305, + "balance_loss_clip": 1.04654825, + "balance_loss_mlp": 1.01512861, + "epoch": 0.9803402873805086, + "flos": 24900459298560.0, + "grad_norm": 1.670339576918651, + "language_loss": 0.76432467, + "learning_rate": 4.029142666398977e-09, + "loss": 0.78630179, + "num_input_tokens_seen": 176011505, + "step": 8153, + "time_per_iteration": 2.6655101776123047 + }, + { + "auxiliary_loss_clip": 0.0116412, + "auxiliary_loss_mlp": 0.01021276, + "balance_loss_clip": 1.04671407, + "balance_loss_mlp": 1.01467204, + "epoch": 0.9804605302711478, + "flos": 22564937082240.0, + "grad_norm": 1.8310761234429906, + "language_loss": 0.80073798, + "learning_rate": 3.979873333778805e-09, + "loss": 0.82259196, + "num_input_tokens_seen": 176029680, + "step": 8154, + "time_per_iteration": 2.5850141048431396 + }, + { + "auxiliary_loss_clip": 0.01194085, + "auxiliary_loss_mlp": 0.01025199, + "balance_loss_clip": 1.04784203, + "balance_loss_mlp": 1.01767087, + "epoch": 0.9805807731617868, + "flos": 38905368382080.0, + "grad_norm": 2.3985405557851256, + "language_loss": 0.73978412, + "learning_rate": 3.930906795824862e-09, + "loss": 0.76197696, + "num_input_tokens_seen": 176050355, + "step": 8155, + "time_per_iteration": 2.8229258060455322 + }, + { + "auxiliary_loss_clip": 0.01170566, + "auxiliary_loss_mlp": 0.01025149, + "balance_loss_clip": 1.04280186, + "balance_loss_mlp": 1.01834273, + "epoch": 0.9807010160524259, + "flos": 17821999578240.0, + "grad_norm": 1.854808790936603, + "language_loss": 0.76694882, + "learning_rate": 3.882243059965207e-09, + "loss": 0.78890598, + "num_input_tokens_seen": 176068070, + "step": 8156, + "time_per_iteration": 2.6135101318359375 + }, + { + "auxiliary_loss_clip": 0.01165298, + "auxiliary_loss_mlp": 0.01023361, + "balance_loss_clip": 1.04215455, + "balance_loss_mlp": 1.01557112, + "epoch": 0.980821258943065, + "flos": 13552975140480.0, + "grad_norm": 2.837776291185424, + "language_loss": 0.66144902, + "learning_rate": 3.833882133582156e-09, + "loss": 0.68333554, + "num_input_tokens_seen": 176083730, + "step": 8157, + "time_per_iteration": 2.7204701900482178 + }, + { + "auxiliary_loss_clip": 0.01178875, + "auxiliary_loss_mlp": 0.01023569, + "balance_loss_clip": 1.04602981, + "balance_loss_mlp": 1.0167681, + "epoch": 0.9809415018337041, + "flos": 21689794120320.0, + "grad_norm": 1.5606343499085202, + "language_loss": 0.78079349, + "learning_rate": 3.785824024012285e-09, + "loss": 0.80281794, + "num_input_tokens_seen": 176102730, + "step": 8158, + "time_per_iteration": 2.660689353942871 + }, + { + "auxiliary_loss_clip": 0.01178245, + "auxiliary_loss_mlp": 0.01022785, + "balance_loss_clip": 1.04589295, + "balance_loss_mlp": 1.01578462, + "epoch": 0.9810617447243432, + "flos": 23294857357440.0, + "grad_norm": 1.4636852305370756, + "language_loss": 0.78341234, + "learning_rate": 3.738068738545541e-09, + "loss": 0.80542266, + "num_input_tokens_seen": 176121815, + "step": 8159, + "time_per_iteration": 2.6831917762756348 + }, + { + "auxiliary_loss_clip": 0.0117911, + "auxiliary_loss_mlp": 0.01025095, + "balance_loss_clip": 1.04547608, + "balance_loss_mlp": 1.01787651, + "epoch": 0.9811819876149822, + "flos": 18332038748160.0, + "grad_norm": 2.7877894228756874, + "language_loss": 0.78595728, + "learning_rate": 3.6906162844265733e-09, + "loss": 0.80799931, + "num_input_tokens_seen": 176138900, + "step": 8160, + "time_per_iteration": 2.6458044052124023 + }, + { + "auxiliary_loss_clip": 0.01181008, + "auxiliary_loss_mlp": 0.0102811, + "balance_loss_clip": 1.04419851, + "balance_loss_mlp": 1.0203259, + "epoch": 0.9813022305056214, + "flos": 22601961025920.0, + "grad_norm": 1.9574457826486655, + "language_loss": 0.70784724, + "learning_rate": 3.643466668853845e-09, + "loss": 0.72993839, + "num_input_tokens_seen": 176156925, + "step": 8161, + "time_per_iteration": 2.5926992893218994 + }, + { + "auxiliary_loss_clip": 0.01185055, + "auxiliary_loss_mlp": 0.01024294, + "balance_loss_clip": 1.04322922, + "balance_loss_mlp": 1.01697123, + "epoch": 0.9814224733962604, + "flos": 25413335642880.0, + "grad_norm": 2.656604954410967, + "language_loss": 0.75275677, + "learning_rate": 3.59661989898008e-09, + "loss": 0.77485025, + "num_input_tokens_seen": 176177980, + "step": 8162, + "time_per_iteration": 3.7035608291625977 + }, + { + "auxiliary_loss_clip": 0.0118438, + "auxiliary_loss_mlp": 0.01024318, + "balance_loss_clip": 1.04260743, + "balance_loss_mlp": 1.01787484, + "epoch": 0.9815427162868995, + "flos": 25007185584000.0, + "grad_norm": 1.620600016386881, + "language_loss": 0.76675308, + "learning_rate": 3.5500759819115934e-09, + "loss": 0.78884006, + "num_input_tokens_seen": 176198345, + "step": 8163, + "time_per_iteration": 2.7131972312927246 + }, + { + "auxiliary_loss_clip": 0.01170636, + "auxiliary_loss_mlp": 0.01023939, + "balance_loss_clip": 1.05052054, + "balance_loss_mlp": 1.01694405, + "epoch": 0.9816629591775387, + "flos": 20662604887680.0, + "grad_norm": 1.8045155389645016, + "language_loss": 0.80896223, + "learning_rate": 3.5038349247094034e-09, + "loss": 0.83090794, + "num_input_tokens_seen": 176215605, + "step": 8164, + "time_per_iteration": 2.561910629272461 + }, + { + "auxiliary_loss_clip": 0.01184078, + "auxiliary_loss_mlp": 0.01026301, + "balance_loss_clip": 1.04321623, + "balance_loss_mlp": 1.0192523, + "epoch": 0.9817832020681777, + "flos": 17712220636800.0, + "grad_norm": 2.3673407339921186, + "language_loss": 0.77282816, + "learning_rate": 3.4578967343878994e-09, + "loss": 0.79493189, + "num_input_tokens_seen": 176231810, + "step": 8165, + "time_per_iteration": 2.614224910736084 + }, + { + "auxiliary_loss_clip": 0.01185071, + "auxiliary_loss_mlp": 0.0102529, + "balance_loss_clip": 1.044945, + "balance_loss_mlp": 1.01836395, + "epoch": 0.9819034449588168, + "flos": 22530032040960.0, + "grad_norm": 1.7381880222856942, + "language_loss": 0.80949593, + "learning_rate": 3.4122614179161733e-09, + "loss": 0.83159953, + "num_input_tokens_seen": 176251770, + "step": 8166, + "time_per_iteration": 3.568035364151001 + }, + { + "auxiliary_loss_clip": 0.01178902, + "auxiliary_loss_mlp": 0.01026278, + "balance_loss_clip": 1.03910923, + "balance_loss_mlp": 1.01945925, + "epoch": 0.9820236878494559, + "flos": 20011221699840.0, + "grad_norm": 1.6767710662693383, + "language_loss": 0.78172684, + "learning_rate": 3.36692898221691e-09, + "loss": 0.80377865, + "num_input_tokens_seen": 176270135, + "step": 8167, + "time_per_iteration": 2.745534896850586 + }, + { + "auxiliary_loss_clip": 0.0117722, + "auxiliary_loss_mlp": 0.01022522, + "balance_loss_clip": 1.04632115, + "balance_loss_mlp": 1.01609373, + "epoch": 0.982143930740095, + "flos": 18807316531200.0, + "grad_norm": 1.6940913907974542, + "language_loss": 0.73424059, + "learning_rate": 3.3218994341668305e-09, + "loss": 0.75623804, + "num_input_tokens_seen": 176289065, + "step": 8168, + "time_per_iteration": 2.608076810836792 + }, + { + "auxiliary_loss_clip": 0.01166449, + "auxiliary_loss_mlp": 0.01023655, + "balance_loss_clip": 1.04854989, + "balance_loss_mlp": 1.0172807, + "epoch": 0.982264173630734, + "flos": 26578026138240.0, + "grad_norm": 1.664801790447085, + "language_loss": 0.75342745, + "learning_rate": 3.2771727805971373e-09, + "loss": 0.77532852, + "num_input_tokens_seen": 176310450, + "step": 8169, + "time_per_iteration": 2.674191474914551 + }, + { + "auxiliary_loss_clip": 0.01194167, + "auxiliary_loss_mlp": 0.01027451, + "balance_loss_clip": 1.03817225, + "balance_loss_mlp": 1.01998234, + "epoch": 0.9823844165213732, + "flos": 22014462176640.0, + "grad_norm": 1.6828712653042561, + "language_loss": 0.76969504, + "learning_rate": 3.232749028292847e-09, + "loss": 0.79191118, + "num_input_tokens_seen": 176327415, + "step": 8170, + "time_per_iteration": 3.5818653106689453 + }, + { + "auxiliary_loss_clip": 0.01167409, + "auxiliary_loss_mlp": 0.01023217, + "balance_loss_clip": 1.0461731, + "balance_loss_mlp": 1.0159421, + "epoch": 0.9825046594120123, + "flos": 21908166854400.0, + "grad_norm": 1.746670418738227, + "language_loss": 0.88504869, + "learning_rate": 3.188628183992792e-09, + "loss": 0.90695494, + "num_input_tokens_seen": 176347680, + "step": 8171, + "time_per_iteration": 2.597752332687378 + }, + { + "auxiliary_loss_clip": 0.01073592, + "auxiliary_loss_mlp": 0.01002314, + "balance_loss_clip": 1.00862908, + "balance_loss_mlp": 1.00127721, + "epoch": 0.9826249023026513, + "flos": 59494610718720.0, + "grad_norm": 0.7375443290746198, + "language_loss": 0.62566406, + "learning_rate": 3.1448102543902844e-09, + "loss": 0.6464231, + "num_input_tokens_seen": 176411595, + "step": 8172, + "time_per_iteration": 4.084103345870972 + }, + { + "auxiliary_loss_clip": 0.0117872, + "auxiliary_loss_mlp": 0.01024395, + "balance_loss_clip": 1.04600787, + "balance_loss_mlp": 1.01781178, + "epoch": 0.9827451451932905, + "flos": 16071031296000.0, + "grad_norm": 1.8437741986892098, + "language_loss": 0.67290282, + "learning_rate": 3.1012952461324515e-09, + "loss": 0.69493401, + "num_input_tokens_seen": 176430570, + "step": 8173, + "time_per_iteration": 2.6712565422058105 + }, + { + "auxiliary_loss_clip": 0.01174511, + "auxiliary_loss_mlp": 0.01024261, + "balance_loss_clip": 1.04769969, + "balance_loss_mlp": 1.01749849, + "epoch": 0.9828653880839295, + "flos": 20262775622400.0, + "grad_norm": 2.317559572061147, + "language_loss": 0.73874009, + "learning_rate": 3.0580831658204575e-09, + "loss": 0.76072776, + "num_input_tokens_seen": 176448150, + "step": 8174, + "time_per_iteration": 2.639312505722046 + }, + { + "auxiliary_loss_clip": 0.01175335, + "auxiliary_loss_mlp": 0.01029569, + "balance_loss_clip": 1.04794943, + "balance_loss_mlp": 1.02257192, + "epoch": 0.9829856309745686, + "flos": 21616141282560.0, + "grad_norm": 1.6985280477393705, + "language_loss": 0.77801013, + "learning_rate": 3.015174020009281e-09, + "loss": 0.8000592, + "num_input_tokens_seen": 176467475, + "step": 8175, + "time_per_iteration": 2.5947630405426025 + }, + { + "auxiliary_loss_clip": 0.01194271, + "auxiliary_loss_mlp": 0.01026522, + "balance_loss_clip": 1.0407629, + "balance_loss_mlp": 1.01973271, + "epoch": 0.9831058738652078, + "flos": 23764209396480.0, + "grad_norm": 1.882900851653232, + "language_loss": 0.74979782, + "learning_rate": 2.9725678152086043e-09, + "loss": 0.77200574, + "num_input_tokens_seen": 176486045, + "step": 8176, + "time_per_iteration": 2.6888725757598877 + }, + { + "auxiliary_loss_clip": 0.01170504, + "auxiliary_loss_mlp": 0.01025898, + "balance_loss_clip": 1.04237342, + "balance_loss_mlp": 1.018659, + "epoch": 0.9832261167558468, + "flos": 11320911072000.0, + "grad_norm": 4.6598027579293, + "language_loss": 0.82817876, + "learning_rate": 2.930264557881257e-09, + "loss": 0.85014284, + "num_input_tokens_seen": 176501230, + "step": 8177, + "time_per_iteration": 2.5639748573303223 + }, + { + "auxiliary_loss_clip": 0.01059441, + "auxiliary_loss_mlp": 0.0100286, + "balance_loss_clip": 1.00833988, + "balance_loss_mlp": 1.00182867, + "epoch": 0.9833463596464859, + "flos": 60000304343040.0, + "grad_norm": 0.8615957806259525, + "language_loss": 0.58203036, + "learning_rate": 2.8882642544452163e-09, + "loss": 0.60265338, + "num_input_tokens_seen": 176565955, + "step": 8178, + "time_per_iteration": 3.2011444568634033 + }, + { + "auxiliary_loss_clip": 0.01171621, + "auxiliary_loss_mlp": 0.01024668, + "balance_loss_clip": 1.04173946, + "balance_loss_mlp": 1.01716065, + "epoch": 0.983466602537125, + "flos": 13626699805440.0, + "grad_norm": 2.4043262092963906, + "language_loss": 0.74322587, + "learning_rate": 2.8465669112716083e-09, + "loss": 0.76518881, + "num_input_tokens_seen": 176583480, + "step": 8179, + "time_per_iteration": 2.7438507080078125 + }, + { + "auxiliary_loss_clip": 0.01176138, + "auxiliary_loss_mlp": 0.01378852, + "balance_loss_clip": 1.0451448, + "balance_loss_mlp": 1.00012279, + "epoch": 0.9835868454277641, + "flos": 22926844563840.0, + "grad_norm": 1.848041080568616, + "language_loss": 0.76552165, + "learning_rate": 2.8051725346858177e-09, + "loss": 0.79107153, + "num_input_tokens_seen": 176603740, + "step": 8180, + "time_per_iteration": 2.6186625957489014 + }, + { + "auxiliary_loss_clip": 0.01169381, + "auxiliary_loss_mlp": 0.01022428, + "balance_loss_clip": 1.04714668, + "balance_loss_mlp": 1.01465559, + "epoch": 0.9837070883184031, + "flos": 27673409341440.0, + "grad_norm": 2.44805092168767, + "language_loss": 0.70601964, + "learning_rate": 2.7640811309674883e-09, + "loss": 0.7279377, + "num_input_tokens_seen": 176623240, + "step": 8181, + "time_per_iteration": 2.6758062839508057 + }, + { + "auxiliary_loss_clip": 0.01184134, + "auxiliary_loss_mlp": 0.01023756, + "balance_loss_clip": 1.04233921, + "balance_loss_mlp": 1.01703227, + "epoch": 0.9838273312090423, + "flos": 29241951425280.0, + "grad_norm": 1.6233628568778975, + "language_loss": 0.80684894, + "learning_rate": 2.7232927063498557e-09, + "loss": 0.82892787, + "num_input_tokens_seen": 176643615, + "step": 8182, + "time_per_iteration": 2.811366558074951 + }, + { + "auxiliary_loss_clip": 0.0117802, + "auxiliary_loss_mlp": 0.01031178, + "balance_loss_clip": 1.04580569, + "balance_loss_mlp": 1.02387917, + "epoch": 0.9839475740996814, + "flos": 40110207304320.0, + "grad_norm": 1.8081966423663087, + "language_loss": 0.69098842, + "learning_rate": 2.682807267020859e-09, + "loss": 0.71308041, + "num_input_tokens_seen": 176666375, + "step": 8183, + "time_per_iteration": 2.7592101097106934 + }, + { + "auxiliary_loss_clip": 0.01176563, + "auxiliary_loss_mlp": 0.01025347, + "balance_loss_clip": 1.0465548, + "balance_loss_mlp": 1.01781297, + "epoch": 0.9840678169903204, + "flos": 24169389788160.0, + "grad_norm": 1.6080297243417396, + "language_loss": 0.62519181, + "learning_rate": 2.642624819121808e-09, + "loss": 0.64721096, + "num_input_tokens_seen": 176686525, + "step": 8184, + "time_per_iteration": 2.5905227661132812 + }, + { + "auxiliary_loss_clip": 0.01184656, + "auxiliary_loss_mlp": 0.01027333, + "balance_loss_clip": 1.04606724, + "balance_loss_mlp": 1.0206964, + "epoch": 0.9841880598809596, + "flos": 14684484447360.0, + "grad_norm": 2.0473214610283117, + "language_loss": 0.61886048, + "learning_rate": 2.6027453687487154e-09, + "loss": 0.64098042, + "num_input_tokens_seen": 176703615, + "step": 8185, + "time_per_iteration": 2.6833956241607666 + }, + { + "auxiliary_loss_clip": 0.01187168, + "auxiliary_loss_mlp": 0.01027326, + "balance_loss_clip": 1.04521227, + "balance_loss_mlp": 1.01970863, + "epoch": 0.9843083027715986, + "flos": 22344768668160.0, + "grad_norm": 2.339712805467144, + "language_loss": 0.53865016, + "learning_rate": 2.5631689219509643e-09, + "loss": 0.56079507, + "num_input_tokens_seen": 176722295, + "step": 8186, + "time_per_iteration": 2.6535589694976807 + }, + { + "auxiliary_loss_clip": 0.01183952, + "auxiliary_loss_mlp": 0.01023204, + "balance_loss_clip": 1.04526758, + "balance_loss_mlp": 1.01642656, + "epoch": 0.9844285456622377, + "flos": 21800111765760.0, + "grad_norm": 1.6581913219823996, + "language_loss": 0.83339113, + "learning_rate": 2.523895484732197e-09, + "loss": 0.85546267, + "num_input_tokens_seen": 176741750, + "step": 8187, + "time_per_iteration": 2.6890547275543213 + }, + { + "auxiliary_loss_clip": 0.01181126, + "auxiliary_loss_mlp": 0.01027303, + "balance_loss_clip": 1.04564691, + "balance_loss_mlp": 1.01956046, + "epoch": 0.9845487885528769, + "flos": 18035380321920.0, + "grad_norm": 2.001929502445723, + "language_loss": 0.75220889, + "learning_rate": 2.4849250630505357e-09, + "loss": 0.77429318, + "num_input_tokens_seen": 176759995, + "step": 8188, + "time_per_iteration": 3.5824944972991943 + }, + { + "auxiliary_loss_clip": 0.01208908, + "auxiliary_loss_mlp": 0.01023361, + "balance_loss_clip": 1.03821921, + "balance_loss_mlp": 1.01584458, + "epoch": 0.9846690314435159, + "flos": 25228610974080.0, + "grad_norm": 2.4290171361189117, + "language_loss": 0.73395789, + "learning_rate": 2.4462576628172528e-09, + "loss": 0.7562806, + "num_input_tokens_seen": 176778625, + "step": 8189, + "time_per_iteration": 2.772386312484741 + }, + { + "auxiliary_loss_clip": 0.01174674, + "auxiliary_loss_mlp": 0.01024635, + "balance_loss_clip": 1.04687834, + "balance_loss_mlp": 1.01743495, + "epoch": 0.984789274334155, + "flos": 18552171248640.0, + "grad_norm": 2.444515862929819, + "language_loss": 0.74344361, + "learning_rate": 2.407893289898766e-09, + "loss": 0.76543665, + "num_input_tokens_seen": 176797655, + "step": 8190, + "time_per_iteration": 2.7096385955810547 + }, + { + "auxiliary_loss_clip": 0.01182078, + "auxiliary_loss_mlp": 0.0102434, + "balance_loss_clip": 1.04001808, + "balance_loss_mlp": 1.01694274, + "epoch": 0.984909517224794, + "flos": 27345437233920.0, + "grad_norm": 1.8233157231016097, + "language_loss": 0.83768666, + "learning_rate": 2.3698319501144202e-09, + "loss": 0.85975081, + "num_input_tokens_seen": 176818640, + "step": 8191, + "time_per_iteration": 2.723256826400757 + }, + { + "auxiliary_loss_clip": 0.01181982, + "auxiliary_loss_mlp": 0.01027724, + "balance_loss_clip": 1.04612458, + "balance_loss_mlp": 1.02013671, + "epoch": 0.9850297601154332, + "flos": 18734058743040.0, + "grad_norm": 1.718872803853701, + "language_loss": 0.73269361, + "learning_rate": 2.3320736492382644e-09, + "loss": 0.75479072, + "num_input_tokens_seen": 176837475, + "step": 8192, + "time_per_iteration": 3.5390846729278564 + }, + { + "auxiliary_loss_clip": 0.01164946, + "auxiliary_loss_mlp": 0.01023564, + "balance_loss_clip": 1.04740572, + "balance_loss_mlp": 1.01683748, + "epoch": 0.9851500030060723, + "flos": 22308247514880.0, + "grad_norm": 1.8868765203865892, + "language_loss": 0.68035388, + "learning_rate": 2.29461839299816e-09, + "loss": 0.70223898, + "num_input_tokens_seen": 176857190, + "step": 8193, + "time_per_iteration": 2.6356780529022217 + }, + { + "auxiliary_loss_clip": 0.01194939, + "auxiliary_loss_mlp": 0.01023556, + "balance_loss_clip": 1.04282737, + "balance_loss_mlp": 1.0167551, + "epoch": 0.9852702458967113, + "flos": 26353691746560.0, + "grad_norm": 1.5837038890620243, + "language_loss": 0.79823667, + "learning_rate": 2.257466187076229e-09, + "loss": 0.82042164, + "num_input_tokens_seen": 176876395, + "step": 8194, + "time_per_iteration": 2.7548089027404785 + }, + { + "auxiliary_loss_clip": 0.01181181, + "auxiliary_loss_mlp": 0.01378719, + "balance_loss_clip": 1.04600382, + "balance_loss_mlp": 1.00013804, + "epoch": 0.9853904887873505, + "flos": 20883599314560.0, + "grad_norm": 1.7872120429366176, + "language_loss": 0.71312457, + "learning_rate": 2.2206170371081854e-09, + "loss": 0.73872358, + "num_input_tokens_seen": 176894980, + "step": 8195, + "time_per_iteration": 2.653217315673828 + }, + { + "auxiliary_loss_clip": 0.0118397, + "auxiliary_loss_mlp": 0.01022661, + "balance_loss_clip": 1.04289985, + "balance_loss_mlp": 1.0153234, + "epoch": 0.9855107316779895, + "flos": 25263444188160.0, + "grad_norm": 1.6640298094988644, + "language_loss": 0.84702671, + "learning_rate": 2.1840709486842247e-09, + "loss": 0.86909306, + "num_input_tokens_seen": 176914600, + "step": 8196, + "time_per_iteration": 3.516422748565674 + }, + { + "auxiliary_loss_clip": 0.01176268, + "auxiliary_loss_mlp": 0.01028167, + "balance_loss_clip": 1.04270804, + "balance_loss_mlp": 1.02075839, + "epoch": 0.9856309745686286, + "flos": 19062102677760.0, + "grad_norm": 1.8349408482560423, + "language_loss": 0.79231858, + "learning_rate": 2.1478279273481335e-09, + "loss": 0.81436288, + "num_input_tokens_seen": 176933085, + "step": 8197, + "time_per_iteration": 2.733304023742676 + }, + { + "auxiliary_loss_clip": 0.01178563, + "auxiliary_loss_mlp": 0.01024117, + "balance_loss_clip": 1.04945707, + "balance_loss_mlp": 1.01722622, + "epoch": 0.9857512174592677, + "flos": 34130758060800.0, + "grad_norm": 2.59026166154589, + "language_loss": 0.80082798, + "learning_rate": 2.1118879785981815e-09, + "loss": 0.82285476, + "num_input_tokens_seen": 176953225, + "step": 8198, + "time_per_iteration": 2.7242214679718018 + }, + { + "auxiliary_loss_clip": 0.01183476, + "auxiliary_loss_mlp": 0.01022104, + "balance_loss_clip": 1.04448378, + "balance_loss_mlp": 1.01528263, + "epoch": 0.9858714603499068, + "flos": 25994693266560.0, + "grad_norm": 1.793047232473552, + "language_loss": 0.79347861, + "learning_rate": 2.0762511078862288e-09, + "loss": 0.81553435, + "num_input_tokens_seen": 176973570, + "step": 8199, + "time_per_iteration": 3.500659465789795 + }, + { + "auxiliary_loss_clip": 0.01192703, + "auxiliary_loss_mlp": 0.0102332, + "balance_loss_clip": 1.04559636, + "balance_loss_mlp": 1.01595569, + "epoch": 0.9859917032405459, + "flos": 23696230907520.0, + "grad_norm": 2.899379761670249, + "language_loss": 0.64809453, + "learning_rate": 2.0409173206186183e-09, + "loss": 0.67025477, + "num_input_tokens_seen": 176992810, + "step": 8200, + "time_per_iteration": 2.761871337890625 + }, + { + "auxiliary_loss_clip": 0.01191569, + "auxiliary_loss_mlp": 0.01025436, + "balance_loss_clip": 1.04565597, + "balance_loss_mlp": 1.01849556, + "epoch": 0.986111946131185, + "flos": 19938287134080.0, + "grad_norm": 2.4469548131838366, + "language_loss": 0.86944544, + "learning_rate": 2.0058866221550617e-09, + "loss": 0.89161545, + "num_input_tokens_seen": 177011050, + "step": 8201, + "time_per_iteration": 2.650571823120117 + }, + { + "auxiliary_loss_clip": 0.01165709, + "auxiliary_loss_mlp": 0.01028739, + "balance_loss_clip": 1.04461479, + "balance_loss_mlp": 1.02124703, + "epoch": 0.9862321890218241, + "flos": 19828831415040.0, + "grad_norm": 1.974588350204907, + "language_loss": 0.74983692, + "learning_rate": 1.971159017809976e-09, + "loss": 0.77178133, + "num_input_tokens_seen": 177029340, + "step": 8202, + "time_per_iteration": 2.574781656265259 + }, + { + "auxiliary_loss_clip": 0.01173606, + "auxiliary_loss_mlp": 0.01025249, + "balance_loss_clip": 1.04539728, + "balance_loss_mlp": 1.0179292, + "epoch": 0.9863524319124631, + "flos": 21652051904640.0, + "grad_norm": 2.23902069462709, + "language_loss": 0.77434671, + "learning_rate": 1.93673451285159e-09, + "loss": 0.79633528, + "num_input_tokens_seen": 177048390, + "step": 8203, + "time_per_iteration": 2.641221284866333 + }, + { + "auxiliary_loss_clip": 0.01087879, + "auxiliary_loss_mlp": 0.01000957, + "balance_loss_clip": 1.00971961, + "balance_loss_mlp": 0.99990219, + "epoch": 0.9864726748031023, + "flos": 52769977920000.0, + "grad_norm": 0.7445506829154726, + "language_loss": 0.56568539, + "learning_rate": 1.9026131125019495e-09, + "loss": 0.58657372, + "num_input_tokens_seen": 177105760, + "step": 8204, + "time_per_iteration": 3.147986888885498 + }, + { + "auxiliary_loss_clip": 0.01171995, + "auxiliary_loss_mlp": 0.01022667, + "balance_loss_clip": 1.04711127, + "balance_loss_mlp": 1.01576447, + "epoch": 0.9865929176937414, + "flos": 23364631526400.0, + "grad_norm": 1.7318722987241366, + "language_loss": 0.87003279, + "learning_rate": 1.8687948219371363e-09, + "loss": 0.8919794, + "num_input_tokens_seen": 177124985, + "step": 8205, + "time_per_iteration": 2.590672016143799 + }, + { + "auxiliary_loss_clip": 0.01171298, + "auxiliary_loss_mlp": 0.01027152, + "balance_loss_clip": 1.04696822, + "balance_loss_mlp": 1.01912355, + "epoch": 0.9867131605843804, + "flos": 21616679986560.0, + "grad_norm": 1.8877533142829543, + "language_loss": 0.8855707, + "learning_rate": 1.835279646287491e-09, + "loss": 0.90755522, + "num_input_tokens_seen": 177142995, + "step": 8206, + "time_per_iteration": 2.617561101913452 + }, + { + "auxiliary_loss_clip": 0.01181358, + "auxiliary_loss_mlp": 0.01025082, + "balance_loss_clip": 1.04697251, + "balance_loss_mlp": 1.01706505, + "epoch": 0.9868334034750196, + "flos": 22271403139200.0, + "grad_norm": 1.8488673377041485, + "language_loss": 0.76348877, + "learning_rate": 1.8020675906371685e-09, + "loss": 0.78555316, + "num_input_tokens_seen": 177162390, + "step": 8207, + "time_per_iteration": 2.6078755855560303 + }, + { + "auxiliary_loss_clip": 0.01201751, + "auxiliary_loss_mlp": 0.01024725, + "balance_loss_clip": 1.04363811, + "balance_loss_mlp": 1.01775455, + "epoch": 0.9869536463656586, + "flos": 25809573548160.0, + "grad_norm": 2.0653553776454525, + "language_loss": 0.75229466, + "learning_rate": 1.7691586600243612e-09, + "loss": 0.7745595, + "num_input_tokens_seen": 177181290, + "step": 8208, + "time_per_iteration": 2.7348906993865967 + }, + { + "auxiliary_loss_clip": 0.01183333, + "auxiliary_loss_mlp": 0.01030074, + "balance_loss_clip": 1.04559731, + "balance_loss_mlp": 1.02266455, + "epoch": 0.9870738892562977, + "flos": 16398500613120.0, + "grad_norm": 2.4083616339285645, + "language_loss": 0.86834085, + "learning_rate": 1.7365528594415202e-09, + "loss": 0.89047492, + "num_input_tokens_seen": 177195360, + "step": 8209, + "time_per_iteration": 2.647700786590576 + }, + { + "auxiliary_loss_clip": 0.01179574, + "auxiliary_loss_mlp": 0.01378942, + "balance_loss_clip": 1.04462111, + "balance_loss_mlp": 1.00015807, + "epoch": 0.9871941321469369, + "flos": 35481358373760.0, + "grad_norm": 1.63048934058672, + "language_loss": 0.6763128, + "learning_rate": 1.7042501938346888e-09, + "loss": 0.70189798, + "num_input_tokens_seen": 177218090, + "step": 8210, + "time_per_iteration": 2.7490975856781006 + }, + { + "auxiliary_loss_clip": 0.01165992, + "auxiliary_loss_mlp": 0.01022415, + "balance_loss_clip": 1.03795886, + "balance_loss_mlp": 1.01566207, + "epoch": 0.9873143750375759, + "flos": 21434217874560.0, + "grad_norm": 1.998398295394396, + "language_loss": 0.76409191, + "learning_rate": 1.6722506681043913e-09, + "loss": 0.78597593, + "num_input_tokens_seen": 177237050, + "step": 8211, + "time_per_iteration": 2.6522607803344727 + }, + { + "auxiliary_loss_clip": 0.01189106, + "auxiliary_loss_mlp": 0.01023156, + "balance_loss_clip": 1.04506159, + "balance_loss_mlp": 1.01588404, + "epoch": 0.987434617928215, + "flos": 16326499800960.0, + "grad_norm": 2.2580099969757796, + "language_loss": 0.69428527, + "learning_rate": 1.640554287104745e-09, + "loss": 0.7164079, + "num_input_tokens_seen": 177255325, + "step": 8212, + "time_per_iteration": 2.645089864730835 + }, + { + "auxiliary_loss_clip": 0.01194071, + "auxiliary_loss_mlp": 0.01025126, + "balance_loss_clip": 1.04014063, + "balance_loss_mlp": 1.01732409, + "epoch": 0.9875548608188541, + "flos": 17851984456320.0, + "grad_norm": 2.531761105230466, + "language_loss": 0.80231476, + "learning_rate": 1.609161055644348e-09, + "loss": 0.82450676, + "num_input_tokens_seen": 177271250, + "step": 8213, + "time_per_iteration": 2.6632680892944336 + }, + { + "auxiliary_loss_clip": 0.01179836, + "auxiliary_loss_mlp": 0.01019887, + "balance_loss_clip": 1.04405856, + "balance_loss_mlp": 1.01288593, + "epoch": 0.9876751037094932, + "flos": 26132876887680.0, + "grad_norm": 2.033104334149321, + "language_loss": 0.68806875, + "learning_rate": 1.5780709784849467e-09, + "loss": 0.71006602, + "num_input_tokens_seen": 177288270, + "step": 8214, + "time_per_iteration": 2.6183319091796875 + }, + { + "auxiliary_loss_clip": 0.01218237, + "auxiliary_loss_mlp": 0.01026259, + "balance_loss_clip": 1.0437659, + "balance_loss_mlp": 1.01907372, + "epoch": 0.9877953466001322, + "flos": 15991344973440.0, + "grad_norm": 1.8798605209925616, + "language_loss": 0.82394421, + "learning_rate": 1.5472840603436565e-09, + "loss": 0.84638917, + "num_input_tokens_seen": 177305500, + "step": 8215, + "time_per_iteration": 3.6705453395843506 + }, + { + "auxiliary_loss_clip": 0.01187489, + "auxiliary_loss_mlp": 0.01022692, + "balance_loss_clip": 1.04586518, + "balance_loss_mlp": 1.01553965, + "epoch": 0.9879155894907714, + "flos": 18806777827200.0, + "grad_norm": 2.2153215112139883, + "language_loss": 0.78136319, + "learning_rate": 1.5168003058900757e-09, + "loss": 0.80346501, + "num_input_tokens_seen": 177323500, + "step": 8216, + "time_per_iteration": 2.6707756519317627 + }, + { + "auxiliary_loss_clip": 0.01194045, + "auxiliary_loss_mlp": 0.01025202, + "balance_loss_clip": 1.04292059, + "balance_loss_mlp": 1.01813936, + "epoch": 0.9880358323814105, + "flos": 22382044007040.0, + "grad_norm": 1.7992894581949619, + "language_loss": 0.91997027, + "learning_rate": 1.4866197197491715e-09, + "loss": 0.94216275, + "num_input_tokens_seen": 177342860, + "step": 8217, + "time_per_iteration": 2.697964906692505 + }, + { + "auxiliary_loss_clip": 0.01179339, + "auxiliary_loss_mlp": 0.01378945, + "balance_loss_clip": 1.04640055, + "balance_loss_mlp": 1.00021553, + "epoch": 0.9881560752720495, + "flos": 15668831733120.0, + "grad_norm": 6.321179891947377, + "language_loss": 0.79101908, + "learning_rate": 1.4567423064988371e-09, + "loss": 0.81660187, + "num_input_tokens_seen": 177360210, + "step": 8218, + "time_per_iteration": 3.5159947872161865 + }, + { + "auxiliary_loss_clip": 0.01169745, + "auxiliary_loss_mlp": 0.01024882, + "balance_loss_clip": 1.04780817, + "balance_loss_mlp": 1.01737237, + "epoch": 0.9882763181626887, + "flos": 21500113374720.0, + "grad_norm": 2.016771692223295, + "language_loss": 0.78179246, + "learning_rate": 1.4271680706718913e-09, + "loss": 0.80373877, + "num_input_tokens_seen": 177377885, + "step": 8219, + "time_per_iteration": 2.5777673721313477 + }, + { + "auxiliary_loss_clip": 0.0117765, + "auxiliary_loss_mlp": 0.01026409, + "balance_loss_clip": 1.0468204, + "balance_loss_mlp": 1.01836789, + "epoch": 0.9883965610533277, + "flos": 28034598551040.0, + "grad_norm": 1.794766049522343, + "language_loss": 0.82886177, + "learning_rate": 1.3978970167543013e-09, + "loss": 0.85090238, + "num_input_tokens_seen": 177398065, + "step": 8220, + "time_per_iteration": 2.6921868324279785 + }, + { + "auxiliary_loss_clip": 0.011707, + "auxiliary_loss_mlp": 0.01026664, + "balance_loss_clip": 1.04187655, + "balance_loss_mlp": 1.01922488, + "epoch": 0.9885168039439668, + "flos": 14098601710080.0, + "grad_norm": 2.5656493335245805, + "language_loss": 0.77608299, + "learning_rate": 1.3689291491867372e-09, + "loss": 0.7980566, + "num_input_tokens_seen": 177416380, + "step": 8221, + "time_per_iteration": 2.637373685836792 + }, + { + "auxiliary_loss_clip": 0.01169476, + "auxiliary_loss_mlp": 0.01020492, + "balance_loss_clip": 1.04691577, + "balance_loss_mlp": 1.01306224, + "epoch": 0.988637046834606, + "flos": 26432013352320.0, + "grad_norm": 2.03140540170762, + "language_loss": 0.73497587, + "learning_rate": 1.3402644723636836e-09, + "loss": 0.75687551, + "num_input_tokens_seen": 177438410, + "step": 8222, + "time_per_iteration": 3.5633809566497803 + }, + { + "auxiliary_loss_clip": 0.01181308, + "auxiliary_loss_mlp": 0.0102465, + "balance_loss_clip": 1.04730344, + "balance_loss_mlp": 1.01730108, + "epoch": 0.988757289725245, + "flos": 25229113764480.0, + "grad_norm": 2.0646172085982033, + "language_loss": 0.83584213, + "learning_rate": 1.311902990633218e-09, + "loss": 0.85790169, + "num_input_tokens_seen": 177457375, + "step": 8223, + "time_per_iteration": 2.7756118774414062 + }, + { + "auxiliary_loss_clip": 0.01171631, + "auxiliary_loss_mlp": 0.01025338, + "balance_loss_clip": 1.03760552, + "balance_loss_mlp": 1.0184598, + "epoch": 0.9888775326158841, + "flos": 26359042872960.0, + "grad_norm": 1.583326610945227, + "language_loss": 0.71618235, + "learning_rate": 1.2838447082978987e-09, + "loss": 0.73815203, + "num_input_tokens_seen": 177478530, + "step": 8224, + "time_per_iteration": 3.6498324871063232 + }, + { + "auxiliary_loss_clip": 0.01170569, + "auxiliary_loss_mlp": 0.01026269, + "balance_loss_clip": 1.04331923, + "balance_loss_mlp": 1.01893783, + "epoch": 0.9889977755065231, + "flos": 24316120846080.0, + "grad_norm": 2.3019457256021316, + "language_loss": 0.83280456, + "learning_rate": 1.2560896296143208e-09, + "loss": 0.85477293, + "num_input_tokens_seen": 177496995, + "step": 8225, + "time_per_iteration": 2.6434216499328613 + }, + { + "auxiliary_loss_clip": 0.0116624, + "auxiliary_loss_mlp": 0.01022897, + "balance_loss_clip": 1.04650187, + "balance_loss_mlp": 1.01580119, + "epoch": 0.9891180183971623, + "flos": 18951066760320.0, + "grad_norm": 4.9984383063192155, + "language_loss": 0.8212049, + "learning_rate": 1.2286377587926722e-09, + "loss": 0.84309626, + "num_input_tokens_seen": 177513785, + "step": 8226, + "time_per_iteration": 2.5264878273010254 + }, + { + "auxiliary_loss_clip": 0.01166396, + "auxiliary_loss_mlp": 0.01025632, + "balance_loss_clip": 1.04621911, + "balance_loss_mlp": 1.01842833, + "epoch": 0.9892382612878013, + "flos": 26176580760960.0, + "grad_norm": 2.011987730927754, + "language_loss": 0.74927878, + "learning_rate": 1.2014890999973992e-09, + "loss": 0.77119899, + "num_input_tokens_seen": 177530705, + "step": 8227, + "time_per_iteration": 2.6528573036193848 + }, + { + "auxiliary_loss_clip": 0.01164247, + "auxiliary_loss_mlp": 0.01024236, + "balance_loss_clip": 1.04453456, + "balance_loss_mlp": 1.01712823, + "epoch": 0.9893585041784404, + "flos": 25449605400960.0, + "grad_norm": 1.6559231927204283, + "language_loss": 0.78607208, + "learning_rate": 1.1746436573472073e-09, + "loss": 0.80795687, + "num_input_tokens_seen": 177552440, + "step": 8228, + "time_per_iteration": 2.6393063068389893 + }, + { + "auxiliary_loss_clip": 0.01191887, + "auxiliary_loss_mlp": 0.01028317, + "balance_loss_clip": 1.04426575, + "balance_loss_mlp": 1.02118802, + "epoch": 0.9894787470690796, + "flos": 20189302352640.0, + "grad_norm": 1.8842272268884908, + "language_loss": 0.69110227, + "learning_rate": 1.1481014349141726e-09, + "loss": 0.71330428, + "num_input_tokens_seen": 177569660, + "step": 8229, + "time_per_iteration": 2.698438882827759 + }, + { + "auxiliary_loss_clip": 0.01188656, + "auxiliary_loss_mlp": 0.01025459, + "balance_loss_clip": 1.04584944, + "balance_loss_mlp": 1.01756132, + "epoch": 0.9895989899597186, + "flos": 24644308435200.0, + "grad_norm": 2.0428386534318443, + "language_loss": 0.84374714, + "learning_rate": 1.121862436724852e-09, + "loss": 0.8658883, + "num_input_tokens_seen": 177588500, + "step": 8230, + "time_per_iteration": 2.681861639022827 + }, + { + "auxiliary_loss_clip": 0.01176984, + "auxiliary_loss_mlp": 0.01024433, + "balance_loss_clip": 1.04836583, + "balance_loss_mlp": 1.01712584, + "epoch": 0.9897192328503577, + "flos": 21799034357760.0, + "grad_norm": 1.9515867007854744, + "language_loss": 0.70486307, + "learning_rate": 1.0959266667598388e-09, + "loss": 0.72687721, + "num_input_tokens_seen": 177607315, + "step": 8231, + "time_per_iteration": 2.630969285964966 + }, + { + "auxiliary_loss_clip": 0.01195536, + "auxiliary_loss_mlp": 0.0102544, + "balance_loss_clip": 1.04362035, + "balance_loss_mlp": 1.01744676, + "epoch": 0.9898394757409968, + "flos": 21325229032320.0, + "grad_norm": 2.0509242926324203, + "language_loss": 0.74483669, + "learning_rate": 1.0702941289533196e-09, + "loss": 0.76704645, + "num_input_tokens_seen": 177625990, + "step": 8232, + "time_per_iteration": 2.691694498062134 + }, + { + "auxiliary_loss_clip": 0.01195523, + "auxiliary_loss_mlp": 0.0102109, + "balance_loss_clip": 1.04459012, + "balance_loss_mlp": 1.01426518, + "epoch": 0.9899597186316359, + "flos": 18545024442240.0, + "grad_norm": 1.9269571809742387, + "language_loss": 0.88811183, + "learning_rate": 1.0449648271939615e-09, + "loss": 0.91027796, + "num_input_tokens_seen": 177642335, + "step": 8233, + "time_per_iteration": 2.6951961517333984 + }, + { + "auxiliary_loss_clip": 0.0120757, + "auxiliary_loss_mlp": 0.01378261, + "balance_loss_clip": 1.04510498, + "balance_loss_mlp": 1.00017786, + "epoch": 0.990079961522275, + "flos": 23766723348480.0, + "grad_norm": 1.487565586559314, + "language_loss": 0.72743779, + "learning_rate": 1.0199387653240243e-09, + "loss": 0.75329614, + "num_input_tokens_seen": 177662025, + "step": 8234, + "time_per_iteration": 2.7644529342651367 + }, + { + "auxiliary_loss_clip": 0.01177421, + "auxiliary_loss_mlp": 0.01028595, + "balance_loss_clip": 1.04435766, + "balance_loss_mlp": 1.02105772, + "epoch": 0.9902002044129141, + "flos": 16399182971520.0, + "grad_norm": 1.9944529497779246, + "language_loss": 0.70931786, + "learning_rate": 9.952159471400267e-10, + "loss": 0.73137796, + "num_input_tokens_seen": 177679065, + "step": 8235, + "time_per_iteration": 2.601717948913574 + }, + { + "auxiliary_loss_clip": 0.01179164, + "auxiliary_loss_mlp": 0.01378472, + "balance_loss_clip": 1.04561949, + "balance_loss_mlp": 1.00018287, + "epoch": 0.9903204473035532, + "flos": 22559657783040.0, + "grad_norm": 1.9046313137839535, + "language_loss": 0.84377748, + "learning_rate": 9.707963763923022e-10, + "loss": 0.86935383, + "num_input_tokens_seen": 177698115, + "step": 8236, + "time_per_iteration": 2.613905429840088 + }, + { + "auxiliary_loss_clip": 0.01181039, + "auxiliary_loss_mlp": 0.01024884, + "balance_loss_clip": 1.04179168, + "balance_loss_mlp": 1.01776171, + "epoch": 0.9904406901941922, + "flos": 16144001775360.0, + "grad_norm": 1.8210725842030644, + "language_loss": 0.79109585, + "learning_rate": 9.466800567854427e-10, + "loss": 0.81315511, + "num_input_tokens_seen": 177716715, + "step": 8237, + "time_per_iteration": 2.6155741214752197 + }, + { + "auxiliary_loss_clip": 0.01192322, + "auxiliary_loss_mlp": 0.01026521, + "balance_loss_clip": 1.04062939, + "balance_loss_mlp": 1.01865292, + "epoch": 0.9905609330848314, + "flos": 26651499408000.0, + "grad_norm": 1.8877837810460636, + "language_loss": 0.68506551, + "learning_rate": 9.228669919778553e-10, + "loss": 0.70725393, + "num_input_tokens_seen": 177735640, + "step": 8238, + "time_per_iteration": 2.734128952026367 + }, + { + "auxiliary_loss_clip": 0.01177852, + "auxiliary_loss_mlp": 0.01024769, + "balance_loss_clip": 1.0420773, + "balance_loss_mlp": 1.01771212, + "epoch": 0.9906811759754705, + "flos": 23111820627840.0, + "grad_norm": 2.3911396030837526, + "language_loss": 0.79567921, + "learning_rate": 8.993571855817617e-10, + "loss": 0.81770551, + "num_input_tokens_seen": 177754470, + "step": 8239, + "time_per_iteration": 2.622840642929077 + }, + { + "auxiliary_loss_clip": 0.01175116, + "auxiliary_loss_mlp": 0.01019847, + "balance_loss_clip": 1.04439664, + "balance_loss_mlp": 1.01246476, + "epoch": 0.9908014188661095, + "flos": 22090593052800.0, + "grad_norm": 2.313181616439988, + "language_loss": 0.74704528, + "learning_rate": 8.761506411638642e-10, + "loss": 0.76899487, + "num_input_tokens_seen": 177773935, + "step": 8240, + "time_per_iteration": 2.64048171043396 + }, + { + "auxiliary_loss_clip": 0.01185549, + "auxiliary_loss_mlp": 0.01023794, + "balance_loss_clip": 1.04576874, + "balance_loss_mlp": 1.01646876, + "epoch": 0.9909216617567487, + "flos": 19242948677760.0, + "grad_norm": 1.7173958804730083, + "language_loss": 0.73999566, + "learning_rate": 8.53247362244236e-10, + "loss": 0.76208913, + "num_input_tokens_seen": 177792745, + "step": 8241, + "time_per_iteration": 3.552093982696533 + }, + { + "auxiliary_loss_clip": 0.01186283, + "auxiliary_loss_mlp": 0.01027033, + "balance_loss_clip": 1.0460186, + "balance_loss_mlp": 1.01974022, + "epoch": 0.9910419046473877, + "flos": 23621213352960.0, + "grad_norm": 1.6708826870611388, + "language_loss": 0.68133056, + "learning_rate": 8.306473522976532e-10, + "loss": 0.70346367, + "num_input_tokens_seen": 177812150, + "step": 8242, + "time_per_iteration": 2.666760206222534 + }, + { + "auxiliary_loss_clip": 0.01167728, + "auxiliary_loss_mlp": 0.01024993, + "balance_loss_clip": 1.04722118, + "balance_loss_mlp": 1.01789737, + "epoch": 0.9911621475380268, + "flos": 22711380831360.0, + "grad_norm": 2.360897566600639, + "language_loss": 0.71645641, + "learning_rate": 8.083506147522623e-10, + "loss": 0.73838365, + "num_input_tokens_seen": 177831545, + "step": 8243, + "time_per_iteration": 2.5995781421661377 + }, + { + "auxiliary_loss_clip": 0.01167656, + "auxiliary_loss_mlp": 0.01028765, + "balance_loss_clip": 1.0435909, + "balance_loss_mlp": 1.02125168, + "epoch": 0.991282390428666, + "flos": 13516956777600.0, + "grad_norm": 1.9519223464120183, + "language_loss": 0.85357082, + "learning_rate": 7.863571529906909e-10, + "loss": 0.87553501, + "num_input_tokens_seen": 177847130, + "step": 8244, + "time_per_iteration": 3.48496413230896 + }, + { + "auxiliary_loss_clip": 0.01073338, + "auxiliary_loss_mlp": 0.01000101, + "balance_loss_clip": 1.00838959, + "balance_loss_mlp": 0.99906427, + "epoch": 0.991402633319305, + "flos": 61830492071040.0, + "grad_norm": 0.7249273106038059, + "language_loss": 0.59655541, + "learning_rate": 7.646669703489372e-10, + "loss": 0.61728978, + "num_input_tokens_seen": 177911440, + "step": 8245, + "time_per_iteration": 3.3448646068573 + }, + { + "auxiliary_loss_clip": 0.01240257, + "auxiliary_loss_mlp": 0.01026766, + "balance_loss_clip": 1.0344286, + "balance_loss_mlp": 1.01932096, + "epoch": 0.9915228762099441, + "flos": 18770148933120.0, + "grad_norm": 1.9613010037592664, + "language_loss": 0.57268304, + "learning_rate": 7.432800701177023e-10, + "loss": 0.59535325, + "num_input_tokens_seen": 177929440, + "step": 8246, + "time_per_iteration": 3.1422152519226074 + }, + { + "auxiliary_loss_clip": 0.01088057, + "auxiliary_loss_mlp": 0.01000974, + "balance_loss_clip": 1.00979304, + "balance_loss_mlp": 0.99990159, + "epoch": 0.9916431191005832, + "flos": 65936660244480.0, + "grad_norm": 0.8027345547690571, + "language_loss": 0.57775629, + "learning_rate": 7.221964555415017e-10, + "loss": 0.59864664, + "num_input_tokens_seen": 177989100, + "step": 8247, + "time_per_iteration": 4.518578290939331 + }, + { + "auxiliary_loss_clip": 0.01184597, + "auxiliary_loss_mlp": 0.01018463, + "balance_loss_clip": 1.044909, + "balance_loss_mlp": 1.0118413, + "epoch": 0.9917633619912223, + "flos": 16581573256320.0, + "grad_norm": 2.366677842835947, + "language_loss": 0.74805063, + "learning_rate": 7.01416129818222e-10, + "loss": 0.77008122, + "num_input_tokens_seen": 178006720, + "step": 8248, + "time_per_iteration": 2.643998146057129 + }, + { + "auxiliary_loss_clip": 0.01204543, + "auxiliary_loss_mlp": 0.0102364, + "balance_loss_clip": 1.04435301, + "balance_loss_mlp": 1.01646948, + "epoch": 0.9918836048818613, + "flos": 25411108999680.0, + "grad_norm": 1.899493857499062, + "language_loss": 0.58701795, + "learning_rate": 6.809390961006745e-10, + "loss": 0.60929978, + "num_input_tokens_seen": 178026850, + "step": 8249, + "time_per_iteration": 2.729001522064209 + }, + { + "auxiliary_loss_clip": 0.01185107, + "auxiliary_loss_mlp": 0.01023053, + "balance_loss_clip": 1.04536843, + "balance_loss_mlp": 1.01555514, + "epoch": 0.9920038477725005, + "flos": 25046867134080.0, + "grad_norm": 1.842780682927369, + "language_loss": 0.68385839, + "learning_rate": 6.607653574948191e-10, + "loss": 0.70594007, + "num_input_tokens_seen": 178047630, + "step": 8250, + "time_per_iteration": 3.742576837539673 + }, + { + "auxiliary_loss_clip": 0.01165143, + "auxiliary_loss_mlp": 0.01022088, + "balance_loss_clip": 1.04165649, + "balance_loss_mlp": 1.01543641, + "epoch": 0.9921240906631396, + "flos": 21829773421440.0, + "grad_norm": 1.7045489344949225, + "language_loss": 0.81882221, + "learning_rate": 6.408949170613187e-10, + "loss": 0.84069449, + "num_input_tokens_seen": 178066895, + "step": 8251, + "time_per_iteration": 2.669550657272339 + }, + { + "auxiliary_loss_clip": 0.01181752, + "auxiliary_loss_mlp": 0.01025907, + "balance_loss_clip": 1.04316616, + "balance_loss_mlp": 1.01841509, + "epoch": 0.9922443335537786, + "flos": 24864225454080.0, + "grad_norm": 1.6107544897130797, + "language_loss": 0.81648731, + "learning_rate": 6.213277778144288e-10, + "loss": 0.83856392, + "num_input_tokens_seen": 178088540, + "step": 8252, + "time_per_iteration": 2.725538730621338 + }, + { + "auxiliary_loss_clip": 0.01209798, + "auxiliary_loss_mlp": 0.01022947, + "balance_loss_clip": 1.03802228, + "balance_loss_mlp": 1.0158, + "epoch": 0.9923645764444178, + "flos": 21613088626560.0, + "grad_norm": 2.402758093490357, + "language_loss": 0.66928005, + "learning_rate": 6.020639427224416e-10, + "loss": 0.69160748, + "num_input_tokens_seen": 178106185, + "step": 8253, + "time_per_iteration": 2.81846284866333 + }, + { + "auxiliary_loss_clip": 0.01183537, + "auxiliary_loss_mlp": 0.01025087, + "balance_loss_clip": 1.04408693, + "balance_loss_mlp": 1.01827455, + "epoch": 0.9924848193350568, + "flos": 25001798544000.0, + "grad_norm": 2.7539625778174535, + "language_loss": 0.72569144, + "learning_rate": 5.831034147076864e-10, + "loss": 0.7477777, + "num_input_tokens_seen": 178123435, + "step": 8254, + "time_per_iteration": 2.7074005603790283 + }, + { + "auxiliary_loss_clip": 0.01069552, + "auxiliary_loss_mlp": 0.00998488, + "balance_loss_clip": 1.0082444, + "balance_loss_mlp": 0.99742079, + "epoch": 0.9926050622256959, + "flos": 68912543151360.0, + "grad_norm": 0.7179933870594163, + "language_loss": 0.55727822, + "learning_rate": 5.644461966463065e-10, + "loss": 0.57795858, + "num_input_tokens_seen": 178191045, + "step": 8255, + "time_per_iteration": 3.318939208984375 + }, + { + "auxiliary_loss_clip": 0.01187247, + "auxiliary_loss_mlp": 0.01031055, + "balance_loss_clip": 1.04663503, + "balance_loss_mlp": 1.0237534, + "epoch": 0.9927253051163349, + "flos": 20923675914240.0, + "grad_norm": 2.3066741600623475, + "language_loss": 0.75692749, + "learning_rate": 5.460922913687049e-10, + "loss": 0.77911049, + "num_input_tokens_seen": 178210135, + "step": 8256, + "time_per_iteration": 2.7790403366088867 + }, + { + "auxiliary_loss_clip": 0.01199001, + "auxiliary_loss_mlp": 0.01378979, + "balance_loss_clip": 1.03924787, + "balance_loss_mlp": 1.00015092, + "epoch": 0.9928455480069741, + "flos": 22308211601280.0, + "grad_norm": 2.4098744672084322, + "language_loss": 0.75722617, + "learning_rate": 5.280417016593208e-10, + "loss": 0.78300595, + "num_input_tokens_seen": 178229925, + "step": 8257, + "time_per_iteration": 2.751054048538208 + }, + { + "auxiliary_loss_clip": 0.01174363, + "auxiliary_loss_mlp": 0.01378465, + "balance_loss_clip": 1.04852796, + "balance_loss_mlp": 1.0001142, + "epoch": 0.9929657908976132, + "flos": 17383889393280.0, + "grad_norm": 1.8481474185972397, + "language_loss": 0.74924314, + "learning_rate": 5.102944302559642e-10, + "loss": 0.77477139, + "num_input_tokens_seen": 178247420, + "step": 8258, + "time_per_iteration": 2.6344611644744873 + }, + { + "auxiliary_loss_clip": 0.01219662, + "auxiliary_loss_mlp": 0.01023596, + "balance_loss_clip": 1.04273129, + "balance_loss_mlp": 1.01553106, + "epoch": 0.9930860337882522, + "flos": 22674680110080.0, + "grad_norm": 2.2802693655036714, + "language_loss": 0.79351157, + "learning_rate": 4.9285047985137e-10, + "loss": 0.81594414, + "num_input_tokens_seen": 178266840, + "step": 8259, + "time_per_iteration": 2.88146710395813 + }, + { + "auxiliary_loss_clip": 0.01177891, + "auxiliary_loss_mlp": 0.01033121, + "balance_loss_clip": 1.04573488, + "balance_loss_mlp": 1.02540851, + "epoch": 0.9932062766788914, + "flos": 28147789284480.0, + "grad_norm": 2.2020092148286365, + "language_loss": 0.74485576, + "learning_rate": 4.757098530916436e-10, + "loss": 0.76696587, + "num_input_tokens_seen": 178287285, + "step": 8260, + "time_per_iteration": 2.8031110763549805 + }, + { + "auxiliary_loss_clip": 0.01178156, + "auxiliary_loss_mlp": 0.01030451, + "balance_loss_clip": 1.0473454, + "balance_loss_mlp": 1.02226126, + "epoch": 0.9933265195695304, + "flos": 20156659868160.0, + "grad_norm": 2.8953399744366233, + "language_loss": 0.7776438, + "learning_rate": 4.5887255257670563e-10, + "loss": 0.79972988, + "num_input_tokens_seen": 178304325, + "step": 8261, + "time_per_iteration": 2.6098265647888184 + }, + { + "auxiliary_loss_clip": 0.01166067, + "auxiliary_loss_mlp": 0.01022897, + "balance_loss_clip": 1.04573393, + "balance_loss_mlp": 1.01522589, + "epoch": 0.9934467624601695, + "flos": 21362037494400.0, + "grad_norm": 2.76444814279582, + "language_loss": 0.77194226, + "learning_rate": 4.4233858086117906e-10, + "loss": 0.79383194, + "num_input_tokens_seen": 178322850, + "step": 8262, + "time_per_iteration": 2.5879740715026855 + }, + { + "auxiliary_loss_clip": 0.012005, + "auxiliary_loss_mlp": 0.01025972, + "balance_loss_clip": 1.04468918, + "balance_loss_mlp": 1.01819968, + "epoch": 0.9935670053508087, + "flos": 19756040503680.0, + "grad_norm": 4.265685297257089, + "language_loss": 0.67527956, + "learning_rate": 4.261079404528356e-10, + "loss": 0.69754434, + "num_input_tokens_seen": 178342330, + "step": 8263, + "time_per_iteration": 2.688669204711914 + }, + { + "auxiliary_loss_clip": 0.01171573, + "auxiliary_loss_mlp": 0.01026142, + "balance_loss_clip": 1.04408681, + "balance_loss_mlp": 1.0187391, + "epoch": 0.9936872482414477, + "flos": 21978838863360.0, + "grad_norm": 2.6732375413362877, + "language_loss": 0.69085908, + "learning_rate": 4.1018063381437205e-10, + "loss": 0.71283621, + "num_input_tokens_seen": 178362715, + "step": 8264, + "time_per_iteration": 2.625208854675293 + }, + { + "auxiliary_loss_clip": 0.01068311, + "auxiliary_loss_mlp": 0.01001322, + "balance_loss_clip": 1.01013267, + "balance_loss_mlp": 1.00027275, + "epoch": 0.9938074911320868, + "flos": 69810667839360.0, + "grad_norm": 0.8655198892929505, + "language_loss": 0.61124349, + "learning_rate": 3.9455666336141167e-10, + "loss": 0.63193983, + "num_input_tokens_seen": 178426495, + "step": 8265, + "time_per_iteration": 3.252532958984375 + }, + { + "auxiliary_loss_clip": 0.01166015, + "auxiliary_loss_mlp": 0.01030345, + "balance_loss_clip": 1.04720902, + "balance_loss_mlp": 1.02338314, + "epoch": 0.9939277340227259, + "flos": 15084170058240.0, + "grad_norm": 7.941569229748052, + "language_loss": 0.83106542, + "learning_rate": 3.7923603146450267e-10, + "loss": 0.85302907, + "num_input_tokens_seen": 178442555, + "step": 8266, + "time_per_iteration": 2.540363311767578 + }, + { + "auxiliary_loss_clip": 0.01195501, + "auxiliary_loss_mlp": 0.01026105, + "balance_loss_clip": 1.0416553, + "balance_loss_mlp": 1.01883876, + "epoch": 0.994047976913365, + "flos": 17712364291200.0, + "grad_norm": 1.9945651358214165, + "language_loss": 0.80757391, + "learning_rate": 3.642187404473418e-10, + "loss": 0.82979, + "num_input_tokens_seen": 178460715, + "step": 8267, + "time_per_iteration": 3.593189001083374 + }, + { + "auxiliary_loss_clip": 0.01178313, + "auxiliary_loss_mlp": 0.01020948, + "balance_loss_clip": 1.04513121, + "balance_loss_mlp": 1.01430154, + "epoch": 0.994168219804004, + "flos": 19171558396800.0, + "grad_norm": 2.06263718158369, + "language_loss": 0.8614347, + "learning_rate": 3.495047925885508e-10, + "loss": 0.88342726, + "num_input_tokens_seen": 178479050, + "step": 8268, + "time_per_iteration": 2.68351411819458 + }, + { + "auxiliary_loss_clip": 0.01178839, + "auxiliary_loss_mlp": 0.01025939, + "balance_loss_clip": 1.04154766, + "balance_loss_mlp": 1.01866448, + "epoch": 0.9942884626946432, + "flos": 17851589406720.0, + "grad_norm": 2.6405266612618417, + "language_loss": 0.82653081, + "learning_rate": 3.350941901199e-10, + "loss": 0.84857857, + "num_input_tokens_seen": 178495970, + "step": 8269, + "time_per_iteration": 2.645376205444336 + }, + { + "auxiliary_loss_clip": 0.01187047, + "auxiliary_loss_mlp": 0.0102334, + "balance_loss_clip": 1.04430008, + "balance_loss_mlp": 1.01570129, + "epoch": 0.9944087055852823, + "flos": 18796578364800.0, + "grad_norm": 2.4164068068579154, + "language_loss": 0.83152288, + "learning_rate": 3.2098693522764066e-10, + "loss": 0.85362673, + "num_input_tokens_seen": 178509170, + "step": 8270, + "time_per_iteration": 3.5725903511047363 + }, + { + "auxiliary_loss_clip": 0.0119067, + "auxiliary_loss_mlp": 0.013786, + "balance_loss_clip": 1.04511595, + "balance_loss_mlp": 1.00018895, + "epoch": 0.9945289484759213, + "flos": 20996969616000.0, + "grad_norm": 1.8459027878825969, + "language_loss": 0.80989826, + "learning_rate": 3.071830300516165e-10, + "loss": 0.83559096, + "num_input_tokens_seen": 178527000, + "step": 8271, + "time_per_iteration": 2.6430280208587646 + }, + { + "auxiliary_loss_clip": 0.01182364, + "auxiliary_loss_mlp": 0.01025345, + "balance_loss_clip": 1.04661405, + "balance_loss_mlp": 1.01780248, + "epoch": 0.9946491913665605, + "flos": 14756952136320.0, + "grad_norm": 2.0938817794236795, + "language_loss": 0.70695853, + "learning_rate": 2.9368247668615234e-10, + "loss": 0.72903562, + "num_input_tokens_seen": 178545590, + "step": 8272, + "time_per_iteration": 2.601221799850464 + }, + { + "auxiliary_loss_clip": 0.01173112, + "auxiliary_loss_mlp": 0.01024614, + "balance_loss_clip": 1.05066466, + "balance_loss_mlp": 1.01659083, + "epoch": 0.9947694342571995, + "flos": 12669931186560.0, + "grad_norm": 2.5712238758048005, + "language_loss": 0.6138978, + "learning_rate": 2.804852771789434e-10, + "loss": 0.63587511, + "num_input_tokens_seen": 178558890, + "step": 8273, + "time_per_iteration": 3.4558651447296143 + }, + { + "auxiliary_loss_clip": 0.01167758, + "auxiliary_loss_mlp": 0.01025037, + "balance_loss_clip": 1.04775929, + "balance_loss_mlp": 1.01821852, + "epoch": 0.9948896771478386, + "flos": 18843442634880.0, + "grad_norm": 1.7481433168478, + "language_loss": 0.55651945, + "learning_rate": 2.675914335321661e-10, + "loss": 0.57844746, + "num_input_tokens_seen": 178577645, + "step": 8274, + "time_per_iteration": 2.572911262512207 + }, + { + "auxiliary_loss_clip": 0.01181519, + "auxiliary_loss_mlp": 0.01025282, + "balance_loss_clip": 1.04599321, + "balance_loss_mlp": 1.01707447, + "epoch": 0.9950099200384778, + "flos": 24900207903360.0, + "grad_norm": 2.481774690392874, + "language_loss": 0.79135966, + "learning_rate": 2.550009477018111e-10, + "loss": 0.81342763, + "num_input_tokens_seen": 178596415, + "step": 8275, + "time_per_iteration": 2.6198384761810303 + }, + { + "auxiliary_loss_clip": 0.01182861, + "auxiliary_loss_mlp": 0.01378789, + "balance_loss_clip": 1.04424131, + "balance_loss_mlp": 1.00019503, + "epoch": 0.9951301629291168, + "flos": 23733613987200.0, + "grad_norm": 2.327373631806293, + "language_loss": 0.62892377, + "learning_rate": 2.4271382159790634e-10, + "loss": 0.6545403, + "num_input_tokens_seen": 178613845, + "step": 8276, + "time_per_iteration": 3.5437428951263428 + }, + { + "auxiliary_loss_clip": 0.01224754, + "auxiliary_loss_mlp": 0.01025466, + "balance_loss_clip": 1.04296875, + "balance_loss_mlp": 1.0185461, + "epoch": 0.9952504058197559, + "flos": 22236893147520.0, + "grad_norm": 1.7734199359909772, + "language_loss": 0.85748255, + "learning_rate": 2.3073005708429406e-10, + "loss": 0.87998474, + "num_input_tokens_seen": 178633490, + "step": 8277, + "time_per_iteration": 2.771969795227051 + }, + { + "auxiliary_loss_clip": 0.01195423, + "auxiliary_loss_mlp": 0.01027267, + "balance_loss_clip": 1.04763007, + "balance_loss_mlp": 1.02075243, + "epoch": 0.995370648710395, + "flos": 21211032718080.0, + "grad_norm": 1.782614850606718, + "language_loss": 0.72086972, + "learning_rate": 2.190496559788535e-10, + "loss": 0.74309659, + "num_input_tokens_seen": 178651775, + "step": 8278, + "time_per_iteration": 2.805229663848877 + }, + { + "auxiliary_loss_clip": 0.01183889, + "auxiliary_loss_mlp": 0.01022697, + "balance_loss_clip": 1.0459367, + "balance_loss_mlp": 1.01519918, + "epoch": 0.9954908916010341, + "flos": 14866731077760.0, + "grad_norm": 2.402507805474652, + "language_loss": 0.76609498, + "learning_rate": 2.0767262005372265e-10, + "loss": 0.78816092, + "num_input_tokens_seen": 178669290, + "step": 8279, + "time_per_iteration": 2.6738874912261963 + }, + { + "auxiliary_loss_clip": 0.01200114, + "auxiliary_loss_mlp": 0.01027637, + "balance_loss_clip": 1.04301143, + "balance_loss_mlp": 1.0203886, + "epoch": 0.9956111344916732, + "flos": 19208259118080.0, + "grad_norm": 1.8799533876331396, + "language_loss": 0.75433218, + "learning_rate": 1.965989510346322e-10, + "loss": 0.77660966, + "num_input_tokens_seen": 178688410, + "step": 8280, + "time_per_iteration": 2.643495559692383 + }, + { + "auxiliary_loss_clip": 0.0119922, + "auxiliary_loss_mlp": 0.01028755, + "balance_loss_clip": 1.04264402, + "balance_loss_mlp": 1.02038634, + "epoch": 0.9957313773823123, + "flos": 20047060494720.0, + "grad_norm": 1.9056604045339693, + "language_loss": 0.7105844, + "learning_rate": 1.8582865060134955e-10, + "loss": 0.7328642, + "num_input_tokens_seen": 178706600, + "step": 8281, + "time_per_iteration": 2.7929728031158447 + }, + { + "auxiliary_loss_clip": 0.01059213, + "auxiliary_loss_mlp": 0.01002382, + "balance_loss_clip": 1.00819016, + "balance_loss_mlp": 1.00135124, + "epoch": 0.9958516202729514, + "flos": 57483253768320.0, + "grad_norm": 0.7839458015894087, + "language_loss": 0.55735171, + "learning_rate": 1.7536172038790098e-10, + "loss": 0.57796764, + "num_input_tokens_seen": 178766910, + "step": 8282, + "time_per_iteration": 3.2462246417999268 + }, + { + "auxiliary_loss_clip": 0.01185297, + "auxiliary_loss_mlp": 0.01026516, + "balance_loss_clip": 1.04459357, + "balance_loss_mlp": 1.01940823, + "epoch": 0.9959718631635904, + "flos": 27782900974080.0, + "grad_norm": 2.2582453595948104, + "language_loss": 0.69268513, + "learning_rate": 1.651981619819054e-10, + "loss": 0.71480328, + "num_input_tokens_seen": 178784060, + "step": 8283, + "time_per_iteration": 2.6857523918151855 + }, + { + "auxiliary_loss_clip": 0.01205429, + "auxiliary_loss_mlp": 0.01024437, + "balance_loss_clip": 1.04168344, + "balance_loss_mlp": 1.01740324, + "epoch": 0.9960921060542296, + "flos": 24024095274240.0, + "grad_norm": 2.2760796713918556, + "language_loss": 0.70764446, + "learning_rate": 1.5533797692546257e-10, + "loss": 0.72994316, + "num_input_tokens_seen": 178802795, + "step": 8284, + "time_per_iteration": 2.7749722003936768 + }, + { + "auxiliary_loss_clip": 0.01171136, + "auxiliary_loss_mlp": 0.010239, + "balance_loss_clip": 1.04326963, + "balance_loss_mlp": 1.01579952, + "epoch": 0.9962123489448687, + "flos": 18697393935360.0, + "grad_norm": 2.0853837829295547, + "language_loss": 0.84578395, + "learning_rate": 1.4578116671404296e-10, + "loss": 0.86773431, + "num_input_tokens_seen": 178821075, + "step": 8285, + "time_per_iteration": 2.610365390777588 + }, + { + "auxiliary_loss_clip": 0.01172589, + "auxiliary_loss_mlp": 0.01027638, + "balance_loss_clip": 1.04746127, + "balance_loss_mlp": 1.02044952, + "epoch": 0.9963325918355077, + "flos": 20010754823040.0, + "grad_norm": 2.240818409535412, + "language_loss": 0.71372676, + "learning_rate": 1.3652773279759777e-10, + "loss": 0.73572904, + "num_input_tokens_seen": 178837725, + "step": 8286, + "time_per_iteration": 2.630824565887451 + }, + { + "auxiliary_loss_clip": 0.01174425, + "auxiliary_loss_mlp": 0.0102694, + "balance_loss_clip": 1.04611921, + "balance_loss_mlp": 1.01969767, + "epoch": 0.9964528347261468, + "flos": 33108488991360.0, + "grad_norm": 1.6322087902586748, + "language_loss": 0.6275413, + "learning_rate": 1.2757767657989305e-10, + "loss": 0.64955497, + "num_input_tokens_seen": 178861515, + "step": 8287, + "time_per_iteration": 2.7020585536956787 + }, + { + "auxiliary_loss_clip": 0.01173635, + "auxiliary_loss_mlp": 0.01024187, + "balance_loss_clip": 1.04627407, + "balance_loss_mlp": 1.01665854, + "epoch": 0.9965730776167859, + "flos": 23109342589440.0, + "grad_norm": 1.8689881570181122, + "language_loss": 0.86872458, + "learning_rate": 1.1893099941850948e-10, + "loss": 0.89070284, + "num_input_tokens_seen": 178880410, + "step": 8288, + "time_per_iteration": 2.629955291748047 + }, + { + "auxiliary_loss_clip": 0.01187194, + "auxiliary_loss_mlp": 0.01023564, + "balance_loss_clip": 1.04278111, + "balance_loss_mlp": 1.01638126, + "epoch": 0.996693320507425, + "flos": 22965843755520.0, + "grad_norm": 2.4005354269295562, + "language_loss": 0.7742939, + "learning_rate": 1.105877026252866e-10, + "loss": 0.7964015, + "num_input_tokens_seen": 178898740, + "step": 8289, + "time_per_iteration": 2.6499722003936768 + }, + { + "auxiliary_loss_clip": 0.01169494, + "auxiliary_loss_mlp": 0.01026643, + "balance_loss_clip": 1.04776502, + "balance_loss_mlp": 1.01900184, + "epoch": 0.996813563398064, + "flos": 13222740476160.0, + "grad_norm": 2.160004844741276, + "language_loss": 0.7215246, + "learning_rate": 1.0254778746565663e-10, + "loss": 0.74348605, + "num_input_tokens_seen": 178914015, + "step": 8290, + "time_per_iteration": 2.6016628742218018 + }, + { + "auxiliary_loss_clip": 0.01194396, + "auxiliary_loss_mlp": 0.01025206, + "balance_loss_clip": 1.04435778, + "balance_loss_mlp": 1.01844406, + "epoch": 0.9969338062887032, + "flos": 14647855553280.0, + "grad_norm": 2.277071057232257, + "language_loss": 0.73414636, + "learning_rate": 9.481125515953259e-11, + "loss": 0.75634235, + "num_input_tokens_seen": 178932075, + "step": 8291, + "time_per_iteration": 2.617684841156006 + }, + { + "auxiliary_loss_clip": 0.01200955, + "auxiliary_loss_mlp": 0.01022001, + "balance_loss_clip": 1.03847289, + "balance_loss_mlp": 1.01470232, + "epoch": 0.9970540491793423, + "flos": 25735741142400.0, + "grad_norm": 1.7278324595469015, + "language_loss": 0.79692489, + "learning_rate": 8.737810688064228e-11, + "loss": 0.81915444, + "num_input_tokens_seen": 178951910, + "step": 8292, + "time_per_iteration": 3.787325859069824 + }, + { + "auxiliary_loss_clip": 0.0118537, + "auxiliary_loss_mlp": 0.01027278, + "balance_loss_clip": 1.04154634, + "balance_loss_mlp": 1.01909423, + "epoch": 0.9971742920699813, + "flos": 21470236237440.0, + "grad_norm": 2.0697871878595393, + "language_loss": 0.79228461, + "learning_rate": 8.024834375608414e-11, + "loss": 0.81441116, + "num_input_tokens_seen": 178970500, + "step": 8293, + "time_per_iteration": 2.644801378250122 + }, + { + "auxiliary_loss_clip": 0.01059484, + "auxiliary_loss_mlp": 0.01002013, + "balance_loss_clip": 1.0083915, + "balance_loss_mlp": 1.0009644, + "epoch": 0.9972945349606205, + "flos": 72211223629440.0, + "grad_norm": 0.8225267315661321, + "language_loss": 0.62761986, + "learning_rate": 7.342196686788149e-11, + "loss": 0.64823484, + "num_input_tokens_seen": 179023665, + "step": 8294, + "time_per_iteration": 3.0456864833831787 + }, + { + "auxiliary_loss_clip": 0.01183365, + "auxiliary_loss_mlp": 0.01027699, + "balance_loss_clip": 1.04792666, + "balance_loss_mlp": 1.0195272, + "epoch": 0.9974147778512595, + "flos": 19678293515520.0, + "grad_norm": 1.913401542465096, + "language_loss": 0.68966448, + "learning_rate": 6.689897725142834e-11, + "loss": 0.71177512, + "num_input_tokens_seen": 179043140, + "step": 8295, + "time_per_iteration": 2.663809061050415 + }, + { + "auxiliary_loss_clip": 0.01186106, + "auxiliary_loss_mlp": 0.01022197, + "balance_loss_clip": 1.04295206, + "balance_loss_mlp": 1.01524115, + "epoch": 0.9975350207418986, + "flos": 15960821391360.0, + "grad_norm": 5.026244878166953, + "language_loss": 0.88488495, + "learning_rate": 6.067937589615545e-11, + "loss": 0.906968, + "num_input_tokens_seen": 179061215, + "step": 8296, + "time_per_iteration": 3.537792444229126 + }, + { + "auxiliary_loss_clip": 0.01086762, + "auxiliary_loss_mlp": 0.01003329, + "balance_loss_clip": 1.00868273, + "balance_loss_mlp": 1.00225627, + "epoch": 0.9976552636325378, + "flos": 59961879768960.0, + "grad_norm": 0.7383981387727844, + "language_loss": 0.57669604, + "learning_rate": 5.476316374575241e-11, + "loss": 0.59759694, + "num_input_tokens_seen": 179124700, + "step": 8297, + "time_per_iteration": 3.179234743118286 + }, + { + "auxiliary_loss_clip": 0.01171026, + "auxiliary_loss_mlp": 0.01030744, + "balance_loss_clip": 1.04916191, + "balance_loss_mlp": 1.02298939, + "epoch": 0.9977755065231768, + "flos": 22487872452480.0, + "grad_norm": 2.051897458386495, + "language_loss": 0.724805, + "learning_rate": 4.9150341697723476e-11, + "loss": 0.74682271, + "num_input_tokens_seen": 179144590, + "step": 8298, + "time_per_iteration": 2.6084563732147217 + }, + { + "auxiliary_loss_clip": 0.01180849, + "auxiliary_loss_mlp": 0.01034874, + "balance_loss_clip": 1.04523814, + "balance_loss_mlp": 1.02739966, + "epoch": 0.9978957494138159, + "flos": 26030280666240.0, + "grad_norm": 1.5311529040275174, + "language_loss": 0.6667701, + "learning_rate": 4.384091060338768e-11, + "loss": 0.68892729, + "num_input_tokens_seen": 179165060, + "step": 8299, + "time_per_iteration": 3.609290838241577 + }, + { + "auxiliary_loss_clip": 0.01173728, + "auxiliary_loss_mlp": 0.0102096, + "balance_loss_clip": 1.04523361, + "balance_loss_mlp": 1.01383698, + "epoch": 0.998015992304455, + "flos": 22637835734400.0, + "grad_norm": 6.205142866100515, + "language_loss": 0.74214953, + "learning_rate": 3.883487126810081e-11, + "loss": 0.76409638, + "num_input_tokens_seen": 179184320, + "step": 8300, + "time_per_iteration": 2.757868528366089 + }, + { + "auxiliary_loss_clip": 0.01165026, + "auxiliary_loss_mlp": 0.01024668, + "balance_loss_clip": 1.04274964, + "balance_loss_mlp": 1.01710415, + "epoch": 0.9981362351950941, + "flos": 18223444955520.0, + "grad_norm": 3.7251224672189505, + "language_loss": 0.79073852, + "learning_rate": 3.41322244516995e-11, + "loss": 0.81263548, + "num_input_tokens_seen": 179202265, + "step": 8301, + "time_per_iteration": 3.5387606620788574 + }, + { + "auxiliary_loss_clip": 0.01191555, + "auxiliary_loss_mlp": 0.01023895, + "balance_loss_clip": 1.04198432, + "balance_loss_mlp": 1.01669455, + "epoch": 0.9982564780857331, + "flos": 33474095573760.0, + "grad_norm": 1.7863487681009464, + "language_loss": 0.6274386, + "learning_rate": 2.9732970866946925e-11, + "loss": 0.64959311, + "num_input_tokens_seen": 179222145, + "step": 8302, + "time_per_iteration": 2.809474468231201 + }, + { + "auxiliary_loss_clip": 0.01177486, + "auxiliary_loss_mlp": 0.01029822, + "balance_loss_clip": 1.0392108, + "balance_loss_mlp": 1.0215373, + "epoch": 0.9983767209763723, + "flos": 15523465392000.0, + "grad_norm": 3.3774602842663626, + "language_loss": 0.78184891, + "learning_rate": 2.563711118175327e-11, + "loss": 0.80392194, + "num_input_tokens_seen": 179239030, + "step": 8303, + "time_per_iteration": 2.7245497703552246 + }, + { + "auxiliary_loss_clip": 0.01189757, + "auxiliary_loss_mlp": 0.01025954, + "balance_loss_clip": 1.0447526, + "balance_loss_mlp": 1.01883435, + "epoch": 0.9984969638670114, + "flos": 19974377324160.0, + "grad_norm": 1.7864284518678066, + "language_loss": 0.8376568, + "learning_rate": 2.184464601717728e-11, + "loss": 0.85981393, + "num_input_tokens_seen": 179257345, + "step": 8304, + "time_per_iteration": 2.646822929382324 + }, + { + "auxiliary_loss_clip": 0.01181516, + "auxiliary_loss_mlp": 0.01030817, + "balance_loss_clip": 1.04846287, + "balance_loss_mlp": 1.02301514, + "epoch": 0.9986172067576504, + "flos": 20375750874240.0, + "grad_norm": 2.2932426033589155, + "language_loss": 0.77969414, + "learning_rate": 1.8355575948758585e-11, + "loss": 0.80181748, + "num_input_tokens_seen": 179275330, + "step": 8305, + "time_per_iteration": 2.617717742919922 + }, + { + "auxiliary_loss_clip": 0.01183859, + "auxiliary_loss_mlp": 0.01026793, + "balance_loss_clip": 1.042732, + "balance_loss_mlp": 1.01914549, + "epoch": 0.9987374496482896, + "flos": 23727903724800.0, + "grad_norm": 2.5220898279438315, + "language_loss": 0.7314226, + "learning_rate": 1.5169901505407424e-11, + "loss": 0.75352913, + "num_input_tokens_seen": 179292395, + "step": 8306, + "time_per_iteration": 2.6745007038116455 + }, + { + "auxiliary_loss_clip": 0.01184265, + "auxiliary_loss_mlp": 0.0102559, + "balance_loss_clip": 1.04520893, + "balance_loss_mlp": 1.01840448, + "epoch": 0.9988576925389286, + "flos": 25044029959680.0, + "grad_norm": 1.9371351860019097, + "language_loss": 0.74012625, + "learning_rate": 1.228762317073695e-11, + "loss": 0.76222479, + "num_input_tokens_seen": 179311225, + "step": 8307, + "time_per_iteration": 2.653017520904541 + }, + { + "auxiliary_loss_clip": 0.0118463, + "auxiliary_loss_mlp": 0.01024208, + "balance_loss_clip": 1.04541373, + "balance_loss_mlp": 1.0171392, + "epoch": 0.9989779354295677, + "flos": 31285627637760.0, + "grad_norm": 2.020134704570394, + "language_loss": 0.7877667, + "learning_rate": 9.70874138195299e-12, + "loss": 0.80985498, + "num_input_tokens_seen": 179333135, + "step": 8308, + "time_per_iteration": 2.7299511432647705 + }, + { + "auxiliary_loss_clip": 0.01166687, + "auxiliary_loss_mlp": 0.01021256, + "balance_loss_clip": 1.04517245, + "balance_loss_mlp": 1.0141573, + "epoch": 0.9990981783202069, + "flos": 19573398823680.0, + "grad_norm": 1.5641074853629167, + "language_loss": 0.74319386, + "learning_rate": 7.433256530076093e-12, + "loss": 0.7650733, + "num_input_tokens_seen": 179353090, + "step": 8309, + "time_per_iteration": 2.63389253616333 + }, + { + "auxiliary_loss_clip": 0.01203914, + "auxiliary_loss_mlp": 0.01021131, + "balance_loss_clip": 1.03832328, + "balance_loss_mlp": 1.01457405, + "epoch": 0.9992184212108459, + "flos": 17199667514880.0, + "grad_norm": 2.294059374309664, + "language_loss": 0.75685048, + "learning_rate": 5.46116896038562e-12, + "loss": 0.77910095, + "num_input_tokens_seen": 179367500, + "step": 8310, + "time_per_iteration": 2.655848979949951 + }, + { + "auxiliary_loss_clip": 0.01182778, + "auxiliary_loss_mlp": 0.01026746, + "balance_loss_clip": 1.044595, + "balance_loss_mlp": 1.01935506, + "epoch": 0.999338664101485, + "flos": 46497853681920.0, + "grad_norm": 2.6447949837675093, + "language_loss": 0.62149072, + "learning_rate": 3.792478972197699e-12, + "loss": 0.64358592, + "num_input_tokens_seen": 179388085, + "step": 8311, + "time_per_iteration": 2.855036735534668 + }, + { + "auxiliary_loss_clip": 0.01165755, + "auxiliary_loss_mlp": 0.0102335, + "balance_loss_clip": 1.04555464, + "balance_loss_mlp": 1.01648986, + "epoch": 0.9994589069921241, + "flos": 15158253859200.0, + "grad_norm": 2.7992954305804365, + "language_loss": 0.69761103, + "learning_rate": 2.4271868181990895e-12, + "loss": 0.71950209, + "num_input_tokens_seen": 179405250, + "step": 8312, + "time_per_iteration": 2.597712755203247 + }, + { + "auxiliary_loss_clip": 0.01176836, + "auxiliary_loss_mlp": 0.01023986, + "balance_loss_clip": 1.04436302, + "balance_loss_mlp": 1.01692331, + "epoch": 0.9995791498827632, + "flos": 12531460256640.0, + "grad_norm": 2.002271588376986, + "language_loss": 0.81167215, + "learning_rate": 1.3652927060014973e-12, + "loss": 0.83368039, + "num_input_tokens_seen": 179420845, + "step": 8313, + "time_per_iteration": 2.61883544921875 + }, + { + "auxiliary_loss_clip": 0.01194745, + "auxiliary_loss_mlp": 0.01021458, + "balance_loss_clip": 1.04284143, + "balance_loss_mlp": 1.01463962, + "epoch": 0.9996993927734023, + "flos": 19245175320960.0, + "grad_norm": 3.745135157331051, + "language_loss": 0.63700694, + "learning_rate": 6.067967965872612e-13, + "loss": 0.65916896, + "num_input_tokens_seen": 179440455, + "step": 8314, + "time_per_iteration": 2.670163869857788 + }, + { + "auxiliary_loss_clip": 0.01194595, + "auxiliary_loss_mlp": 0.01025355, + "balance_loss_clip": 1.04551172, + "balance_loss_mlp": 1.01816702, + "epoch": 0.9998196356640414, + "flos": 62952804518400.0, + "grad_norm": 3.125619007478784, + "language_loss": 0.76896489, + "learning_rate": 1.5169920497548615e-13, + "loss": 0.7911644, + "num_input_tokens_seen": 179465075, + "step": 8315, + "time_per_iteration": 3.0748207569122314 + }, + { + "auxiliary_loss_clip": 0.01125555, + "auxiliary_loss_mlp": 0.01013576, + "balance_loss_clip": 1.02518666, + "balance_loss_mlp": 1.00912786, + "epoch": 0.9999398785546805, + "flos": 50922375073920.0, + "grad_norm": 1.2667324894521959, + "language_loss": 0.550879, + "learning_rate": 0.0, + "loss": 0.57227039, + "num_input_tokens_seen": 179513955, + "step": 8316, + "time_per_iteration": 3.160606861114502 + }, + { + "epoch": 0.9999398785546805, + "num_input_tokens_seen": 179513955, + "step": 8316, + "total_flos": 6.996749092776837e+17, + "train_loss": 0.7900429457827449, + "train_runtime": 25339.6831, + "train_samples_per_second": 13.128, + "train_steps_per_second": 0.328 + } + ], + "logging_steps": 1.0, + "max_steps": 8316, + "num_input_tokens_seen": 179513955, + "num_train_epochs": 1, + "save_steps": 1664, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6.996749092776837e+17, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +} diff --git a/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/training_args.bin b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d7f731d29d10a6c61aab6032e62a2fbb94916cc0 --- /dev/null +++ b/CompeteSMoE/competesmoe_versions/Softplus_diversity_3loss_downrouter_competesmoe/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08e3b31b2e43867fda701cb65885c9d6cff09128141cbe9810e04a29a24d10b8 +size 7992